Carbon footprint intelligent matching and rapid accounting system and method based on BOM table

Abstract

The present application relates to a kind of carbon footprint intelligent matching and fast accounting system and method based on BOM table, the system includes BOM table extension data correlation unit, intelligent matching unit, missing data intelligent replacement unit, whole-process fast calculation unit and carbon footprint and cost comprehensive output unit, by automatically parsing BOM table and calling enterprise interface to obtain supply chain, cost and energy consumption data, realize data automatic acquisition and extension;Adopt natural language processing technology to standardize and intelligently match material name, combine multi-dimensional pedigree matrix to filter optimal emission factor;For unmatched data, intelligent replacement is carried out based on rule base and quantification uncertainty;Adopt parallel computing engine to realize whole life cycle carbon emission fast accounting, and the result with confidence interval is output through Monte Carlo simulation, finally, carbon cost efficiency index, stage contribution degree and supplier comprehensive evaluation report are generated by fusing cost data, significantly improve the degree of automation, calculation efficiency and result reliability of accounting.

Description

Technical Field

[0001] This invention relates to the fields of green supply chain management and product carbon emission accounting technology, and in particular to a system and method for intelligent matching and rapid calculation of carbon footprint based on BOM (Bill of Materials). Background Technology

[0002] Accurate and efficient calculation of product carbon footprints has become a core requirement for enterprises to fulfill their environmental responsibilities, address green trade barriers, and optimize supply chain management. Currently, the industry mainly relies on Life Cycle Assessment (LCA) methodologies and specialized software (such as SimaPro, GaBi, and OpenLCA) for calculations. However, existing technologies exhibit the following systemic flaws when addressing enterprise-level, large-scale, and high-frequency calculation needs, severely limiting their practical application effectiveness: a. Lack of automation and standardization in data collection and integration: Existing methods rely heavily on manual labor, manually collecting material, energy consumption, and supply chain logistics data from the enterprise's fragmented systems. This process is cumbersome, error-prone, and inefficient, making it difficult to guarantee the integrity and timeliness of the data source.

[0003] b. Insufficient intelligence and accuracy in carbon emission factor matching: Faced with massive and heterogeneous carbon emission factor databases, existing technologies lack intelligent data association capabilities. Firstly, they cannot effectively handle the diversity and non-standardization of material names (such as a mixture of synonyms, abbreviations, and specifications), leading to matching failures or errors. Secondly, when the same material name corresponds to multiple factors (due to differences in region, process, and year), there is a lack of objective and quantitative automatic screening mechanisms, relying on subjective human judgment, which is inefficient and inconsistent.

[0004] c. Inadequate mechanisms for managing missing data and uncertainty: When specific material factors are missing from the core database, existing methods typically resort to arbitrary approaches such as manually finding approximate values, subjectively adding them, or simply ignoring them. This not only undermines the integrity of the accounting model but also, due to the lack of unified substitution rules and uncertainty quantification mechanisms, results in low reliability, poor traceability, and an inability to meet international standards for data quality.

[0005] d. Fragmentation of accounting process and computing performance bottleneck: Traditional LCA software is complex to operate, requiring manual import and association of data between different modules and serial execution of calculations. The process is fragmented and cannot achieve one-stop rapid output. For products with complex structures, the calculation time is long and cannot support business scenarios such as rapid product design iteration, multi-solution comparison or large-scale supply chain screening.

[0006] e. Lack of integrated decision support for environmental performance and economic benefits: The output of existing tools is usually limited to single-dimensional carbon emission reports and fails to be deeply integrated and correlated with product cost data; enterprises cannot intuitively identify key links with high carbon emissions and high costs, nor can they conduct comprehensive evaluation of suppliers in terms of both cost and carbon emissions, resulting in a disconnect between carbon management decisions and economic benefit objectives, making it difficult to drive substantial green supply chain optimization and cost reduction and emission reduction.

[0007] Therefore, there is an urgent need for a system and method that can span the entire product lifecycle, achieve automatic data association and intelligent matching, support rapid and accurate accounting, and provide integrated carbon cost decision support to address the pressing challenges currently facing enterprises. Summary of the Invention

[0008] The technical problem to be solved by this invention is to provide a carbon footprint intelligent matching and rapid calculation system based on a Bill of Materials (BOM) in order to address the problems of manual data collection, subjective matching process, arbitrary handling of missing data, slow calculation process, and single decision-making dimension in the existing technologies mentioned above.

[0009] The technical solution adopted by this invention to solve its technical problem is: a carbon footprint intelligent matching and rapid calculation system based on BOM (Bill of Materials), comprising: The BOM table extended data association unit is used to parse the product bill of materials file and automatically obtain the supply chain information, cost information and energy-consuming equipment binding information of the materials by calling the enterprise system's application programming interface to form an extended bill of materials dataset. The intelligent matching unit, connected to the BOM table extended data association unit, is used to standardize the material name through natural language processing. When the matching confidence of the material name with the carbon emission factor database is lower than a preset threshold, the optimal carbon emission factor is selected based on a weighted score of multiple preset dimensions. The missing data intelligent substitution unit is connected to the intelligent matching unit and is used to select candidate factors from the substitution factor library for materials that have not been successfully matched, according to preset rules, and calculate the corresponding uncertainty coefficient. The end-to-end rapid calculation unit is connected to the missing data intelligent replacement unit to calculate the carbon emissions of each stage of the product's entire life cycle in parallel, and performs uncertainty propagation calculation through Monte Carlo simulation to output carbon footprint results with confidence intervals. The carbon footprint and cost integrated output unit is connected to the full-process rapid calculation unit to integrate carbon emission results and cost data, calculate the carbon cost efficiency index, stage contribution comparison and supplier comprehensive evaluation, and generate a comprehensive analysis report.

[0010] By combining and sequentially connecting the BOM table extended data association unit, intelligent matching unit, missing data intelligent replacement unit, full-process rapid calculation unit, and carbon footprint and cost comprehensive output unit, a complete carbon footprint accounting closed-loop system is formed, realizing automated, rapid, and integrated carbon footprint accounting.

[0011] Furthermore, the BOM table extended data association unit includes: The document parsing and standardization module is used to parse bill of materials (BOM) files and extract key fields. The automatic material details acquisition module is used to obtain detailed material specifications and purchase prices by calling the interface of the enterprise resource planning system based on the material code. The energy and equipment binding module is used to query preset energy-equipment mapping rules and obtain equipment energy consumption data from the manufacturing execution system; The supply chain information extension module is used to call the supplier relationship management system interface to obtain supplier coordinates and call the geographic information system service interface to calculate transportation distance.

[0012] Through four modules—file parsing, detailed data collection, energy binding, and supply chain expansion—data is automatically and structurally extracted and integrated from the enterprise's multi-source heterogeneous systems. This transforms tedious manual work into automated system calls, ensuring the integrity and consistency of data sources and preparing for accurate accounting in the future.

[0013] Furthermore, the intelligent matching unit includes: The text preprocessing module is used to perform text cleaning, word segmentation, and stop word removal on material names; The thesaurus mapping module is used to query a pre-built thesaurus and map non-standard terms to standard terms; The composite similarity calculation module is used to calculate string similarity scores. and semantic similarity score ; The genealogy matrix filtering module is used to perform weighted scoring based on multiple preset dimensions when the matching confidence is lower than the threshold.

[0014] The intelligent matching process is defined to include four major steps: preprocessing, synonym mapping, composite calculation, and genealogy screening. A standardized processing flow from the original name to the selection of the optimal factor is established.

[0015] Furthermore, the genealogy matrix screening module calculates the comprehensive score of candidate factors based on the following formula. : , in, Score the similarity between names. Score for the authority of the data source. The score is based on the degree of regional suitability. Score for time fit. The production process matching score is used to determine the production process matching score. , , , , The preset weights are for each dimension.

[0016] When the confidence of simple matching is insufficient, the algorithm objectively and quantitatively scores the candidate factors from five dimensions: name, data source, region, time, and process. This transforms subjective selection that relies on human experience into automated and interpretable objective decision-making based on a multi-dimensional rule model.

[0017] Furthermore, the name similarity score The composite similarity calculation module calculates the similarity using the following formula: , Where α and β are preset weighting coefficients, and α+β=1.

[0018] This ensures that the system can effectively handle differences in descriptions such as "Φ12 steel bar" and "hot-rolled ribbed steel bar", resolving the problem of matching failures caused by the diversity and non-standardization of material names.

[0019] Furthermore, the missing data intelligent replacement unit includes: The missing item detection module is used to identify material records that are not matched. The candidate retrieval module is used to query the substitution rule database and retrieve candidate substitution factors based on material attributes; The similarity assessment module is used to calculate a comprehensive similarity score for candidate substitutes. ; The uncertainty quantification module is used to calculate the uncertainty coefficient U according to the following formula: , in, γ and δ are preset weighting coefficients used to assess the quality of alternative factor data sources.

[0020] Through modules for missing data detection, candidate retrieval, similarity assessment, and uncertainty quantification, the system finds the most suitable replacement factor based on a preset rule base. An uncertainty coefficient is assigned to the replacement result using a formula, ensuring the continuity of the accounting process, the traceability of the results, and the scientific rigor, thus meeting high standards of data quality requirements.

[0021] Furthermore, the end-to-end high-speed computing unit includes: A parallel computing engine is used to perform the following computations synchronously: Carbon emissions during the raw material acquisition stage Where n is the total number of materials. Let i be the amount of material i used. Let be the carbon emission factor of material i; Carbon emissions during transportation Where m is the number of materials with transportation information. Let j be the weight of material j. For its transportation distance, The carbon emission factor of its transportation method; Carbon emissions during the production phase Where p is the total number of energy types, Let k be the amount of energy consumed. , where k is the carbon emission factor for energy. The uncertainty propagation module is used to calculate the total carbon footprint using Monte Carlo simulation based on the uncertainty coefficients of each input data. The confidence interval.

[0022] The parallel computing engine can simultaneously calculate carbon emissions in the three stages of raw materials, transportation, and production. It also has an uncertainty propagation module to perform Monte Carlo simulations of uncertainties in the preceding stages. Through parallelization technology, it greatly accelerates the traditional serial and time-consuming calculation process and outputs results with confidence intervals, achieving rapid and scientific quantification of carbon footprint.

[0023] Furthermore, the integrated carbon footprint and cost output unit includes: The cost calculation and allocation module is used to calculate material costs, energy costs, transportation costs, and total product production costs based on material usage, purchase price, energy consumption, and transportation rates. ; The comprehensive evaluation index calculation module is used to calculate the carbon cost efficiency index. And the proportion of carbon emissions at each stage of the life cycle. and cost percentage ,in and These represent the emissions and costs of stage s, respectively. The supplier evaluation module is used to aggregate the carbon emissions and procurement costs of suppliers to generate comprehensive supplier evaluation data. The visualization report generation module is used to render carbon footprint details, cost details, and comprehensive evaluation indicators into a comprehensive analysis report with data traceability capabilities.

[0024] The carbon footprint and cost integrated output unit collects costs, generates a carbon cost efficiency index and a stage contribution comparison through the comprehensive evaluation index calculation module, and outputs decision insights using the supplier evaluation module and the visualization report generation module. This upgrades the tool from a simple accounting reporter to a strategic decision-making system that supports green procurement, process optimization and supply chain management.

[0025] It also provides a method for intelligent matching and rapid calculation of carbon footprint based on BOM (Bill of Materials), including: S1. Parse the product bill of materials file and automatically obtain the supply chain, cost and energy consumption information of the materials by calling the enterprise system interface to form an extended bill of materials dataset; S2. Standardize the material name. When the confidence level of the material name matching the carbon emission factor database is lower than the preset threshold, select the optimal carbon emission factor based on a weighted score of multiple preset dimensions. S3. For materials that are not successfully matched, select candidate factors from the alternative factor library according to preset rules for substitution, and calculate the corresponding uncertainty coefficient. S4. Perform parallel calculations on carbon emissions at each stage of the product's entire lifecycle, and use Monte Carlo simulation to perform uncertainty propagation calculations, outputting carbon footprint results with confidence intervals. S5 integrates carbon emission results and cost data, calculates the carbon cost efficiency index, compares stage contributions, and provides a comprehensive evaluation of suppliers, generating a comprehensive analysis report.

[0026] Furthermore, step S1 specifically includes: S101. Parse the product bill of materials file, extract the material name, material code, usage and unit of measurement fields, and convert them into standard data objects for internal system use. S102. Based on the material code, call the application programming interface of the enterprise's internal resource planning system or manufacturing execution system to obtain the detailed specifications and purchase price of the material. S103. Based on the energy entries in the bill of materials, query the preset energy-equipment mapping rules, and obtain the production energy consumption time series data of the corresponding equipment from the manufacturing execution system, and bind the energy consumption to the specific production equipment. S104. Based on the material code, call the interface of the supplier relationship management system to obtain the main supplier information and coordinates, and call the geographic information system service interface to calculate the transportation distance from the supplier to the production site. S105. Based on the standardized material and energy names, query the local or external carbon emission factor database to obtain a preliminary list of carbon emission data entries. S106. Provide a data verification interface for users to review and revise automatically associated data, and persistently store the final data in an extended bill of materials data table.

[0027] Furthermore, step S2 specifically includes: S201. Perform text cleaning, case unification, word segmentation, and stop word removal on the original names of materials or energy to obtain a core name word sequence; S202. Query the pre-built thesaurus and map the core name word sequence to standard terms; S203. Perform a composite similarity calculation between the standard terminology and the entry names in the carbon emission factor database to obtain the string similarity score. and semantic similarity score ; S204. Determine whether the highest similarity score has reached the preset automatic matching threshold. If not, trigger the lineage matrix filtering based on name similarity score. Data source authority score Regional adaptability score Time fit score Matching score with production process The overall score for each candidate factor is calculated as follows: And select the factor with the highest score: , in, , , , , , The preset weights are for each dimension.

[0028] Furthermore, step S3 specifically includes: S301. Scan the extended bill of materials data table, identify material records with a status of unmatched, and create alternative processing tasks; S302. Based on the properties of the material, access the substitution rule database and retrieve candidate substitution factors according to the preset substitution priority rules. S303. For each candidate substitution factor, a multi-dimensional weighted evaluation is performed based on industry similarity, geographical similarity, technological similarity, and name semantic similarity to calculate a comprehensive similarity score. ; S304. Select the candidate replacement factor with the highest comprehensive similarity score as the replacement value, and then base the replacement value on its score. and data source quality rating The uncertainty coefficient U is calculated using the following formula: , Where γ and δ are preset weighting coefficients; S305. Write the substitution value and uncertainty coefficient into the extended bill of materials data table, and record the substitution decision log.

[0029] Furthermore, step S4 specifically includes: S401. Responding to the user's calculation request, load the specified version of the extended bill of materials dataset that has undergone data processing from the database; S402. Using a parallel computing framework, the following calculations are executed synchronously: Carbon emissions during the raw material acquisition stage Where n is the total number of materials. Let i be the amount of material i used. Let be the carbon emission factor of material i; Carbon emissions during transportation Where m is the number of materials with transportation information. Let j be the weight of material j. For its transportation distance, The carbon emission factor of its transportation method; Carbon emissions during the production phase Where p is the total number of energy types, Let k be the amount of energy consumed. , where k is the carbon emission factor for energy. S403. Based on the uncertainty coefficients of each input data, Monte Carlo simulation is used to calculate the uncertainty propagation and obtain the total carbon footprint. The probability distribution and confidence interval; S404. Summarize the results of each stage and generate a result object that includes total emissions, emissions in each stage, uncertainty range and data quality identifier; S405. Persistently store the result object and automatically generate a visual carbon footprint calculation report.

[0030] Furthermore, step S5 specifically includes: S501. Align and correlate the carbon footprint results output by the full-process rapid calculation unit with the material cost, energy cost, and transportation cost data obtained by the BOM extended data association unit. S502, Calculate the total production cost of the product

[0031] S503, Calculate the carbon cost efficiency index Calculate the carbon emission share at each stage of the life cycle. and cost percentage ,in, and These represent the emissions and costs of stage s, respectively. By aggregating the carbon emissions and procurement costs of suppliers, a comprehensive supplier evaluation chart is generated. S504: Call the report engine to render carbon footprint details, cost details, carbon cost efficiency index, stage contribution comparison, and supplier comprehensive evaluation into an interactive or static comprehensive analysis report. S505. Store the generated report files and user operation audit logs.

[0032] An electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, when the processor executes the program, it implements the carbon footprint intelligent matching and rapid calculation method based on the BOM table as described in the above-described scheme.

[0033] A computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements the carbon footprint intelligent matching and rapid calculation method based on the BOM table as described above.

[0034] The beneficial effects of this invention are: This invention achieves automatic collection of material, supply chain, and cost data through automatic BOM parsing and multi-source system data expansion technology; through a parallelized full-process fast calculation engine, the accounting of each stage of the life cycle is changed from serial to concurrent execution; the entire process from data input to report output can be completed within minutes, and the overall efficiency is more than 100 times higher than that of traditional methods. This invention introduces an intelligent matching unit that integrates natural language processing. It solves the problem of material name diversity through text cleaning and synonym mapping. Based on a weighted scoring model of a multi-dimensional genealogy matrix, it systematically incorporates key quality dimensions such as data source authority, regional suitability, time validity, and production process matching degree in addition to name similarity for comprehensive evaluation. This model transforms the factor selection decision-making process from an operation that relies on personal experience to an automated and interpretable process based on clear rules and quantitative algorithms. This improves the overall matching accuracy by about 12% to 15% compared to the traditional manual method, and greatly enhances the reliability of the calculation results. This invention utilizes a rule-driven intelligent replacement unit for missing data to automatically retrieve and select the optimal replacement factor based on multi-dimensional similarity in industry, region, and technology. It also pioneered the use of a quantitative model to calculate and label the uncertainty coefficient of each replacement operation. Combined with Monte Carlo simulation in the entire process calculation, it achieves the full-link propagation and quantification of uncertainty from data input to the final carbon footprint result. The final output is a carbon footprint report with confidence intervals, which meets international standards for data quality. This invention, through a carbon footprint and cost integrated output unit, automatically integrates and calculates material, energy, and logistics costs while completing carbon accounting, thereby generating a carbon cost efficiency index and a phased carbon-cost contribution comparison matrix. This enables enterprises to accurately identify the dual hotspots of high carbon emissions and high costs. At the same time, the system can automatically generate a supplier cost-carbon emission integrated evaluation quadrant chart, intuitively supporting green procurement decisions. This invention records detailed decision-making paths, scoring criteria, and data sources for both automatic matching and substitution operations, and provides a user-friendly interface for manual review and intervention, fully meeting the transparency requirements of internal auditing and external certification. It supports the import of BOM files in multiple formats and can flexibly access carbon emission factor databases from different sources. Its matching weights and substitution rules can be continuously optimized and adaptively adjusted based on historical data and user feedback through machine learning mechanisms, ensuring that the system can adapt to the accounting characteristics of different industries. Attached Figure Description

[0035] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0036] Figure 1 This is the overall architecture diagram of the system of the present invention.

[0037] Figure 2 This is a structural diagram of the extended data association unit of the BOM table in the system of this invention.

[0038] Figure 3 This is a structural diagram of the intelligent matching unit in the system of the present invention.

[0039] Figure 4 This is a structural diagram of the intelligent data replacement unit in the system of this invention.

[0040] Figure 5 This is a flowchart of the full-process fast calculation unit in the system of this invention.

[0041] Figure 6 This is a structural diagram of the carbon footprint and cost integrated output unit in the system of this invention.

[0042] In the diagram: 100, BOM table extended data association unit; 200, intelligent matching unit; 300, missing data intelligent replacement unit; 400, full-process rapid calculation unit; 500, carbon footprint and cost comprehensive output unit; 101, document parsing and standardization module; 102, automatic material details collection module; 103, energy and equipment binding module; 104, supply chain information extension module; 105, manual verification and persistence module; 201, text preprocessing module; 202, synonym mapping module; 203, composite similarity calculation module; 204, genealogy matrix screening module; 301, missing data detection module; 302, candidate retrieval module; 303, similarity assessment module; 304, uncertainty quantification module; 501, cost calculation and allocation module; 502, comprehensive evaluation index calculation module; 503, visualization report generation module; 504, data storage and management module. Detailed Implementation

[0043] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0044] like Figure 1As shown, a carbon footprint intelligent matching and rapid calculation system based on a Bill of Materials (BOM) includes a BOM extended data association unit 100, an intelligent matching unit 200, a missing data intelligent replacement unit 300, a full-process rapid calculation unit 400, and a carbon footprint and cost comprehensive output unit 500. The BOM extended data association unit 100 receives product BOM files uploaded by users and acts as a data hub, communicating with the enterprise's resource planning system (ERP), manufacturing execution system (MES), supplier relationship management system (SRM), and geographic information system (GIS) to automatically capture and integrate an extended bill of materials dataset rich in material, cost, supplier, transportation, and energy consumption information. The extended bill of materials dataset flows into the intelligent matching unit 200, which is responsible for associating material names with built-in or external carbon emission factor databases. First, the names are cleaned and standardized. If the confidence of a direct match is high enough, it is automatically selected; otherwise, its core phylogenetic matrix screening algorithm is activated to comprehensively evaluate candidate factors from multiple dimensions, ultimately objectively selecting the optimal match and recording the entire decision-making process. For material items still marked as "unmatched" after processing by the intelligent matching unit 200, the system will trigger the missing data intelligent substitution unit 300. This unit automatically retrieves and selects the most similar substitution factor based on a preset rule base, and calculates an uncertainty coefficient for this substitution to quantify the additional error risk brought by the substitution data. After completing the matching and substitution of all data, the system enters the full-process rapid calculation unit 400. This unit employs parallel computing technology to simultaneously calculate the carbon emissions of the product at each stage of its life cycle, including raw material acquisition, transportation, and production. More importantly, the full-process rapid calculation unit 400 gathers the uncertainty information of all data points from the preceding steps, performs probability propagation through Monte Carlo simulation, and ultimately outputs a total carbon footprint result with a confidence interval, rather than just a single value. Finally, the Carbon Footprint and Cost Integrated Output Unit 500 integrates environmental and economic data, utilizes cost information obtained from the BOM extension unit to calculate various costs such as materials, energy, and transportation, and combines them with carbon footprint results to generate advanced analytical indicators such as the Carbon Cost Efficiency Index (CCEI), a comparison of carbon-cost contribution at different stages, and a quadrant chart for supplier carbon emission-cost integrated evaluation. It also automatically generates a structured, interactive, and comprehensive analytical report with a complete data traceability chain, providing intuitive support for management decisions.

[0045] like Figure 2As shown, the BOM table extended data association unit 100 consists of multiple collaborative modules. First, the file parsing and standardization module 101 receives the BOM file (supporting Excel, XML, JSON, etc. formats) uploaded by the user through the web interface, calls parsing libraries such as Apache POI to read the file content, and identifies and extracts key fields such as "material code," "material name," "usage," and "unit." Then, the system converts this potentially inconsistent raw data into internally unified standard data objects (e.g., standardizing units of measurement and cleaning illegal characters) and temporarily stores it in a transitional database table. Next, the material details automatic acquisition module 102 starts, using the standardized "material code" obtained in the previous step as the key, and initiates an HTTPS request to the enterprise ERP system (such as SAP or Yonyou U8) through a pre-configured RESTful API interface. An example request is: GET https: / / {erp_host} / api / materials / {material_id}, with an authentication token attached. The ERP system returns the material's JSON details, from which the module parses the detailed specifications, inventory number, and latest purchase price, and associates this information with the corresponding material object.

[0046] For energy consumption items listed in the BOM (e.g., "45 kWh of electricity for production"), the energy and equipment binding module 103 begins operation. Internally, it maintains an "Energy-Equipment Mapping Rule Table," defining the main equipment and their energy types corresponding to different production processes (e.g., "mixing" and "curing"). The module matches the rules based on the energy item description and then obtains the power sampling sequence of the specified equipment during the relevant production period via the OPC UA protocol or by querying the MES system's time-series database (e.g., InfluxDB). By integrating the power sequence over time, the abstract "45 kWh of electricity" is precisely bound and allocated to the specific "mixing station motor" and "curing kiln blower," calculating the individual energy consumption values ​​for each piece of equipment.

[0047] Simultaneously, the supply chain information extension module 104 operates in parallel. Also based on the "material code," it calls the SRM system's SOAP or REST interface to obtain the main supplier information for the material, including the supplier's name, unified social credit code, and the latitude and longitude coordinates of its registered location. Subsequently, the module calls the route planning interface of a third-party map service (such as the Gaode Map API), inputs the supplier coordinates and the company's production base coordinates, and automatically calculates the road transport distance (unit: kilometers). If the SRM system does not specify the transport method for the material, the module will query the built-in "transportation method rule base" based on the material category (such as "bulk cargo" or "precision instruments") and automatically assign a default method (such as "heavy-duty diesel truck").

[0048] Finally, the manual verification and persistence module 105 provides a visual web interface that centrally displays all the automatically associated data in tabular form. LCA engineers can review the accuracy of the data on this interface and manually revise or supplement any questionable or automatically retrieved fields. The system frontend performs basic verification (e.g., usage must be positive). After user confirmation, all data is persistently stored in the extended_bom_table of the system's main database via an ORM framework (such as MyBatis), forming a verified, authoritative version of the data that can be used for all subsequent calculations.

[0049] like Figure 3 As shown, the intelligent matching unit 200 begins its work with the text preprocessing module 201, which reads the original material name (e.g., "rebar Φ12") from the extended_bom_table. It first uses regular expressions to remove special symbols (e.g., "Φ", "-"), units ("mm", "kg"), and other irrelevant characters from the name, obtaining "rebar". Then, it calls a string processing function to convert all characters to lowercase. For Chinese names, it further calls a word segmentation component (e.g., HanLP) for segmentation. For example, "high-efficiency polycarboxylate superplasticizer" is segmented into [high-efficiency, polycarboxylate, superplasticizer], and meaningless words such as "high-efficiency" are filtered out according to the built-in stop word list, resulting in the core word element [polycarboxylate, superplasticizer].

[0050] The synonym mapping module 202 then starts, querying a pre-built industry synonym database (the `synonym_dict` table). This table stores a large number of synonyms and mapping relationships in key-value pairs, such as ("rebar", "threaded steel"), ("threaded steel", "hot-rolled ribbed steel bar"). The module matches the core terminology with the keys in this table, and if a match is found, it replaces it with the standard term. For example, the terminology [rebar] is mapped to the standard term [hot-rolled ribbed steel bar].

[0051] Next, the composite similarity calculation module 203 compares the standardized names with all entry names in the carbon emission factor database (such as the factor_main table). It employs a composite algorithm: on one hand, it uses the Levenshtein distance or Jaro-Winkler algorithm to calculate string similarity scores. To tolerate spelling errors and abbreviations; on the other hand, pre-trained word vector models (such as Word2Vec) or BERT models are used to convert names into high-dimensional vectors and calculate semantic similarity scores. This was done to understand the semantic equivalence between "cement" and "cement". Finally, a name similarity score was calculated. The result is obtained by weighted calculation of both: , where α and β are configurable weights (default 0.4 and 0.6).

[0052] The pedigree matrix screening module 204 is the core decision-making component, and the system presets a confidence threshold (e.g., 0.85). If module 203 calculates the highest... If the threshold is reached or exceeded, the candidate factor is directly selected. If it is not reached, the lineage matrix screening algorithm is activated. This algorithm no longer relies solely on the name, but calculates a comprehensive score for each candidate factor. The calculation formula is as follows: , in, (Data source authority score) Scoring is based on a predefined list, such as government database = 1.0, industry authoritative report = 0.8, enterprise LCA report = 0.6. (Regional Adaptability Score) The judgment factor is the degree of matching between the applicable region and the material's place of origin. For example, the score is 1.0 for the province / city, 0.6 for the whole country, and 0.4 for the international region. (Time Adaptability Score) is based on the difference between the year the data was published and the current year. For example, within 1 year = 1.0, within 5 years = 0.6. (Production process matching score) Evaluation factor describes the degree of consistency between the process and the actual process of the materials. For example, complete matching = 1.0. , , , , Set the preset weights for each dimension (e.g., 0.25, 0.25, 0.20, 0.15, 0.15).

[0053] The module calculates all candidate factors. Then, the data is sorted in descending order, and the highest score is automatically selected as the final match. If the highest score is still below another minimum acceptance threshold (e.g., 0.5), the material is marked as UNMATCHED. The matching results (factor ID, matching type, and scores for each item) are written back to extended_bom_table. At the same time, the complete matching decision path, timestamps, and other information are recorded in matching_log_table, achieving full traceability.

[0054] like Figure 4As shown, the missing data intelligent replacement unit 300 is responsible for processing the "unmatched" entries left by the intelligent matching unit, ensuring that the accounting chain is not broken. The missing data detection module 301 periodically scans the extended_bom_table through scheduled tasks or event listening, and filters out all records with the match_status field being UNMATCHED. For each such record, the system creates a replacement task and puts it into the processing queue.

[0055] The candidate retrieval module 302 accesses the substitution rule database (substitution_rule_db) based on the material's attributes (such as material type and industry classification code GB / T 4754). This database contains a structured substitution_pool table storing a wide range of general factors (such as "average value of ferrous metal smelting industry" and "average factor of electricity in East China"), and a rule_table defining priority order (e.g., Rule 1: average value of similar materials in the same industry > Rule 2: average value of the next higher level industry > Rule 3: approximate value from a globally recognized database). Based on the rules, the module executes an SQL query or calls a search engine (such as Elasticsearch) to retrieve all candidate substitution datasets that meet the criteria.

[0056] The similarity assessment module 303 performs a multi-dimensional evaluation of each retrieved candidate substitution factor and calculates a comprehensive similarity score. The evaluation dimensions and weights are shown in the following examples: Industry similarity (weight 0.3): Compares the degree of matching between the material industry code and the applicable industry code of the candidate factor; Geographic similarity (weight 0.25): Compare the coordinates of the material's origin with the geographic range defined by the candidate factors; Technical similarity (weight 0.25): Compare the production process description of the material with the technical description of the candidate factors; Name semantic similarity (weight 0.2): The cosine similarity between names is calculated again using the word vector model; .

[0057] Uncertainty quantification module 304 makes a decision based on the results of module 303, and selects... The candidate factor with the highest score is used as the final replacement value. Then, crucially, it is based on that factor's... The score and its own data source quality rating The uncertainty coefficient U for this substitution is calculated using the following function: Here, γ and δ are preset weighting coefficients (each defaulting to 0.5). A larger U value indicates higher uncertainty in the substitution. For example, selecting the global industry average as the substitution might result in a higher U value (e.g., 35%). Finally, the substitution result writing module 305 updates the selected substitution factor value and its calculated U value to the carbon_factor field of the corresponding material in the extended_bom_table and marks the status as SUBSTITUTED. Simultaneously, a detailed log entry is recorded in the substitution_log_table, including the original material ID, substitution time, selected substitution factor ID, and so on. The score, U-value calculation process, and the triggering rule ID.

[0058] like Figure 5 As shown, the execution of the full-process fast computing unit 400 is a highly automated and parallel process, including the following steps: S401 Initialization and Data Loading: The user selects the product BOM version to be calculated on the front-end interface and triggers the calculation. After receiving the request, the calculation engine controller generates a globally unique task ID, and then loads all the data of the specified version from the extended_bom_table through an efficient SQL query (using SELECT … FOR UPDATE to avoid dirty reads). This data has been matched or replaced and contains uncertainty information.

[0059] S402, Parallelized Computation Stage: The core of the engine uses Java's ForkJoinPool or ParallelStreams technology to break down and execute computational tasks at each stage of the lifecycle synchronously. Raw material acquisition stage calculation: One calculation thread traverses all material records and executes the formula. Where n is the total number of materials. Let i be the amount of material i used. Let be the carbon emission factor of material i. The calculation result (e.g., 498.87 kg CO2e) is stored in a cache (e.g., Redis) with the key stage:raw_material.

[0060] Transportation phase calculation: Another thread iterates through the materials with transportation information and executes the formula. Where m is the number of materials with transportation information. The weight of material j (converted from usage and density). For its transportation distance, The transportation mode factor is used (e.g., heavy truck: 0.18 kg CO2e / t·km). The results (e.g., 37.11 kg CO2e) are cached with stage:transportation as the key.

[0061] Production phase calculation: The third thread processes energy consumption records and executes the formula. ,in, Let k be the amount of energy consumed. Energy factors (e.g., East China Power Grid electricity: 0.7035 kg CO2e / kWh). Results (e.g., 55.06 kg CO2e) are cached with stage:production as the key.

[0062] S403: Uncertainty Propagation Calculation – This step is the core embodiment of scientific rigor. The system retrieves the uncertainty coefficients Uo and Uo for each input data point (specifically, the substitution factor) from the logs. Then, a Monte Carlo simulation is initiated (using the Apache Commons Math library). For example, for a substitution factor with an uncertainty of ±30%, the system randomly draws tens of thousands of sample values ​​from its normal distribution N (mean, standard deviation). In each simulation iteration, these random samples are used to recalculate the uncertainty coefficients. , , The total carbon footprint of the simulation is obtained by summing the results. After a large number of iterations (e.g., 10,000), a probability distribution of the total carbon footprint is obtained. The mean (point estimate) and standard deviation of this distribution are then analyzed, and the 95% confidence interval (e.g., 591.56 ± 23.6 kg CO2e) is calculated.

[0063] S404: Results aggregation. The results aggregator retrieves the results from the cache for each stage and performs a simple summation to obtain the mean of the total carbon footprint. At the same time, a structured summary is generated, including a breakdown of the total stages, the percentage of each stage, and the quality label for each data point (such as "Matched from CLCD", "Substitution - Global Average, U=35%").

[0064] S405: Result Output and Persistence. The final structured result object is stored in the database's `calculation_result_table` via an ORM framework, associated with the BOM version. Simultaneously, the system calls a template engine (such as FreeMarker) to populate the results into HTML or Word report templates, generating visual reports (pie charts, bar charts). The system pushes the calculation completion status and result summary to the user interface in real time via WebSocket or SSE technology. Reports can be downloaded in PDF or Word format.

[0065] like Figure 6As shown, the carbon footprint and cost integrated output unit 500 transforms environmental data into business insights. The cost calculation and allocation module 501 works first, retrieving the material usage and purchase price from the extended_bom_table to calculate the material cost. Obtain the unit price of energy from the enterprise's financial system interface and calculate the energy cost. Calculate transportation costs based on the logistics rate table. Finally, the total production cost of the product is calculated. .

[0066] The comprehensive evaluation index calculation module 502 receives the total carbon footprint from the end-to-end calculation unit. and the total cost from module 501 Calculate the Carbon Cost Efficiency Index (CCEI): The lower this value, the better the overall economic and environmental benefits of the product. It also calculates the carbon emission percentage at each stage. and cost percentage By comparing these two percentages, key areas for optimization can be clearly identified. For example, it might be discovered that "the transportation phase contributes 6% of carbon emissions but accounts for 25% of costs," indicating significant potential for logistics cost optimization. Furthermore, this module aggregates the cumulative carbon emissions of all materials supplied by a supplier, using the supplier ID as the dimension. and cumulative procurement costs This prepares data for generating a supplier evaluation chart.

[0067] The visualization report generation module 503 calls upon a reporting engine (such as JasperReports or ECharts). It binds data such as carbon footprint details, cost details, CCEI values, stage contribution comparison matrices, and supplier "cost-carbon emissions" scatter plots to a pre-built report template. The report supports two output formats: an interactive web report (HTML5, allowing users to dynamically filter and drill down data) and a static document report (generated as a .docx file via Apache POI or a .pdf file via iText). Each data item in the report is clickable; clicking it displays a pop-up window showing complete traceability information for that data. For example, the carbon emission value of a material can trace its matching factor sources, matching scores, or the basis for substitution and the uncertainty calculation process.

[0068] The data storage and management module 504 is responsible for storing the final generated report JSON model, rendered report files (PDF / DOCX), and chart snapshots to a distributed file system (such as MinIO) or cloud storage, and creating an index in the metadata database report_table. All user operations such as viewing, downloading, and data tracing of the report are recorded in audit_log_table to meet compliance and audit requirements. Specific Implementation Example 1: 1. Background and Scenarios of the Implementation Examples This embodiment uses the high-strength precast concrete slab (model: HC-PC-2024) produced by Jiangsu Green Building Materials Co., Ltd. as an example to demonstrate how the above-mentioned carbon footprint intelligent matching and rapid calculation system based on BOM can quickly calculate the product's full life cycle carbon footprint and cost.

[0070] Product: High-strength precast concrete slab (HC-PC-2024); Functional unit: 1 cubic meter (m³); System Configuration: The system is deployed on the enterprise server and has built-in the "China Product Life Cycle Greenhouse Gas Emission Coefficient Database (2023)" and the "IPCC Sixth Assessment Report" factor database. It has also successfully completed data interface integration with the enterprise's internal Yonyou U8 (ERP), Siemens Simatic IT (MES) and self-developed SRM system.

[0071] 2. Specific input data required for implementation 2.1 Initial BOM (User Upload): Users upload a file named HC-PC-2024_BOM.xlsx via the system's web interface. The content is as follows:

[0072] Additional notes on the BOM table: Hierarchical structure: Level 0: Represents the final product.

[0073] Level 1: Represents sub-assemblies, purchased parts, packaging materials, and directly consumed energy and generated waste that directly make up the final product.

[0074] Level 2: Represents the original raw materials that make up each sub-assembly.

[0075] Stage Classification: Raw material acquisition: The stage of mining, producing, and transporting all materials to the factory.

[0076] Product manufacturing: The processes of mixing, pouring, curing, and packaging carried out in a factory, including the energy and resources directly consumed.

[0077] Waste disposal: The treatment of waste generated during the production process.

[0078] 2.2 The system automatically associates and expands data. After parsing the HC-PC-2024_Full_BOM.xlsx file, the system automatically retrieves the following extended information from the internal system interface, forming a comprehensive product environment - economic dataset: a. Material unit price obtained from the ERP system

[0079] b. Supply chain information obtained from SRM and GIS systems

[0080] c. Energy-consuming device binding information obtained from the MES system

[0081] d. Other automatically supplemented information Waste disposal method: For WST-500 production waste, the system automatically associates its disposal method with the enterprise's preset waste management rules, which is to transport it to the municipal designated construction waste recycling point, which is 20 km away, and the transportation method is light truck (diesel), with a disposal cost of 50 yuan / ton.

[0082] Data quality labeling: The system automatically labels RM005 (carbon emission factor of clean water) as local database - default value, and RM007 (recycled polypropylene) as preferred recommendation - recycled material.

[0083] Output: At this point, the system has expanded a simple BOM into a complete dataset containing materials, usage, units, suppliers, transportation methods, transportation distances, prices, energy-consuming equipment, and waste disposal methods, laying a solid data foundation for subsequent intelligent matching and accurate calculations. All automatically acquired data can be manually reviewed and corrected through the interactive interface.

[0084] 3. Specific implementation steps and data flow Step 1: BOM parsing and data expansion association User operation: The user uploads the HC-PC-2024_Full_BOM.xlsx file through the web interface.

[0085] System execution: The BOM parsing unit calls the Apache POI library to parse the Excel file and identify the product hierarchy (Level 0 product, Level 1 assembly / purchased parts / energy, Level 2 raw materials). The system automatically creates a product structure tree and identifies all material, energy, and waste entries.

[0086] Data flow: Input: Original BOM file.

[0087] Processing: Parse the file and build a structured data model.

[0088] Output: The preliminary parsed BOM data tree is temporarily stored in the database staging_bom_table.

[0089] Step 2: Intelligent matching of carbon emission factors System execution: (a) Name preprocessing and standardization The original name in the BOM table is: "Hot-rolled ribbed steel bar HRB400 Φ12".

[0090] The system performs the following preprocessing: Remove redundant symbols and unify the diameter symbol (Φ12 is stored separately as a specification attribute).

[0091] The thesaurus maps "hot-rolled ribbed steel bars" to the standard term "threaded steel bars".

[0092] The final standardized name is: "HRB400 rebar".

[0093] (ii) Candidate factor retrieval The system calls the factor database to retrieve candidate entries and returns sample results:

[0094] (III) Similarity Calculation The system calculates string similarity for candidate entries. semantic similarity and according to the formula α=0.4, β=0.6 The overall name similarity score was obtained, and the calculation results are as follows:

[0095] (iv) Genealogical matrix scoring The system further calculates scores for each dimension of the candidate items: data source authority. Regional adaptability Time Adaptability Process compatibility And according to the formula: ,(in, =0.25, =0.25, =0.20, =0.15, =0.15) to obtain the overall score.

[0096] Take candidate A as an example: = 0.938; =0.90; =1.00; =0.95; =0.85, Calculate item by item: 0.25 × 0.938 = 0.2345 0.25 × 0.90 = 0.2250 0.20 × 1.00 = 0.2000 0.15 × 0.95 = 0.1425 0.15 × 0.85 = 0.1275 Summation: = 0.2345+0.2250+0.2000+0.1425+0.1275= 0.9295.

[0097] Similarly: Candidate B ≈0.8805.

[0098] Candidate C ≈0.7495.

[0099] (v) Matching determination The system's automatic matching threshold is set to 0.85. Candidate A: 0.9295 ≥ 0.85 → Meets the criteria; Candidate B: 0.8805 ≥ 0.85 → Meets the criteria, but the score is lower than A; Candidate C: 0.7495 < 0.85 → Not suitable; Therefore, the system ultimately selected candidate A (F002: HRB400 rebar) as the matching factor for this material.

[0100] (vi) Uncertainty Calculation The system is based on the formula: U = 0.5 × (1 )+0.5×(1 ) Calculate the uncertainty, where, Score the data quality.

[0101] For candidate A: = ( + + + ) / 4 = (0.90+1.00+0.95+0.85) / 4=0.925, 1 =1 0.9295 = 0.0705 1 =1 0.925 = 0.075 U=0.5×0.0705+0.5×0.075=0.07275, That is, the uncertainty is approximately ±7.275%.

[0102] (vii) Final output Matching factor: F002, value = 2.05 kg CO2e / kg Match confidence level: 0.9295 Uncertainty level: ±7.275%.

[0103] Confidence interval: 1.90, 2.20 kgCO2e / kg (rounded).

[0104] For "Recycled Polypropylene (PP)," the system automatically identifies the keyword "recycled," prioritizes matching recycled material factors (such as "recycled polypropylene" in Ecoinvent), and marks it as "Preferred Recommendation - Recycled Material." For "Pine Wood Raw Material," the system automatically selects the "Sawn Timber (Average of East China)" factor based on the material's origin (Huzhou City, Zhejiang Province), which has the highest regional suitability. Data Flow: Input: Standardized material names from the staging_bom_table.

[0105] Processing: Intelligent matching and genealogy matrix filtering.

[0106] Output: Enhanced BOM data, including matching carbon emission factor IDs, stored in extended_bom_table.

[0107] Step 3: Intelligent replacement of missing data System Execution: For "plastic maintenance pad block", there is no direct match in the database. The system automatically triggers the "intelligent replacement of missing data" process. This process includes: standardization, alternative candidate retrieval, alternative priority judgment, uncertainty assignment, alternative value output, result recording, and manual review.

[0108] A. Standardization and Triggering Original BOM name: "Plastic maintenance pad block".

[0109] Name standardization: Remove redundant symbols, unify word forms, and map to the standard category "plastic products".

[0110] Because no entry matching "plastic maintenance pad block" was found in the main factor table, the alternative process was triggered.

[0111] B. Alternative Candidate Retrieval and Priority Rules The system retrieves substitution factors according to the following priorities: 1) Neighboring entries of the same category, with the same technology, and in the same region; 2) Entries of the same category but different specifications; 3) Category average (country / region average); 4) Global average by category (as a last resort).

[0112] In this example, the search results showed no identical entries, so the category "Plastics (General) - Global Average" was ultimately selected as the alternative factor source.

[0113] C. Principle of Assigning Values ​​Based on Uncertainty When the substitution factor is adopted as a category global average, a higher uncertainty is assigned according to the system rules to reflect the generalization of data sources and process / formulation differences.

[0114] Assignment logic: Global average fundamental uncertainty =30%; Penalty for lack of specific manufacturing process, density, and additive information. =5%, Ultimate substitution uncertainty =35%, The above =30% =5% comes from the system's built-in "Alternative Factor Uncertainty Rule Base", which is defined in the "Alternative Rule Database (S302, S304)" of the technical solution.

[0115] D. Application of substitution factor values ​​and calculation of confidence intervals Assuming a substitution factor v = 2.50 kgCO2e / kg, the relative uncertainty is... =35%, Lower bound factor value =2.50×(1–0.35)=1.625 kgCO2e / kg, Upper bound factor value =2.50×(1+0.35)=3.375 kgCO2e / kg, If the mass of "plastic curing pad" in the BOM is m = 0.5 kg, then: The mean E = 2.50 × 0.5 = 1.25 kgCO2e. The Lower World =1.625 × 0.5 = 0.8125 kgCO2e, Upper Realm =3.375 × 0.5 = 1.6875 kgCO2e, The above is used for uncertainty propagation in LCA.

[0116] E. Manual Review Triggering Rules and Execution Review threshold: If there is uncertainty after substitution If the percentage is greater than 20% or replaced with a global average, a review will be automatically triggered. This example... =35% > 20%, therefore a manual review task is triggered. The interface displays: original name, standardized name, candidate alternative, alternative value, uncertainty, and confidence interval. Manual review can: accept, modify, or upload internal factors.

[0117] F. Logs and traceability records G. Uncertainty propagation in overall LCA results Alternative terms are included in the Monte Carlo sampling according to their distribution (mean v, standard deviation σ = v × ... Together with other entries, the total carbon footprint is calculated; the front end can first return point estimates and uncertainties, and the back end can asynchronously update the Monte Carlo results.

[0118] Data flow: Input: No matching material entry.

[0119] Processing: Alternative rule query, similarity assessment, uncertainty labeling.

[0120] Output: Replacement factors marked with uncertainty, updated back to extended_bom_table.

[0121] Step 4: Rapid Calculation Throughout the Entire Process Input: The complete extended BOM dataset after the aforementioned steps (all materials have been matched or replaced with carbon emission factors).

[0122] System execution: The end-to-end fast computing engine starts, loads data, and performs parallel computing according to the following lifecycle stages.

[0123] I. Calculations during the raw material acquisition stage Calculation formula: Carbon emission per material = Material usage × Carbon emission factor Calculation process:

[0124] II. Calculation of Transportation Stages Calculation formula: Carbon emissions from single material transportation = Material weight (tons) × Transportation distance (km) × Emission factor of transportation mode (kg CO2e / t·km).

[0125] Public data: Transportation emission factors are taken from the "Guidelines for the Compilation of Provincial Greenhouse Gas Inventories": Heavy-duty diesel trucks: 0.18 kg CO2e / t·km; Dump trucks: 0.18 kg CO2e / t·km; Vans: 0.16 kg CO2e / t·km; Light trucks: 0.16 kg CO2e / t·km.

[0126] Calculation process:

[0127] III. Production Stage Calculation Calculation formula: Carbon emissions per energy source = Energy consumption × Carbon emission factor Calculation process:

[0128] IV. Calculation of Waste Disposal Phase Calculation formula: Carbon emissions from waste treatment = Transportation emissions + Disposal process emissions. Calculation process:

[0129] V. Summary and Uncertainty Propagation Computation Total carbon footprint calculation: Total carbon footprint = raw materials + transportation + production + waste = 498.874 + 37.112 + 55.0575 + 0.516 = 591.5595 kg CO2e ≈ 591.56 kg CO2e Uncertainty propagation calculation (Monte Carlo simulation): The main source of uncertainty is the substitution factor of "PC water-reducing agent" (0.85±30%).

[0130] Simulation process: 10,000 iterative simulations were performed. In each iteration, the factor value of the water-reducing agent was randomly selected from a normal distribution with a mean of 0.85 and a standard deviation of 0.255.

[0131] Simulation results: A total carbon footprint of 10,000 was obtained, with a mean of 591.56 kgCO2e and a standard deviation of ±11.8 kgCO2e.

[0132] Final statement (95% confidence interval): The product's total lifecycle carbon footprint is 591.56 ± 23.6 kg CO2e.

[0133] Output: The full-process fast calculation engine packages the above-mentioned stage details, sums, and uncertainty ranges into a result object.

[0134] Step 5: Cost Calculation and Overall Output Input: (1) The carbon footprint calculation results above; (2) Material details data from the BOM table extended data association module.

[0135] i. Cost Calculation and Allocation Calculation process: a) Raw material cost calculation:

[0136] b) Energy cost calculation:

[0137] c) Transportation cost calculation: Rates: Heavy trucks: 1.2 yuan / t·km; Dump trucks: 1.1 yuan / t·km; Box trucks / Light trucks: 1.5 yuan / t·km.

[0138] Calculation formula: Single material transportation cost = Material weight (t) × Transportation distance (km) × Transportation rate (yuan / t·km)

[0139] Total transportation cost: 243.98 yuan. d) Waste disposal cost calculation: Waste disposal cost = (5 / 1000) × 50 × 1000 = 0.25 yuan.

[0140] e) Total Cost Summary: Total production cost = Raw material cost + Energy cost + Transportation cost + Waste disposal cost =659.465+65.95+243.98+0.25 =969.645 yuan ≈969.65 yuan.

[0141] ii. Calculation of Comprehensive Evaluation Index Carbon Cost Efficiency Index (CCEI): CCEI = Total Carbon Footprint of Product / Total Cost of Product =591.56 / 969.65 ≈0.6101 kg CO2e / yuan Phase contribution analysis (carbon vs. cost):

[0142] iii. Comprehensive Output Report Comprehensive Analysis Report on Product Carbon Footprint and Cost Product Information: High-strength precast concrete slab (HC-PC-2024), functional unit: 1m³.

[0143] Environmental performance: Life cycle carbon footprint = 591.56 ± 23.6 kg CO2e (95% confidence interval).

[0144] Economic performance: Total production cost = 969.65 yuan.

[0145] Key decision-making indicator: Carbon Cost Efficiency Index (CCEI) = 0.6101 kg CO2e / yuan.

[0146] The raw material stage is the absolute core of carbon and cost reduction (accounting for 84.3% of carbon emissions and 68.01% of costs). The transportation stage's cost share (25.16%) is nearly four times its carbon emission share (6.3%), indicating that current logistics costs are extremely high. Optimizing the supply chain layout, negotiating transportation prices, or changing transportation methods have significant potential for cost reduction. Using recycled polypropylene (bed blocks) and recycled steel bars (if available) is an excellent direction for carbon reduction, but the cost impact needs to be comprehensively assessed.

[0147] Data traceability: The report allows you to click to view the source and calculation process of any data.

[0148] Output the structured data, visualization charts, and downloadable PDF report.

[0149] This embodiment verifies the efficiency and reliability of the BOM-based intelligent carbon footprint matching and rapid calculation system when handling complex BOM structures. It can automatically complete the full lifecycle data association of multi-level materials and output a report containing carbon footprint, cost, uncertainty, and decision indicators within one minute, which is more than 100 times more efficient than traditional methods. Through intelligent matching and substitution mechanisms, the results are more accurate and the uncertainty is quantified (591.56±23.6 kgCO2e). At the same time, it reveals the hidden problem of high cost and low carbon emissions in the transportation stage, provides a clear optimization sequence and decision basis, and fully meets the traceability requirements of international standards.

[0150] Example 2 1. Experimental Objective To verify the technical effectiveness of the system of the present invention in terms of efficiency and accuracy of carbon footprint accounting, a comparative experiment was conducted between the intelligent carbon footprint matching and rapid accounting system based on BOM and the traditional manual matching + general LCA software (such as SimaPro / GaBi).

[0151] 2. Experimental Environment Hardware configuration: CPU: Intel Xeon Silver 4310×16 core; Memory: 64 GB; Hard drive: 1 TB NVMe SSD; Operating system: Linux Ubuntu 22.04.

[0152] Software configuration: A carbon footprint intelligent matching and rapid calculation system based on BOM: deployed on a local server with a parallel computing engine enabled (16 threads). Comparison system: manual factor selection + general LCA tool (single-threaded calculation mode).

[0153] 3. Experimental Data Select the BOM (Bill of Materials) for three typical product categories: Precast concrete components: Approximately 150 items in the Bill of Materials (BOM); Prefabricated steel structure components: approximately 600 BOM items; Mechanical and electrical equipment subsystem: approximately 2,000 BOM items; The carbon emission factor database contains approximately 50,000 entries (including factors from multiple regions, processes, and years).

[0154] 4. Test Indicators Calculation time: The total time from importing the BOM to outputting the full lifecycle carbon footprint results; Matching accuracy: The consistency rate between automatically matched factors and human expert annotation results; Uncertainty analysis capability: Whether it can output confidence intervals and their processing time.

[0155] 5. Experimental Results

[0156] In addition, the system performs 10,000 Monte Carlo simulations, assuming that the input data follows a normal distribution and that the convergence condition is a standard deviation change of less than 1%, and outputs carbon footprint results with a 95% confidence interval. In contrast, traditional methods require manual step-by-step parameter setting, which takes more than 2 hours.

[0157] 6. Experimental results show that: The BOM-based intelligent carbon footprint matching and rapid calculation system achieves higher matching accuracy (an average improvement of approximately 12%–15%) through semantic expansion and phylogenetic matrix filtering. By employing a parallel computing engine and caching mechanism, it reduces computation time from hours to seconds to minutes, achieving a performance improvement of over 100 times. The system supports uncertainty propagation and confidence interval output, significantly outperforming traditional systems that require manual setup and single simulations. Therefore, the BOM-based system demonstrates outstanding and substantial technical advantages in efficiency, accuracy, and functional completeness, meeting the needs of rapid carbon footprint calculation for large-scale products.

[0158] Example 3: Database Structure and Lineage Matrix 1. Carbon emission factor database structure The carbon emission factor database in the BOM-based intelligent carbon footprint matching and rapid calculation system adopts a relational data structure, and its core tables include: Factor Main Table: Factor_ID: Primary key, unique identifier; Name: Factor name (e.g., "Ordinary Silicate Cement P.O42.5"); Category: The category to which the product belongs (e.g., "cement", "steel", "electricity"); Unit: Functional unit (e.g., "kg CO2e / kg"); Value: Carbon emission factor numerical value; Source_ID: Foreign key, associated with the source table.

[0159] Factor_Source table: Source_ID: Primary key; Source_Name: Source name (e.g., "IPCC 2019"); Authority_Level: Authority level (0–1, the higher the value, the more authoritative). Publication_Year: Publication date; Region: Applicable geographical area (e.g., "East China", "National Average", "Europe").

[0160] Process Information Sheet (Factor_Process): Process_ID: Primary key; Factor_ID: Foreign key; Process_Type: Process type (e.g., "wet process", "dry process", "converter"); Tech_Score: Process compatibility score (0–1).

[0161] Synonym Dict: Synonym_ID: Primary key; Standard_Term: Standard terminology (e.g., "reinforcing bar"); Synonym: a word or phrase that is synonymous with something else (such as "rebar" or "steel bar"). 2. Sample Data (Simplified Table) Table 1: Factor_Main

[0162] Table 2: Factor_Source

[0163] Table 3: Factor_Process

[0164] Table 4: Synonym_Dict

[0165] 3. Example of genealogical matrix calculation If the material name "steel bar HRB400" appears in the BOM, the system will search for candidate factor F002: threaded steel (HRB400).

[0166] Similarity and weighting are calculated as follows: String similarity ( Levenshtein distance similarity = 0.92; Semantic similarity ( (): Based on word vector cosine similarity = 0.95; Overall name similarity ( ): =0.4×0.92+0.6×0.95=0.938.

[0167] Data source authority ( ): 0.9 (Source: China Carbon Emission Factor Handbook 2023); Regional adaptability ( ): 1.0 (Region: China, exact match); Time fit ( ): 0.95 (Published in 2023, approximately the current year); Process compatibility ( ): 0.85 (Process: converter); Genealogical matrix comprehensive score ( ): =0.25×0.938+0.25×0.9+0.20×1.0+0.15×0.95+0.15×0.85≈0.929 ≈0.929; Judgment: If the threshold is set to 0.85, then the automatic matching is considered successful.

[0168] 4. Technical Effects: Through structured database management and phylogenetic matrix weighting, the carbon footprint intelligent matching and rapid calculation system based on BOM tables supports rapid retrieval of factors across regions, years, and processes; provides traceable matching basis; and significantly improves the accuracy and interpretability of factor matching.

[0169] Based on the above-described preferred embodiments of the present invention, and through the foregoing description, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A carbon footprint intelligent matching and rapid calculation system based on a Bill of Materials (BOM), characterized in that, include: The BOM table extended data association unit is used to parse the product bill of materials file and automatically obtain the supply chain information, cost information and energy-consuming equipment binding information of the materials by calling the enterprise system's application programming interface to form an extended bill of materials dataset. The intelligent matching unit, connected to the BOM table extended data association unit, is used to standardize the material name through natural language processing. When the matching confidence of the material name with the carbon emission factor database is lower than a preset threshold, the optimal carbon emission factor is selected based on a weighted score of multiple preset dimensions. The missing data intelligent substitution unit is connected to the intelligent matching unit and is used to select candidate factors from the substitution factor library for materials that have not been successfully matched, according to preset rules, and calculate the corresponding uncertainty coefficient. The end-to-end rapid calculation unit is connected to the missing data intelligent replacement unit to calculate the carbon emissions of each stage of the product's entire life cycle in parallel, and performs uncertainty propagation calculation through Monte Carlo simulation to output carbon footprint results with confidence intervals. The carbon footprint and cost integrated output unit is connected to the full-process rapid calculation unit to integrate carbon emission results and cost data, calculate the carbon cost efficiency index, stage contribution comparison and supplier comprehensive evaluation, and generate a comprehensive analysis report.

2. The intelligent carbon footprint matching and rapid calculation system based on BOM (Bill of Materials) according to claim 1, characterized in that, The BOM table extended data association unit includes: The document parsing and standardization module is used to parse bill of materials (BOM) files and extract key fields. The automatic material details acquisition module is used to obtain detailed material specifications and purchase prices by calling the interface of the enterprise resource planning system based on the material code. The energy and equipment binding module is used to query preset energy-equipment mapping rules and obtain equipment energy consumption data from the manufacturing execution system; The supply chain information extension module is used to call the supplier relationship management system interface to obtain supplier coordinates and call the geographic information system service interface to calculate transportation distance.

3. The intelligent carbon footprint matching and rapid calculation system based on BOM (Bill of Materials) according to claim 1, characterized in that, The intelligent matching unit includes: The text preprocessing module is used to perform text cleaning, word segmentation, and stop word removal on material names; The thesaurus mapping module is used to query a pre-built thesaurus and map non-standard terms to standard terms; The composite similarity calculation module is used to calculate string similarity scores. and semantic similarity score ; The genealogy matrix filtering module is used to perform weighted scoring based on multiple preset dimensions when the matching confidence is lower than the threshold.

4. The intelligent carbon footprint matching and rapid calculation system based on BOM (Bill of Materials) according to claim 3, characterized in that, The genealogy matrix screening module calculates the comprehensive score of candidate factors based on the following formula. : ， in, Score the similarity between names. Score for the authority of the data source. The score is based on the degree of regional suitability. Score for time fit. The production process matching score is used to determine the production process matching score. , , , , The preset weights are for each dimension.

5. The intelligent carbon footprint matching and rapid calculation system based on BOM (Bill of Materials) according to claim 4, characterized in that, Name similarity score The composite similarity calculation module calculates the similarity using the following formula: ， Where α and β are preset weighting coefficients, and α+β=1.

6. The intelligent carbon footprint matching and rapid calculation system based on BOM (Bill of Materials) according to claim 1, characterized in that, The missing data intelligent replacement unit includes: The missing item detection module is used to identify material records that are not matched. The candidate retrieval module is used to query the substitution rule database and retrieve candidate substitution factors based on material attributes; The similarity assessment module is used to calculate a comprehensive similarity score for candidate substitutes. ; The uncertainty quantification module is used to calculate the uncertainty coefficient U according to the following formula: ， in, γ and δ are preset weighting coefficients used to assess the quality of alternative factor data sources.

7. The intelligent carbon footprint matching and rapid calculation system based on BOM (Bill of Materials) according to claim 1, characterized in that, The end-to-end high-speed calculation unit includes: A parallel computing engine is used to perform the following computations synchronously: Carbon emissions during the raw material acquisition stage Where n is the total number of materials. Let i be the amount of material i used. Let be the carbon emission factor of material i; Carbon emissions during transportation Where m is the number of materials with transportation information. Let j be the weight of material j. For its transportation distance, The carbon emission factor of its transportation method; Carbon emissions during the production phase Where p is the total number of energy types, Let k be the amount of energy consumed. , where k is the carbon emission factor for energy. The uncertainty propagation module is used to calculate the total carbon footprint using Monte Carlo simulation based on the uncertainty coefficients of each input data. The confidence interval.

8. The intelligent carbon footprint matching and rapid calculation system based on BOM (Bill of Materials) according to claim 1, characterized in that, The integrated carbon footprint and cost output unit includes: The cost calculation and allocation module is used to calculate material costs, energy costs, transportation costs, and total product production costs based on material usage, purchase price, energy consumption, and transportation rates. ; The comprehensive evaluation index calculation module is used to calculate the carbon cost efficiency index. And the proportion of carbon emissions at each stage of the life cycle. and cost percentage ,in and These represent the emissions and costs of stage s, respectively. The supplier evaluation module is used to aggregate the carbon emissions and procurement costs of suppliers to generate comprehensive supplier evaluation data. The visualization report generation module is used to render carbon footprint details, cost details, and comprehensive evaluation indicators into a comprehensive analysis report with data traceability capabilities.

9. A method for intelligent matching and rapid calculation of carbon footprint based on BOM (Bill of Materials), characterized in that, include: S1. Parse the product bill of materials file and automatically obtain the supply chain, cost and energy consumption information of the materials by calling the enterprise system interface to form an extended bill of materials dataset; S2. Standardize the material name. When the confidence level of the material name matching the carbon emission factor database is lower than the preset threshold, select the optimal carbon emission factor based on a weighted score of multiple preset dimensions. S3. For materials that are not successfully matched, select candidate factors from the alternative factor library according to preset rules for substitution, and calculate the corresponding uncertainty coefficient. S4. Perform parallel calculations on carbon emissions at each stage of the product's entire lifecycle, and use Monte Carlo simulation to perform uncertainty propagation calculations, outputting carbon footprint results with confidence intervals. S5 integrates carbon emission results and cost data, calculates the carbon cost efficiency index, compares stage contributions, and provides a comprehensive evaluation of suppliers, generating a comprehensive analysis report.

10. The method for intelligent matching and rapid calculation of carbon footprint based on BOM table according to claim 9, characterized in that, Step S1 is as follows: S101. Parse the product bill of materials file, extract the material name, material code, usage and unit of measurement fields, and convert them into standard data objects for internal system use. S102. Based on the material code, call the application programming interface of the enterprise's internal resource planning system or manufacturing execution system to obtain the detailed specifications and purchase price of the material. S103. Based on the energy entries in the bill of materials, query the preset energy-equipment mapping rules, and obtain the production energy consumption time series data of the corresponding equipment from the manufacturing execution system, and bind the energy consumption to the specific production equipment. S104. Based on the material code, call the interface of the supplier relationship management system to obtain the main supplier information and coordinates, and call the geographic information system service interface to calculate the transportation distance from the supplier to the production site. S105. Based on the standardized material and energy names, query the local or external carbon emission factor database to obtain a preliminary list of carbon emission data entries. S106. Provide a data verification interface for users to review and revise automatically associated data, and persistently store the final data in an extended bill of materials data table.

11. The method for intelligent matching and rapid calculation of carbon footprint based on BOM table according to claim 9, characterized in that, Step S2 is as follows: S201. Perform text cleaning, case unification, word segmentation, and stop word removal on the original names of materials or energy to obtain a core name word sequence; S202. Query the pre-built thesaurus and map the core name word sequence to standard terms; S203. Perform a composite similarity calculation between the standard terminology and the entry names in the carbon emission factor database to obtain the string similarity score. and semantic similarity score ; S204. Determine whether the highest similarity score has reached the preset automatic matching threshold. If not, trigger the lineage matrix filtering based on name similarity score. Data source authority score Regional adaptability score Time fit score Matching score with production process The overall score for each candidate factor is calculated as follows: And select the factor with the highest score: ， in, , , , , , The preset weights are for each dimension.

12. The method for intelligent matching and rapid calculation of carbon footprint based on BOM table according to claim 9, characterized in that, Step S3 is as follows: S301. Scan the extended bill of materials data table, identify material records with a status of unmatched, and create alternative processing tasks; S302. Based on the properties of the material, access the substitution rule database and retrieve candidate substitution factors according to the preset substitution priority rules. S303. For each candidate substitution factor, a multi-dimensional weighted evaluation is performed based on industry similarity, geographical similarity, technological similarity, and name semantic similarity to calculate a comprehensive similarity score. ; S304. Select the candidate replacement factor with the highest comprehensive similarity score as the replacement value, and then base the replacement value on its score. and data source quality rating The uncertainty coefficient U is calculated using the following formula: ， Where γ and δ are preset weighting coefficients; S305. Write the substitution value and uncertainty coefficient into the extended bill of materials data table, and record the substitution decision log.

13. The method for intelligent matching and rapid calculation of carbon footprint based on BOM table according to claim 9, characterized in that, Step S4 is as follows: S401. Responding to the user's calculation request, load the specified version of the extended bill of materials dataset that has undergone data processing from the database; S402. Using a parallel computing framework, the following calculations are executed synchronously: Carbon emissions during the raw material acquisition stage Where n is the total number of materials. Let i be the amount of material i used. Let be the carbon emission factor of material i; Carbon emissions during transportation Where m is the number of materials with transportation information. Let j be the weight of material j. For its transportation distance, The carbon emission factor of its transportation method; Carbon emissions during the production phase Where p is the total number of energy types, Let k be the amount of energy consumed. , where k is the carbon emission factor for energy. S403. Based on the uncertainty coefficients of each input data, Monte Carlo simulation is used to calculate the uncertainty propagation and obtain the total carbon footprint. The probability distribution and confidence interval; S404. Summarize the results of each stage and generate a result object that includes total emissions, emissions in each stage, uncertainty range and data quality identifier; S405. Persistently store the result object and automatically generate a visual carbon footprint calculation report.

14. The method for intelligent matching and rapid calculation of carbon footprint based on BOM table according to claim 9, characterized in that, Step S5 is as follows: S501. Align and correlate the carbon footprint results output by the full-process rapid calculation unit with the material cost, energy cost, and transportation cost data obtained by the BOM extended data association unit. S502, Calculate the total production cost of the product S503, Calculate the carbon cost efficiency index Calculate the carbon emission share at each stage of the life cycle. and cost percentage ,in, and These represent the emissions and costs of stage s, respectively. By aggregating the carbon emissions and procurement costs of suppliers, a comprehensive supplier evaluation chart is generated. S504: Call the report engine to render carbon footprint details, cost details, carbon cost efficiency index, stage contribution comparison, and supplier comprehensive evaluation into an interactive or static comprehensive analysis report. S505. Store the generated report files and user operation audit logs.

15. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the carbon footprint intelligent matching and rapid calculation method based on BOM table as described in any one of claims 9 to 14.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the intelligent matching and rapid calculation method for carbon footprint based on BOM as described in any one of claims 9 to 14.

Images