Carbon data processing method and device for steel industry carbon measurement review scene

Abstract

This invention relates to the field of carbon emission technology, specifically to a carbon data processing method and apparatus for carbon measurement and review scenarios in the steel industry. The method includes: identifying, classifying, and integrating raw multi-source data to generate carbon data governance units; preprocessing the carbon data governance units and inputting them into a governance status diagnostic model for diagnosis, performing measurement support verification and accounting support verification to obtain a first labeling result and a first verification result; generating a problem list and rectification path based on the above results; then rectifying the preprocessed carbon data governance units, and inputting the rectified carbon data governance units into the governance status diagnostic model for review and verification, until the rectified carbon data governance units are restored to an auditable state. The method proposed in this application establishes a closed loop of identification—classification and integration—governance—diagnosis—rectification—review, solving the problems of weak data correlation and lack of vertical governance methods in existing carbon data processing methods.

Description

Technical Field

[0001] This invention relates to the field of carbon emission technology, specifically to a carbon data processing method and apparatus for carbon metering and auditing scenarios in the steel industry. Background Technology

[0002] With the expansion of the national carbon market and the continuous strengthening of regulatory requirements for key emission units, steel companies, in addition to conducting carbon emission accounting and reporting, also need to undergo carbon measurement audits to systematically prove that the sources of their carbon-related data are clear, the values ​​are reasonable, the evidence is complete, the standards are consistent, and the process is traceable. The steel industry has a long production process chain, including long processes such as coking, sintering or pelletizing, blast furnace, converter, continuous casting, and rolling, as well as short processes mainly involving scrap steel and electric arc furnaces. It also involves cross-process scenarios such as lime kilns, self-owned power plants, by-product gas recovery and utilization, steam pipelines, and auxiliary systems. This results in a large amount of production data, energy data, measurement data, laboratory data, financial settlement data, ledger data, and document evidence data being generated within the company.

[0003] However, in carbon measurement and auditing scenarios, not all data generated by enterprises directly constitutes auditable carbon data. What truly enters the auditing process are vertically integrated carbon data objects related to enterprise boundaries, process boundaries, emission sources, activity levels, emission factors, carbon content, inventory changes, original records, responsible entities, and supporting documentation. These carbon data objects need to be identified, classified, correlated, integrated, governed, and bound to evidence across different systems to form a data foundation usable for auditing, verification, and rectification review. However, current technologies suffer from weak correlations between auditable carbon data in carbon measurement and auditing scenarios and lack vertical governance methods, specifically as follows:

[0004] (1) Existing solutions mostly focus on general production measurement or general carbon emission accounting, lacking a carbon data object identification and classification mechanism for carbon measurement review scenarios, and cannot answer the questions of "which data belongs to carbon data, which process or emission source, and which type of review object".

[0005] (2) Existing solutions often focus on collection and storage, lacking a carbon data governance mechanism that integrates measurement data, test data, ledger data, settlement data and original evidence, resulting in weak correlation between data, chaotic versions and unclear responsibilities.

[0006] (3) Even if existing technologies perform data cleaning or rule verification, they are mostly general integrity checks, scope checks and consistency checks, lacking vertical governance methods that address the characteristics of the steel industry, such as the coexistence of continuous and batch processes, the cross-process flow of by-product gas, and significant inventory changes.

[0007] (4) Existing technologies have a lot of diagnostic capabilities for measuring instruments and accounting deviations, but they do not reconstruct multi-source data such as measurement, testing, process, finance and ledger into "auditable carbon data governance units", so it is difficult to output truly relevant conclusions for carbon measurement auditing.

[0008] (5) Existing technologies make it difficult to incorporate issues such as carbon data misclassification, source conflicts, lack of evidence, incomplete governance status and accounting boundary omissions, incorrect activity level values, and mismatched emission factors into the same review and diagnosis framework.

[0009] Therefore, there is an urgent need for a technical solution for carbon measurement review of key emission units in the steel industry. First, the multi-source data of enterprises should be identified and reconstructed into a reviewable carbon data object system and carbon data governance unit. Then, carbon data governance diagnosis should be carried out on this basis, supplemented by carbon balance diagnosis of measurement, process and control system, so as to generate diagnostic conclusions and rectification paths that are truly suitable for carbon measurement review scenarios. Summary of the Invention

[0010] The purpose of this application is to propose a carbon data processing method and apparatus for the carbon metering and auditing scenario in the steel industry, addressing the aforementioned technical problems.

[0011] In a first aspect, the present invention provides a carbon data processing method for a carbon metering review scenario in the steel industry, comprising the following steps:

[0012] Obtain raw multi-source data, perform data object identification and classification on the raw multi-source data, and obtain classified carbon data;

[0013] Carbon data from different sources but pointing to the same process, the same emission source or the same medium are merged to generate carbon data governance units;

[0014] The carbon data governance unit is preprocessed based on the process type, value retrieval method, time granularity, source quantity, and inventory changes to obtain the preprocessed carbon data governance unit.

[0015] A governance status diagnosis model is constructed. The governance status diagnosis model is used to mark carbon data governance units according to the set diagnosis content and obtain the marking results. The preprocessed carbon data governance units are input into the governance status diagnosis model to obtain the first marking result.

[0016] The preprocessed carbon data governance unit is subjected to metrological support verification and accounting support verification to obtain the first verification result.

[0017] A problem list and rectification path are generated based on the first marking result and the first verification result;

[0018] Based on the problem list and rectification path, the preprocessed carbon data governance unit is rectified to obtain the rectified carbon data governance unit.

[0019] Input the rectified carbon data governance unit into the governance status diagnosis model and review and verify it to obtain the second marking result and the second verification result. If the second marking result indicates that the governance is complete and the second verification result indicates that all are passed, the rectified carbon data governance unit is restored to the reviewable state. Otherwise, generate a problem list and rectification path based on the second marking result and the second verification result, and repeat the previous step until the rectified carbon data governance unit is restored to the reviewable state.

[0020] As a preferred method, the raw multi-source data is subjected to data object identification and classification to obtain classified carbon data, specifically including:

[0021] Based on the boundaries of steel enterprises, process boundaries, emission source catalogs, medium catalogs, product catalogs, and by-product catalogs, identify data objects belonging to the carbon metering review scenario in the original multi-source data to obtain the identified carbon data.

[0022] A carbon data object classification system is constructed, which includes basic attribute carbon data, measurement and value-based carbon data, accounting support carbon data, evidence and traceability carbon data, and governance status carbon data. Basic attribute carbon data includes enterprise boundaries, process boundaries, emission sources, facilities, media, products, by-products, and solid waste destinations. Measurement and value-based carbon data includes flow rate, weight, electricity, steam volume, composition, temperature, pressure, laboratory values, and inventory values. Accounting support carbon data includes activity levels, emission factors, carbon content, inventory changes, material destinations, and product destinations. Evidence and traceability carbon data includes original records, instrument certificates, inspection reports, system documents, responsible departments, responsible positions, generation time, and modification records. Governance status carbon data includes whether there are source conflicts, whether the evidence is complete, whether responsibilities are clearly defined, and whether the rectification loop has been completed.

[0023] The identified carbon data is classified according to the carbon data object classification system to obtain the classified carbon data.

[0024] Preferably, carbon data from different sources but pointing to the same process, the same emission source, or the same medium are merged to generate a carbon data governance unit, specifically including:

[0025] In the categories of basic attribute carbon data, measurement value carbon data, accounting support carbon data, evidence and traceability carbon data, and governance status carbon data, the classified carbon data pointing to the same process, the same emission source, or the same medium are merged to obtain several groups of merged carbon data.

[0026] Each set of merged carbon data constitutes a carbon data governance unit. Each carbon data governance unit includes: process identifier, emission source identifier, medium identifier, time window identifier, data source identifier, value retrieval method identifier, evidence chain identifier, responsible entity identifier, and governance status identifier.

[0027] Preferably, the carbon data governance unit is preprocessed according to the process type, value retrieval method, time granularity, source quantity, and inventory changes to obtain a preprocessed carbon data governance unit, specifically including:

[0028] Preprocessing includes basic preprocessing and condition-triggered special preprocessing. The basic preprocessing steps include time synchronization, unit unification, format normalization, and source merging. Basic preprocessing is performed on all carbon data governance units to obtain intermediate carbon data governance units.

[0029] Condition-triggered special preprocessing specifically includes:

[0030] In response to the determination that the intermediate carbon data governance unit corresponds to a continuous process and that asynchronous sampling exists, process event anchor point alignment is performed to obtain the preprocessed carbon data governance unit.

[0031] In response to the determination that the intermediate carbon data governance unit involves both continuous process data and batch process data, a dual index mapping is performed to obtain the preprocessed carbon data governance unit.

[0032] In response to the determination that there are two or more sources for the same intermediate carbon data governance unit, source priority fusion and multi-source three-party cross-alignment processing are performed to obtain the preprocessed carbon data governance unit.

[0033] In response to the determination that the intermediate carbon data governance unit includes inventory change data, wherein the inventory change data includes gas holders, raw material and fuel warehouses, and lime warehouses, inventory constraint completion is performed to obtain the preprocessed carbon data governance unit.

[0034] In response to determining the start-up or shutdown of the equipment corresponding to the intermediate carbon data governance unit, load fluctuations or process switching processes are performed, and operating condition segmentation and adaptive threshold processing of operating condition segments are executed to obtain the preprocessed carbon data governance unit.

[0035] Specifically, process event anchor point alignment includes:

[0036] Get the set of process event timestamps The process event timestamps include shift change time, furnace start time, furnace end time, blast furnace tapping time, converter blowing start time, converter blowing end time, electric furnace energization start time, electric furnace tapping time, gas holder switchover time, and equipment start / stop time. n represents the number of process event timestamps.

[0037] For continuous time series data, the sampling time t is mapped to the process event anchor point with the smallest time distance and within a preset tolerance interval. , ;

[0038] For batch or furnace type data, the batch or furnace type data will be directly assigned to the process event interval with the highest overlap with the start and end time of the batch.

[0039] For other carbon data in the intermediate carbon data governance unit, when the time distance between the sampling time t and any process event anchor point exceeds the preset tolerance range, the other carbon data will be retained in the original time coordinate and marked as an unanchored unit.

[0040] Dual index mapping specifically includes:

[0041] Construct a dual-index structure, which includes a time index T and a process batch index B. The time index T is established in units of fixed time windows, while the process batch index B is established in units of furnace number, casting number, ladle number, or batch number.

[0042] For intermediate carbon data governance units, the time window to which the carbon data belongs is first determined based on the time index, and then the process batch to which the carbon data belongs is determined based on the process batch index.

[0043] For continuous data that cannot be directly assigned to a batch, it is mapped to the corresponding batch through process event anchors and production allocation coefficients.

[0044] Source priority fusion and multi-source three-party cross-alignment processing specifically include:

[0045] For the same intermediate carbon data governance unit, the primary source is selected according to the priority chain of direct measurement, production records, laboratory testing, manual ledgers, and financial settlement, while the lower priority sources are retained as verification copies.

[0046] For the same intermediate carbon data governance unit, the primary source data, ledger source data, and settlement or voucher source data are aligned three-way, and the relative deviation index between the sources is calculated, as shown in the following formula:

[0047] ;

[0048] in, The index representing the relative deviation between the i-th source value and the j-th source value within the same intermediate carbon data governance unit; This represents the value of the i-th source within the same time window; This represents the value of the j-th source within the same time window; This represents the absolute difference between two source values; Indicates taking , The maximum value in the positive correction term ε is used as the normalized denominator; ε represents the preset positive correction term to prevent the denominator from being zero;

[0049] Inventory constraint completion specifically includes:

[0050] When the primary source of inventory change data is missing within a certain time window, the missing amount is calculated as shown in the following formula:

[0051] ;

[0052] in, Indicates the amount of missing data. Indicates beginning inventory. Indicates upstream output. Indicates the amount purchased. Represents ending inventory. This indicates downstream consumption. Indicates export volume;

[0053] If the missing amount is not within the preset reasonable range, the completion will not be performed, and the inventory change data will be marked as an unrepairable anomaly; otherwise, the completion will be performed based on the missing amount.

[0054] The specific steps of working condition segmentation and adaptive threshold processing for working condition segments include:

[0055] Based on equipment start-up and shutdown, production load, furnace status, gas holder inventory changes, and steam pipeline switching status, carbon data is divided into stable operating conditions, switching operating conditions, shutdown operating conditions, and abnormal operating conditions. The stable operating conditions refer to the period when equipment operation is continuous and stable, key process parameters fluctuate within a preset stable range, and the rate of change in output and energy consumption is below the stable threshold. The switching operating conditions refer to the period during shift changes, furnace changes, raw material changes, gas holder changes, steam changes, or rapid equipment load adjustments. The shutdown operating conditions refer to the period when major equipment is shut down, there is no effective output, and the flow rate of key media remains below the shutdown threshold. The abnormal operating conditions refer to the period when equipment failures, sensor malfunctions, prolonged data interruptions, parameter mutations, or residual abnormalities exceeding the threshold occur.

[0056] Different integrity thresholds, continuity deviation thresholds, and source deviation thresholds are configured for different operating conditions. Data is then supplemented based on the operating condition in which the carbon data is located and the corresponding integrity threshold, continuity deviation threshold, and source deviation threshold.

[0057] As a preferred option, the processing procedure of the governance status diagnosis model is as follows:

[0058] The overall quality score of the carbon data governance unit is calculated using the following formula:

[0059] ;

[0060] in, This represents the overall quality score of the u-th carbon data governance unit; The integrity sub-score is calculated based on the ratio of the expected number of samples to the actual number of samples. The range reasonableness score is calculated by combining the process allowable range, the instrument range range, and the historical stable operation quantile range. The continuous sub-score is calculated based on the deviation of the time interval between adjacent records and the number of breakpoints within the sliding window. The traceability score is calculated based on the completeness of the original record, generation time, responsible party, and modification traces. The sub-score for evidence completeness is calculated based on the binding of certificates, ledgers, inspection reports, and institutional materials. The governance consistency sub-score is calculated based on the degree of consistency in process attribution, emission source attribution, medium attribution, and time attribution. , , , , and To preset weights, ;

[0061] Calculate the relative deviation index of carbon data governance units ;

[0062] The residuals of the process constraints for calculating the activity level data of the carbon data governance unit are calculated using the following formula:

[0063] ;

[0064] in, This represents the residual vector within the time window t; A represents a column vector consisting of all activity level data within a time window t; A represents the process constraint matrix, and A represents the coefficients in the material balance equation or energy balance equation. Each row corresponds to a balance equation, with the coefficient of input or output terms being +1 and the coefficient of consumption or output terms being -1. Represents a vector of constant terms;

[0065] The diagnostic criteria include whether the classification is correct, whether the integration is complete, whether the sources are conflicting, whether the evidence is complete, whether the responsible party is clear, whether the time attribution is correct, whether the process attribution is correct, and whether the verification and rectification loop has been completed. These criteria are: completeness sub-score, scope reasonableness sub-score, continuity sub-score, traceability sub-score, evidence completeness sub-score, governance consistency sub-score, relative deviation index, and process constraint residuals. Anomaly judgment thresholds are set for these criteria.

[0066] The integrity sub-score, scope reasonableness sub-score, continuity sub-score, traceability sub-score, evidence integrity sub-score, governance consistency sub-score, relative deviation index, and process constraint residual are compared with their corresponding anomaly judgment thresholds to obtain the comparison results.

[0067] Based on the set diagnostic content and comparison results, determine whether there are any anomalies in the carbon data governance unit, and mark the carbon data governance unit to obtain the marking results. The markings include missing anomalies, range anomalies, jump anomalies, source conflict anomalies, link interruption anomalies, evidence missing anomalies, process logic anomalies, and governance integrity.

[0068] As a preferred embodiment, the preprocessed carbon data governance unit undergoes metrological support verification and accounting support verification to obtain the first verification result, which specifically includes:

[0069] The preprocessed carbon data governance unit determines whether the relevant measuring instruments are equipped, whether they are within the verification or calibration validity period, whether the range and accuracy are suitable, and whether the traceability chain of the measurement value is complete. If the relevant measuring instruments are equipped, are within the verification or calibration validity period, the range and accuracy are suitable, and the traceability chain of the measurement value is complete, the metrological support verification passes; otherwise, the metrological support verification fails.

[0070] The specific verification of accounting support includes:

[0071] Based on the actual production boundaries of steel enterprises, coking, sintering or pelletizing, blast furnace, converter or electric furnace, lime kiln, self-owned power plant, gas holder, steam pipeline network and external purchase and sales nodes are abstracted as process control body nodes, and a directed network is formed by the medium or material flow path to obtain the process control body network. Each process control body node is established with a one-to-one mapping relationship with the pre-processed carbon data governance unit.

[0072] The carbon flow rate of the control volume is calculated based on the process control volume network, as shown in the following formula:

[0073] ;

[0074] in, This represents the carbon flow rate of the carrying medium or material m within the time window t; This indicates the quantity of medium or material m within the time window t; This indicates the carbon content coefficient, carbon mass fraction, or equivalent carbon factor corresponding to the medium or material;

[0075] For each process control node v, the carbon balance residual is calculated within the time window t, as shown in the following formula:

[0076] ;

[0077] in, Indicates the carbon balance residual; This represents the total amount of carbon elements flowing into control node v; This represents the carbon content converted from inventory changes, which is the difference between the carbon content of the inventory at the beginning of the time window and the carbon content of the inventory at the end of the time window. This represents the total amount of carbon elements flowing out of the control volume node v; Indicates the amount of carbon emitted into the atmosphere; Indicates the amount of carbon entering the product, by-product, or solid waste destination;

[0078] The carbon balance residual rate of control node v within time window t is calculated based on the absolute value of the carbon balance residual and the total amount of carbon flowing into control node v, as shown in the following formula:

[0079] ;

[0080] in, Indicates the carbon balance residual rate. To prevent the use of a pre-defined positive correction term when the denominator is zero;

[0081] like If the residual rate exceeds the preset threshold, it is determined that the preprocessed carbon data governance unit corresponding to the process control unit node has a risk of non-compliance in the accounting process, the accounting support verification fails, and the preprocessed carbon data governance unit is recalculated using alternative sources and the deviation contribution rate is quantified; otherwise, the accounting support verification passes. The specific process of recalculating alternative sources and quantifying the deviation contribution rate includes:

[0082] For the activity level data involved in the accounting, the substitution source values ​​are called for recalculation to obtain the substitution residual of the kth suspicious variable, denoted as: ;

[0083] The variable correction benefit is calculated as shown in the following formula:

[0084] ;

[0085] in, Indicates the variable-adjusted return value. This represents the residual obtained by recalculating the activity level data without invoking alternative source values;

[0086] Normalizing the returns of each variable yields the deviation contribution rate, as shown in the following formula:

[0087] ;

[0088] in, The bias contribution rate of the kth suspicious variable within the time window t. This represents the sum of the variable-corrected returns of all questionable variables within the time window t;

[0089] If the measurement support verification and accounting support verification of the preprocessed carbon data governance unit are both passed, then the first verification result is that the preprocessed carbon data governance unit can be used for carbon measurement review; otherwise, the first verification result is that the preprocessed carbon data governance unit cannot be used for carbon measurement review.

[0090] As a preferred option, the accounting support verification also includes cross-validation of material balance and energy balance, and identification of boundary omissions and misjudgments of emission sources;

[0091] The process of cross-validating material and energy balance includes:

[0092] When the carbon balance residual exceeds the preset carbon balance residual threshold or the carbon balance residual rate exceeds the preset residual rate threshold, the material balance residual and energy balance residual will be calculated in the corresponding process of the preprocessed carbon data governance unit corresponding to the carbon balance residual.

[0093] If the carbon balance residual exceeds a preset carbon balance residual threshold and the material balance residual exceeds a preset material balance residual threshold, then the activity level value is determined to be incorrect or the accounting boundary is omitted.

[0094] In response to the determination that the carbon balance residual exceeds the preset carbon balance residual threshold and the energy balance residual exceeds the preset first energy balance threshold, the input gas quantity, calorific value or component value is determined to be abnormal.

[0095] In response to the determination that the carbon balance residual exceeds a preset carbon balance residual threshold and the energy balance residual is less than a preset second energy balance threshold, the classification of emission factor, carbon content or carbon destination is determined to be incorrect.

[0096] The process for handling boundary omissions and misidentification of emission sources specifically includes:

[0097] Establish a process-emission source-medium-destination template to obtain a preset template;

[0098] If a process control node exhibits persistent residuals in the same direction and its carbon flow pattern matches a preset template, then a risk of boundary omission is identified.

[0099] In response to determining that a carbon stream is recorded as a product or by-product destination, and that the flow characteristics of the carbon stream conform to the fuel combustion or by-product gas emission pattern, it is determined that there is a risk of misjudgment of the emission source.

[0100] As a preferred option, the problem list includes carbon data classification errors, carbon data source conflicts, missing carbon data evidence, incomplete carbon data governance status, insufficient measurement support, abnormal process logic, misjudgment of emission sources, incorrect activity level values, and mismatch in emission factor selection.

[0101] The rectification path includes supplementing evidence, correcting classification relationships, reconstructing mappings, replacing sources, and re-verifying or recalculating.

[0102] As a preferred approach, the review and verification process includes classification verification, evidence binding verification, and accounting support verification.

[0103] Secondly, the present invention provides a carbon data processing device for carbon metering review scenarios in the steel industry, comprising:

[0104] The identification and classification module is configured to acquire raw multi-source data, perform data object identification and classification on the raw multi-source data, and obtain classified carbon data.

[0105] The carbon data governance unit reconstruction module is configured to merge classified carbon data from different sources that point to the same process, the same emission source, or the same medium to generate a carbon data governance unit.

[0106] The preprocessing module is configured to preprocess the carbon data governance unit according to the process type, value retrieval method, time granularity, source quantity and inventory changes of the carbon data governance unit to obtain the preprocessed carbon data governance unit.

[0107] The governance status diagnosis module is configured to build a governance status diagnosis model. The governance status diagnosis model is used to mark carbon data governance units according to the set diagnosis content and obtain the marking results. The preprocessed carbon data governance units are input into the governance status diagnosis model to obtain the first marking result.

[0108] The support verification module is configured to perform measurement support verification and accounting support verification on the preprocessed carbon data governance unit to obtain the first verification result.

[0109] The problem list and rectification path generation module is configured to generate a problem list and rectification path based on the first marking result and the first verification result;

[0110] The rectification module is configured to rectify the preprocessed carbon data governance unit according to the problem list and rectification path, so as to obtain the rectified carbon data governance unit.

[0111] The review module is configured to input the governance status diagnosis model into the rectified carbon data governance unit and review and verify it to obtain the second marking result and the second verification result. If the second marking result indicates that the governance is complete and the second verification result indicates that all are passed, the rectified carbon data governance unit is restored to the reviewable state. Otherwise, a problem list and rectification path are generated based on the second marking result and the second verification result, and the steps of the rectification module are repeated until the rectified carbon data governance unit is restored to the reviewable state.

[0112] Compared with the prior art, the present invention has the following beneficial effects:

[0113] (1) The carbon data processing method proposed in this application for the carbon measurement review scenario of the steel industry obtains a carbon data governance unit that is reviewable, traceable and manageable after identifying, classifying and integrating a large amount of production and operation data of enterprises. It also establishes a closed loop of identification-classification and integration-governance-diagnosis-rectification-review, which can provide steel enterprises with a complete closed loop from problem identification to rectification and review, and significantly improve the efficiency and pertinence of enterprises in carrying out carbon measurement review self-inspection, pre-review and rectification.

[0114] (2) The carbon data processing method proposed in this application for the carbon measurement review scenario in the steel industry takes the carbon measurement review object as the center, first constructs a carbon data governance unit with binding relationships of process, emission source, medium, time window, source, evidence chain and responsible entity, and then introduces a dual index mapping alignment, inventory constraint completion and control volume residual positioning mechanism to address the characteristics of the steel industry such as the coexistence of continuous processes and batch processes, significant inventory changes and cross-process flow of by-product gas. This forms a special technical path method suitable for the carbon measurement review scenario, which is more in line with the carbon measurement review scenario.

[0115] (3) The governance status diagnosis model proposed in this application for carbon data processing method for carbon measurement review scenario in the steel industry combines a three-level method of unit-level comprehensive quality scoring, source consistency verification based on relative deviation index and process constraint residual analysis to realize data quality compliance diagnosis of carbon data governance unit, and improve the data quality compliance diagnosis from general rule verification to a quantifiable, gradeable and traceable governance diagnosis process.

[0116] (4) The carbon data processing method proposed in this application for the carbon measurement review scenario in the steel industry further proposes measurement support verification and accounting support verification as supporting technologies for carbon data governance after the governance status diagnosis. It can further determine whether the carbon data governance unit is real, consistent and usable for review, so that the compliance diagnosis of the accounting process has an implementable mathematical framework.

[0117] (5) The carbon data processing method proposed in this application for the carbon measurement review scenario in the steel industry can simultaneously output problems such as carbon data classification errors, source conflicts, missing evidence, incomplete governance status, omission of accounting boundaries, and incorrect activity level values ​​for carbon data governance units through governance status diagnosis, measurement support verification and accounting support verification, taking into account both the data governance perspective and the review diagnosis perspective. Attached Figure Description

[0118] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0119] Figure 1 This is a flowchart illustrating a carbon data processing method for a carbon metering review scenario in the steel industry, as an embodiment of this application.

[0120] Figure 2 This is a schematic diagram illustrating the process of generating classified carbon data and carbon data governance unit from multi-source data in an embodiment of this application.

[0121] Figure 3 This is a schematic diagram of the structure of the carbon data governance unit in an embodiment of this application;

[0122] Figure 4 This is a schematic diagram illustrating the governance status diagnosis of the carbon data governance unit in an embodiment of this application.

[0123] Figure 5 This is a schematic diagram of carbon element flow balance verification and deviation positioning in the process control system network according to an embodiment of this application.

[0124] Figure 6 This is a schematic diagram illustrating the closed loop of problem generation, rectification, and review in an embodiment of this application.

[0125] Figure 7 This is a schematic diagram of a carbon data processing device for a carbon metering review scenario in the steel industry, as an embodiment of this application.

[0126] Figure 8 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0127] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0128] Figure 1 This application illustrates an embodiment of a carbon data processing method for a carbon metering review scenario in the steel industry, comprising the following steps:

[0129] S1. Obtain raw multi-source data, perform data object identification and classification on the raw multi-source data, and obtain classified carbon data.

[0130] In a specific embodiment, data object identification and classification are performed on the original multi-source data to obtain classified carbon data, specifically including:

[0131] Based on the boundaries of steel enterprises, process boundaries, emission source catalogs, medium catalogs, product catalogs, and by-product catalogs, identify data objects belonging to the carbon metering review scenario in the original multi-source data to obtain the identified carbon data.

[0132] A carbon data object classification system is constructed, which includes basic attribute carbon data, measurement and value-based carbon data, accounting support carbon data, evidence and traceability carbon data, and governance status carbon data. Basic attribute carbon data includes enterprise boundaries, process boundaries, emission sources, facilities, media, products, by-products, and solid waste destinations. Measurement and value-based carbon data includes flow rate, weight, electricity, steam volume, composition, temperature, pressure, laboratory values, and inventory values. Accounting support carbon data includes activity levels, emission factors, carbon content, inventory changes, material destinations, and product destinations. Evidence and traceability carbon data includes original records, instrument certificates, inspection reports, system documents, responsible departments, responsible positions, generation time, and modification records. Governance status carbon data includes whether there are source conflicts, whether the evidence is complete, whether responsibilities are clearly defined, and whether the rectification loop has been completed.

[0133] The identified carbon data is classified according to the carbon data object classification system to obtain the classified carbon data.

[0134] Specifically, the method proposed in this application targets the carbon measurement and review scenario for key emission units in the steel industry, and is applicable to integrated steel enterprises including long-process steel mills, short-process steel mills, and integrated steel enterprises that include by-product gas systems, self-owned power plants, gas holders, and steam pipeline networks. When implementing the method proposed in this application, it can be integrated with production DCS systems, MES systems, energy management systems, metering instrument management systems, laboratory testing systems, carbon ledger systems, and financial settlement systems. The collected relevant data will be reconstructed into carbon data governance units for classification, integration, verification, diagnosis, rectification, and review.

[0135] In a preferred embodiment, the method proposed in this application receives continuous process data at fixed time periods and receives batch process data by furnace number, casting number, ladle number, or batch number. For continuous process data, hourly or shift-level time windows are preferred; for batch process data, furnace-level or casting-level process batch indexes are preferred.

[0136] In step S1, based on the carbon data object classification system, the identified carbon data is mapped to basic attribute carbon data, measurement value carbon data, accounting support carbon data, evidence and traceability carbon data, and governance status carbon data. Data that does not fall within the scope of carbon measurement review is not included in the subsequent carbon data governance process. Therefore, the method proposed in this application does not simply collect data according to system source, but first establishes a mapping relationship between data—process—emission source—medium—review object to ensure that the subsequent steps are performed on reviewable carbon data, rather than a simple summary of general production data.

[0137] S2 merges classified carbon data from different sources that point to the same process, the same emission source, or the same medium to generate a carbon data governance unit.

[0138] In a specific embodiment, carbon data from different sources but pointing to the same process, the same emission source, or the same medium are merged to generate a carbon data governance unit, specifically including:

[0139] In the categories of basic attribute carbon data, measurement value carbon data, accounting support carbon data, evidence and traceability carbon data, and governance status carbon data, the classified carbon data pointing to the same process, the same emission source, or the same medium are merged to obtain several groups of merged carbon data.

[0140] Each set of merged carbon data constitutes a carbon data governance unit. Each carbon data governance unit includes: process identifier, emission source identifier, medium identifier, time window identifier, data source identifier, value retrieval method identifier, evidence chain identifier, responsible entity identifier, and governance status identifier.

[0141] Specifically, when the same activity level data has multiple sources in different systems, the direct measurement source is prioritized, while the laboratory source, ledger source, settlement source, etc., are used as verification copies, forming a primary and secondary governance unit structure. In the embodiments of this application, cross-system integration is performed on the carbon data governance unit, and relevant data from the production DCS system, energy management system, metering instrument management system, laboratory testing system, carbon ledger system, and financial settlement system are uniformly mapped to the carbon data governance unit; at the same time, original records, verification or calibration certificates, test reports, inventory records, system documents, responsible departments, responsible positions, generation time, and modification traces are bound to the corresponding carbon data governance unit. Through steps S1 and S2, the carbon data and supporting materials originally scattered in multiple systems are reconstructed into the smallest reviewable and traceable governance object, so that subsequent diagnosis no longer judges only the numerical value, but judges the combination of numerical value + source + evidence + responsibility. The method proposed in this application executes all subsequent diagnostic and remedial steps based on the carbon data governance unit, rather than directly judging the scattered original data points. This allows data values, sources, evidence, responsible parties, and governance status to be uniformly incorporated into the same minimum governance object for processing, thereby improving the accuracy of problem identification and remediation positioning in carbon measurement review scenarios.

[0142] S3. Based on the process type, value retrieval method, time granularity, source quantity, and inventory changes of the carbon data governance unit, the carbon data governance unit is preprocessed to obtain the preprocessed carbon data governance unit.

[0143] In a specific embodiment, the carbon data governance unit is preprocessed according to its process type, value retrieval method, time granularity, source quantity, and inventory changes to obtain a preprocessed carbon data governance unit, specifically including:

[0144] Preprocessing includes basic preprocessing and condition-triggered special preprocessing. The basic preprocessing steps include time synchronization, unit unification, format normalization, and source merging. Basic preprocessing is performed on all carbon data governance units to obtain intermediate carbon data governance units.

[0145] Condition-triggered special preprocessing specifically includes:

[0146] In response to the determination that the intermediate carbon data governance unit corresponds to a continuous process and that asynchronous sampling exists, process event anchor point alignment is performed to obtain the preprocessed carbon data governance unit.

[0147] In response to the determination that the intermediate carbon data governance unit involves both continuous process data and batch process data, a dual index mapping is performed to obtain the preprocessed carbon data governance unit.

[0148] In response to the determination that there are two or more sources for the same intermediate carbon data governance unit, source priority fusion and multi-source three-party cross-alignment processing are performed to obtain the preprocessed carbon data governance unit.

[0149] In response to the determination that the intermediate carbon data governance unit includes inventory change data, wherein the inventory change data includes gas holders, raw material and fuel warehouses, and lime warehouses, inventory constraint completion is performed to obtain the preprocessed carbon data governance unit.

[0150] In response to determining the start-up or shutdown of the equipment corresponding to the intermediate carbon data governance unit, load fluctuations or process switching processes are performed, and operating condition segmentation and adaptive threshold processing of operating condition segments are executed to obtain the preprocessed carbon data governance unit.

[0151] Specifically, process event anchor point alignment includes:

[0152] Get the set of process event timestamps The process event timestamps include shift change time, furnace start time, furnace end time, blast furnace tapping time, converter blowing start time, converter blowing end time, electric furnace energization start time, electric furnace tapping time, gas holder switchover time, and equipment start / stop time. n represents the number of process event timestamps.

[0153] For continuous time series data, the sampling time t is mapped to the process event anchor point with the smallest time distance and within a preset tolerance interval. , ;

[0154] For batch or furnace type data, the batch or furnace type data will be directly assigned to the process event interval with the highest overlap with the start and end time of the batch.

[0155] For other carbon data in the intermediate carbon data governance unit, when the time distance between the sampling time t and any process event anchor point exceeds the preset tolerance range, the other carbon data will be retained in the original time coordinate and marked as an unanchored unit.

[0156] Dual index mapping specifically includes:

[0157] Construct a dual-index structure, which includes a time index T and a process batch index B. The time index T is established in units of fixed time windows, while the process batch index B is established in units of furnace number, casting number, ladle number, or batch number.

[0158] For intermediate carbon data governance units, the time window to which the carbon data belongs is first determined based on the time index, and then the process batch to which the carbon data belongs is determined based on the process batch index.

[0159] For continuous data that cannot be directly assigned to a batch, it is mapped to the corresponding batch through process event anchors and production allocation coefficients.

[0160] Source priority fusion and multi-source three-party cross-alignment processing specifically include:

[0161] For the same intermediate carbon data governance unit, the primary source is selected according to the priority chain of direct measurement, production records, laboratory testing, manual ledgers, and financial settlement, while the lower priority sources are retained as verification copies.

[0162] For the same intermediate carbon data governance unit, the primary source data, ledger source data, and settlement or voucher source data are aligned three-way, and the relative deviation index between the sources is calculated, as shown in the following formula:

[0163] ;

[0164] in, The index representing the relative deviation between the i-th source value and the j-th source value within the same intermediate carbon data governance unit; This represents the value of the i-th source within the same time window; This represents the value of the j-th source within the same time window; This represents the absolute difference between two source values; Indicates taking , The maximum value in the positive correction term ε is used as the normalized denominator; ε represents the preset positive correction term to prevent the denominator from being zero;

[0165] Inventory constraint completion specifically includes:

[0166] When the primary source of inventory change data is missing within a certain time window, the missing amount is calculated as shown in the following formula:

[0167] ;

[0168] in, Indicates the amount of missing data. Indicates beginning inventory. Indicates upstream output. Indicates the amount purchased. Represents ending inventory. This indicates downstream consumption. Indicates export volume;

[0169] If the missing amount is not within the preset reasonable range, the completion will not be performed, and the inventory change data will be marked as an unrepairable anomaly; otherwise, the completion will be performed based on the missing amount.

[0170] The specific steps of working condition segmentation and adaptive threshold processing for working condition segments include:

[0171] Based on equipment start-up and shutdown, production load, furnace status, gas holder inventory changes, and steam pipeline switching status, carbon data is divided into stable operating conditions, switching operating conditions, shutdown operating conditions, and abnormal operating conditions. The stable operating conditions refer to the period when equipment operation is continuous and stable, key process parameters fluctuate within a preset stable range, and the rate of change in output and energy consumption is below the stable threshold. The switching operating conditions refer to the period during shift changes, furnace changes, raw material changes, gas holder changes, steam changes, or rapid equipment load adjustments. The shutdown operating conditions refer to the period when major equipment is shut down, there is no effective output, and the flow rate of key media remains below the shutdown threshold. The abnormal operating conditions refer to the period when equipment failures, sensor malfunctions, prolonged data interruptions, parameter mutations, or residual abnormalities exceeding the threshold occur.

[0172] Different integrity thresholds, continuity deviation thresholds, and source deviation thresholds are configured for different operating conditions. Data is then supplemented based on the operating condition in which the carbon data is located and the corresponding integrity threshold, continuity deviation threshold, and source deviation threshold.

[0173] Specifically, the preprocessing of carbon data governance units does not involve applying the same processing to all carbon data governance units. Instead, it employs a two-tiered mechanism combining basic preprocessing and condition-triggered dedicated preprocessing. This application addresses the challenge of unified governance of multi-source, multi-granularity data in the steel industry, which involves both continuous and batch processes, through process event anchor point alignment, dual-index mapping, source priority fusion, and inventory constraint completion.

[0174] Specifically, the method proposed in this application addresses the problem of coexistence of continuous and batch processes and inconsistent sampling granularity in steel enterprises. It establishes a dual-index mapping mechanism combining time index and process batch index, and integrates process event anchor points to achieve cross-source and cross-granularity data alignment. This resolves the difficulty in unifying the attribution of continuous data from blast furnaces, coking plants, and self-owned power plants with batch data from converters, electric furnaces, and continuous casting. For objects with inventory changes, such as gas holders, raw material and fuel warehouses, and lime warehouses, an inventory constraint-based solution is used to complete missing values. For different operating conditions such as equipment start-up and shutdown, load fluctuations, and medium switching, a condition segmentation and adaptive threshold processing mechanism is adopted to avoid misjudging normal process switching as abnormal.

[0175] Furthermore, for multi-source three-way cross-alignment processing, when the relative deviation index between sources... When the deviation index threshold is exceeded, the corresponding carbon data governance unit is marked as a source conflict candidate unit and then passed into the subsequent governance status diagnosis model.

[0176] For adaptive threshold processing of operating conditions, different integrity thresholds, continuity deviation thresholds, and source deviation thresholds are configured for different operating conditions, enabling the preprocessing of the carbon data governance unit to adapt to the fluctuation characteristics under different operating conditions in the steel industry. In the embodiments of this application, the integrity threshold, continuity deviation threshold, and source deviation threshold corresponding to the stable operating condition are greater than the integrity threshold, continuity deviation threshold, and source deviation threshold corresponding to the switching operating condition, so as to avoid misjudging normal switching as abnormal; the shutdown operating condition focuses on verifying shutdown consistency, and no longer uses continuity breakpoints as the main basis for anomalies; the abnormal operating condition adopts the most stringent traceability strategy, and simultaneously marks missing, source conflict, and process constraint residuals.

[0177] In the embodiments of this application, the preprocessing of the carbon data governance unit further includes sliding window jump determination, used to analyze the continuity of carbon data and further determine the degree of deviation of the time interval between adjacent records and the number of breakpoints within the sliding window. The sliding window jump determination specifically includes: for continuous time series data, simultaneously calculating the first-order difference, second-order difference, and z-score within a preset sliding window; when the current sampling point simultaneously meets at least two of the following conditions—first-order difference exceeding the threshold, second-order difference exceeding the threshold, or z-score exceeding the threshold—the continuous time series data is identified as a jump anomaly to avoid misjudging normal operating condition disturbances as anomalies. Specifically, the formula for calculating the z-score is as follows:

[0178] ;

[0179] in, This represents the sampled value at the current time t. This represents the mean of the sampled values ​​within the sliding window w. This represents the standard deviation of the sampled values ​​within the sliding window w, where w represents the length of the sliding window. When z exceeds a preset z-score threshold, it indicates that the current sampled value deviates abnormally from the historical level within the most recent window.

[0180] Furthermore, for continuous time series data, the sliding window length w is preferably 3 to 12 sampling points. When the sampling period is 1 hour, a 3-hour, 6-hour, or 12-hour window is preferred. When the window standard deviation... If the value is less than the preset minimum, the z-score will no longer be calculated. Instead, the first and second differences will be used to determine whether there is an abnormal jump, so as to avoid situations where the denominator is zero or abnormally amplified.

[0181] For inventory constraint completion, execution is only performed if the following conditions are met simultaneously:

[0182] (1) At least one of the beginning inventory or ending inventory is available;

[0183] (2) At least two of the following sources of value are available: upstream output, downstream consumption, external purchases, and external sales.

[0184] (3) The solution falls within the intersection of the instrument range, the process allowable range, and the historical stable operation quantile range.

[0185] If the above conditions are not met, automatic completion will not be performed. Instead, the carbon data governance unit will be marked as an unrepairable anomaly and transferred to the manual review process.

[0186] The carbon data processing method proposed in this application for the carbon measurement and review scenario in the steel industry identifies, classifies, and integrates a large amount of enterprise production and operation data to obtain a carbon data governance unit that is auditable, traceable, and governable. Specifically, unlike existing data processing solutions that only target the status diagnosis of measuring instruments, general data cleaning, or general carbon accounting, this application does not simply superimpose existing rules. Instead, it centers on the carbon measurement and review object, first constructing a carbon data governance unit with binding relationships between processes, emission sources, media, time windows, sources, evidence chains, and responsible entities. Then, considering the characteristics of the steel industry, such as the coexistence of continuous and batch processes, significant inventory changes, and the cross-process flow of by-product gas, it introduces a dual-index mapping alignment, inventory constraint completion, and control volume residual positioning mechanism, thereby forming a dedicated technical path suitable for the carbon measurement and review scenario. By establishing a vertical carbon data governance method centered on classification, integration, governance, and evidence binding, it is more suitable for the carbon measurement and review scenario.

[0187] S4. Construct a governance status diagnostic model. The governance status diagnostic model is used to mark carbon data governance units according to the set diagnostic content and obtain the marking results. Input the preprocessed carbon data governance units into the governance status diagnostic model to obtain the first marking result.

[0188] In a specific embodiment, the processing procedure of the governance status diagnosis model is as follows:

[0189] The overall quality score of the carbon data governance unit is calculated using the following formula:

[0190] ;

[0191] in, This represents the overall quality score of the u-th carbon data governance unit; The integrity sub-score is calculated based on the ratio of the expected number of samples to the actual number of samples. The range reasonableness score is calculated by combining the process allowable range, the instrument range range, and the historical stable operation quantile range. The continuous sub-score is calculated based on the deviation of the time interval between adjacent records and the number of breakpoints within the sliding window. The traceability score is calculated based on the completeness of the original record, generation time, responsible party, and modification traces. The sub-score for evidence completeness is calculated based on the binding of certificates, ledgers, inspection reports, and institutional materials. The governance consistency sub-score is calculated based on the degree of consistency in process attribution, emission source attribution, medium attribution, and time attribution. , , , , and To preset weights, ;

[0192] Calculate the relative deviation index of carbon data governance units ;

[0193] The residuals of the process constraints for calculating the activity level data of the carbon data governance unit are calculated using the following formula:

[0194] ;

[0195] in, This represents the residual vector within the time window t; A represents a column vector consisting of all activity level data within a time window t; A represents the process constraint matrix, and A represents the coefficients in the material balance equation or energy balance equation. Each row corresponds to a balance equation, with the coefficient of input or output terms being +1 and the coefficient of consumption or output terms being -1. Represents a vector of constant terms;

[0196] The diagnostic criteria include whether the classification is correct, whether the integration is complete, whether the sources are conflicting, whether the evidence is complete, whether the responsible party is clear, whether the time attribution is correct, whether the process attribution is correct, and whether the verification and rectification loop has been completed. These criteria are: completeness sub-score, scope reasonableness sub-score, continuity sub-score, traceability sub-score, evidence completeness sub-score, governance consistency sub-score, relative deviation index, and process constraint residuals. Anomaly judgment thresholds are set for these criteria.

[0197] The integrity sub-score, scope reasonableness sub-score, continuity sub-score, traceability sub-score, evidence integrity sub-score, governance consistency sub-score, relative deviation index, and process constraint residual are compared with their corresponding anomaly judgment thresholds to obtain the comparison results.

[0198] Based on the set diagnostic content and comparison results, determine whether there are any anomalies in the carbon data governance unit, and mark the carbon data governance unit to obtain the marking results. The markings include missing anomalies, range anomalies, jump anomalies, source conflict anomalies, link interruption anomalies, evidence missing anomalies, process logic anomalies, and governance integrity.

[0199] Specifically, the governance status diagnosis model constructed in step S4 combines a three-level approach: unit-level comprehensive quality scoring, source consistency verification based on relative deviation index, and process constraint residual analysis. This enables data quality compliance diagnosis of carbon data governance units, elevating data quality compliance diagnosis from general rule verification to a quantifiable, gradable, and traceable governance diagnosis process.

[0200] In a preferred embodiment, the preset weights for the overall quality score can be set as: integrity weight. Scope of reasonableness weight Continuity weight Traceability weight Weight of Evidence Integrity Governance Consistency Weight When a carbon data governance unit is classified as a high-risk review target, the weighting of evidence integrity and traceability can be increased.

[0201] When performing source consistency verification based on relative deviation index: For multiple source data of the same carbon data object, calculate the relative deviation index between the sources. When the relative deviation index exceeds the preset deviation index threshold and lasts for more than K time windows, it is determined that there is a source conflict anomaly in the carbon data governance unit. Further, the source of the anomaly is determined by combining the status of the main source instrument, the ledger entry status and the settlement data time lag.

[0202] In the embodiments of this application, the deviation index threshold is preferably determined based on the medium type, metering accuracy level, and historical stable operation statistics: (1) For high-stability, continuous metering data, the deviation threshold is preferably set to 0.01 to 0.03; (2) For data involving conversion or manual input, the deviation threshold is preferably set to 0.03 to 0.08; (3) For switching operating conditions or shutdown operating conditions, the upper limit of the above threshold can be appropriately relaxed. For example, the deviation index threshold for continuous metering data of by-product gas is set to 0.03, and it is required that the source conflict anomaly is determined only after K=3 consecutive time windows exceed the threshold, so as to reduce the probability of false alarms caused by occasional fluctuations.

[0203] When performing process constraint residual analysis: For carbon data governance units related to coal gas, steam, electricity, raw materials, and main products and by-products, process constraint residuals are calculated to identify process logic inconsistencies. When any component in the residual vector exceeds the corresponding threshold, it is determined that there is a process logic anomaly in the carbon data governance unit, and the set of suspicious variables is narrowed down according to the residual direction.

[0204] In the embodiments of this application, the specific meanings of each marker are as follows: Missing Anomaly: Theoretically, a record should exist, but the actual record is missing, and the duration or proportion of the missing record exceeds a preset threshold; Range Anomaly: The sampled value exceeds the instrument's range or the process's allowable range; Jump Anomaly: The sampled value simultaneously satisfies at least two of the following within the sliding window: first-order difference exceeding the threshold, second-order difference exceeding the threshold, and z-score anomaly; Source Conflict Anomaly: The deviation index between the main source and the secondary source of the same governance unit exceeds the threshold for K consecutive time windows; Link Interruption Anomaly: The data acquisition chain is interrupted, resulting in no valid uploaded records within a consecutive time window; Evidence Missing Anomaly: Unable to bind the original record, certificate, responsible entity, or generation time; Process Logic Anomaly: The process constraint residual or the carbon balance residual of the control body continuously exceeds the threshold. Furthermore, based on the duration of the anomaly, the scope of its impact, and the size of the residual, the markers can be further refined into three levels: mild, moderate, and severe. When an anomaly lasts for a short period and its impact is limited to a single governance unit, it is classified as mild; when an anomaly spans multiple time windows or affects multiple governance units in the same process, it is classified as moderate; when an anomaly renders a governance unit unreviewable or directly affects the accounting conclusion, it is classified as severe.

[0205] In embodiments of this application, the labeling of carbon data governance units also includes unclear attribution, governance gaps, and requiring verification. For example, carbon data governance units whose original records, responsible parties, or generation times cannot be traced are labeled as units lacking evidence; carbon data governance units whose corresponding processes or emission sources cannot be clearly identified are labeled as units with unclear attribution.

[0206] S5, perform metrological support verification and accounting support verification on the preprocessed carbon data governance unit to obtain the first verification result.

[0207] In a specific embodiment, the preprocessed carbon data governance unit undergoes metrological support verification and accounting support verification to obtain a first verification result, which specifically includes:

[0208] The preprocessed carbon data governance unit determines whether the relevant measuring instruments are equipped, whether they are within the verification or calibration validity period, whether the range and accuracy are suitable, and whether the traceability chain of the measurement value is complete. If the relevant measuring instruments are equipped, are within the verification or calibration validity period, the range and accuracy are suitable, and the traceability chain of the measurement value is complete, the metrological support verification passes; otherwise, the metrological support verification fails.

[0209] The specific verification of accounting support includes:

[0210] Based on the actual production boundaries of steel enterprises, coking, sintering or pelletizing, blast furnace, converter or electric furnace, lime kiln, self-owned power plant, gas holder, steam pipeline network and external purchase and sales nodes are abstracted as process control body nodes, and a directed network is formed by the medium or material flow path to obtain the process control body network. Each process control body node is established with a one-to-one mapping relationship with the pre-processed carbon data governance unit.

[0211] The carbon flow rate of the control volume is calculated based on the process control volume network, as shown in the following formula:

[0212] ;

[0213] in, This represents the carbon flow rate of the carrying medium or material m within the time window t; This indicates the quantity of medium or material m within the time window t; This indicates the carbon content coefficient, carbon mass fraction, or equivalent carbon factor corresponding to the medium or material;

[0214] For each process control node v, the carbon balance residual is calculated within the time window t, as shown in the following formula:

[0215] ;

[0216] in, Indicates the carbon balance residual; This represents the total amount of carbon elements flowing into control node v; This represents the carbon content converted from inventory changes, which is the difference between the carbon content of the inventory at the beginning of the time window and the carbon content of the inventory at the end of the time window. This represents the total amount of carbon elements flowing out of the control volume node v; Indicates the amount of carbon emitted into the atmosphere; Indicates the amount of carbon entering the product, by-product, or solid waste destination;

[0217] The carbon balance residual rate of control node v within time window t is calculated based on the absolute value of the carbon balance residual and the total amount of carbon flowing into control node v, as shown in the following formula:

[0218] ;

[0219] in, Indicates the carbon balance residual rate. To prevent the use of a pre-defined positive correction term when the denominator is zero;

[0220] like If the residual rate exceeds the preset threshold, it is determined that the preprocessed carbon data governance unit corresponding to the process control unit node has a risk of non-compliance in the accounting process, the accounting support verification fails, and the preprocessed carbon data governance unit is recalculated using alternative sources and the deviation contribution rate is quantified; otherwise, the accounting support verification passes. The specific process of recalculating alternative sources and quantifying the deviation contribution rate includes:

[0221] For the activity level data involved in the accounting, the substitution source values ​​are called for recalculation to obtain the substitution residual of the kth suspicious variable, denoted as: ;

[0222] The variable correction benefit is calculated as shown in the following formula:

[0223] ;

[0224] in, Indicates the variable-adjusted return value. This represents the residual obtained by recalculating the activity level data without invoking alternative source values;

[0225] Normalizing the returns of each variable yields the deviation contribution rate, as shown in the following formula:

[0226] ;

[0227] in, The bias contribution rate of the kth suspicious variable within the time window t. This represents the sum of the variable-corrected returns of all questionable variables within the time window t;

[0228] If the measurement support verification and accounting support verification of the preprocessed carbon data governance unit are both passed, then the first verification result is that the preprocessed carbon data governance unit can be used for carbon measurement review; otherwise, the first verification result is that the preprocessed carbon data governance unit cannot be used for carbon measurement review.

[0229] Specifically, measurement support verification and accounting support verification are only invoked for preprocessed carbon data governance units that still require further verification after governance status diagnosis. These verifications are used to verify the authenticity, consistency, correct attribution, and auditability of the carbon data governance units and serve as technical means to determine the governance status of carbon data.

[0230] Taking a long-process steel plant as an example, the construction of the process control body network is further explained. The process control body network is constructed based on the nodes of coking—sintering / pelletizing—blast furnace—converter—continuous casting—rolling—self-owned power plant—gas holder—steam pipeline network—purchased and sold externally. Among them, coke oven gas, blast furnace gas, converter gas, steam, electricity, limestone, coke, pulverized coal, scrap steel and molten iron are respectively used as the carrying medium or material flow edge between nodes. Each process control body node establishes a mapping relationship with the corresponding carbon data governance unit.

[0231] The method proposed in this application constructs a process control system network based on carbon data governance units. By utilizing process constraint residuals, carbon balance residuals of process control system nodes, and recalculation results of alternative sources, it is possible to locate suspicious variables and rank the contribution rates of deviations in abnormal data, thereby generating problem conclusions and rectification priorities for carbon measurement review.

[0232] Specifically, the alternative source values ​​are derived from secondary source data within the same carbon data governance unit. Secondary source data includes laboratory sources, ledger sources, settlement sources, inventory estimates, or estimates derived by inversely calculating the conservation relationships of adjacent processes. For the k-th suspected variable, the carbon balance residual at the process control node is recalculated after replacing the primary source value with its alternative source value, yielding the alternative residual. Then calculate the variable-corrected return value. ,in The larger the value, the more significant the decrease in residuals after replacing the variable, and the more likely that variable is the main source of the problem. Specifically, when... If this indicates that replacing the variable did not improve the residual, it is preferable to record the variable correction benefit value as 0 and no longer prioritize it for rectification. Then calculate the deviation contribution rate. ,in, This represents the deviation contribution rate of the k-th suspicious variable within the time window t. When multiple suspicious variables coexist, the priority of rectification is ranked according to the deviation contribution rate, which helps to determine the priority review targets. The sum of the variable correction benefits of all suspicious variables... At that time, the deviation contribution rate is no longer calculated. Instead, the time window is marked as an unlocated anomaly that requires manual review in order to avoid the denominator being zero.

[0233] The carbon data processing method proposed in this application for the carbon measurement review scenario in the steel industry calculates carbon element flow balance and residual rate in the accounting support verification. It also proposes alternative source recalculation and deviation contribution rate quantification methods, which enable compliance diagnosis of the accounting process to have an implementable mathematical framework and can directly support the identification of boundary omissions, emission source misjudgments and activity level errors.

[0234] In specific embodiments, the accounting support verification also includes cross-validation of material balance and energy balance, and identification of boundary omissions and emission source misjudgments;

[0235] The process of cross-validating material and energy balance includes:

[0236] When the carbon balance residual exceeds the preset carbon balance residual threshold or the carbon balance residual rate exceeds the preset residual rate threshold, the material balance residual and energy balance residual will be calculated in the corresponding process of the preprocessed carbon data governance unit corresponding to the carbon balance residual.

[0237] If the carbon balance residual exceeds a preset carbon balance residual threshold and the material balance residual exceeds a preset material balance residual threshold, then the activity level value is determined to be incorrect or the accounting boundary is omitted.

[0238] In response to the determination that the carbon balance residual exceeds the preset carbon balance residual threshold and the energy balance residual exceeds the preset first energy balance threshold, the input gas quantity, calorific value or component value is determined to be abnormal.

[0239] In response to the determination that the carbon balance residual exceeds a preset carbon balance residual threshold and the energy balance residual is less than a preset second energy balance threshold, the classification of emission factor, carbon content or carbon destination is determined to be incorrect.

[0240] The process for handling boundary omissions and misidentification of emission sources specifically includes:

[0241] Establish a process-emission source-medium-destination template to obtain a preset template;

[0242] If a process control node exhibits persistent residuals in the same direction and its carbon flow pattern matches a preset template, then a risk of boundary omission is identified.

[0243] In response to determining that a carbon stream is recorded as a product or by-product destination, and that the flow characteristics of the carbon stream conform to the fuel combustion or by-product gas emission pattern, it is determined that there is a risk of misjudgment of the emission source.

[0244] Specifically, material balance equations and energy balance equations are established for blast furnaces, converters, electric furnaces, lime kilns, self-owned power plants, and by-product gas systems, respectively. These are then used to identify boundary omissions, misclassification of emission sources, abnormal activity level values, and emission factor mismatches based on carbon balance results. For example, in the control system of a self-owned power plant, an energy balance relationship can be established between blast furnace gas input, converter gas input, power generation, steam output, and boiler efficiency. If both the carbon balance residual and the energy balance residual exceed their corresponding thresholds, the input gas quantity, calorific value, or component value is considered abnormal. If the carbon balance residual is large while the energy balance is essentially closed, the emission factor, carbon content, or carbon destination classification is considered incorrect.

[0245] The carbon data processing method proposed in this application for the carbon measurement and review scenario in the steel industry further proposes measurement-supported verification and accounting-supported verification as supporting technologies for carbon data governance after governance status diagnosis. These technologies can further determine whether the carbon data governance unit is authentic, consistent, and usable for review. Through governance status diagnosis, measurement-supported verification, and accounting-supported verification, the method can simultaneously output information to the carbon data governance unit regarding issues such as incorrect carbon data classification, source conflicts, missing evidence, incomplete governance status, omissions in accounting boundaries, and incorrect activity level values, thus taking into account both data governance and review diagnosis perspectives.

[0246] S6. Generate a problem list and rectification path based on the first marking result and the first verification result.

[0247] In specific implementations, the problem list includes carbon data classification errors, carbon data source conflicts, missing carbon data evidence, incomplete carbon data governance status, insufficient measurement support, abnormal process logic, misjudgment of emission sources, incorrect activity level values, and mismatched emission factor selection.

[0248] The rectification path includes supplementing evidence, correcting classification relationships, reconstructing mappings, replacing sources, and re-verifying or recalculating.

[0249] Specifically, when generating the problem list, for each problem item, the corresponding process, emission source, medium, time period, problem category, degree of impact, associated carbon data governance unit, evidence gap, possible responsible party, and rectification path are output.

[0250] S7. Based on the problem list and rectification path, rectify the preprocessed carbon data governance unit to obtain the rectified carbon data governance unit.

[0251] S8. Input the governance status diagnosis model into the rectified carbon data governance unit and review and verify it to obtain the second marking result and the second verification result. If the second marking result is that the governance is complete and the second verification result is that all are passed, the rectified carbon data governance unit is restored to the reviewable state. Otherwise, generate a problem list and rectification path based on the second marking result and the second verification result, and repeat the previous step until the rectified carbon data governance unit is restored to the reviewable state.

[0252] In specific implementations, the review and verification process includes classification verification, evidence binding verification, and accounting support verification.

[0253] Specifically, the method proposed in this application re-executes governance status diagnosis, classification verification, evidence binding verification, and control system carbon balance verification on the rectified carbon data governance units, forming a comparison record before and after rectification, until the governance status of the relevant carbon data governance units is restored to an auditable state, thereby achieving a closed loop of identification—classification and integration—governance—diagnosis—rectification—review. Therefore, the carbon data processing method proposed in this application for the carbon measurement review scenario in the steel industry can provide steel enterprises with a complete closed loop from problem identification to rectification review, significantly improving the efficiency and pertinence of enterprises' self-inspection, pre-review, and rectification improvement in carbon measurement review.

[0254] In summary, the carbon data processing method proposed in this application for the carbon measurement review scenario in the steel industry converts multi-source heterogeneous data of steel enterprises into reviewable carbon data governance objects, and further realizes classification verification, source conflict identification, evidence missing identification, process logic diagnosis and accounting support diagnosis, ultimately forming a technical solution applicable to carbon measurement review of key emission units in the steel industry.

[0255] The following example, using a long-process steel enterprise, further illustrates the carbon data processing method proposed in this application for the carbon metering review scenario in the steel industry.

[0256] A certain integrated steel enterprise includes coking, sintering, blast furnace ironmaking, converter steelmaking, continuous casting, a self-owned power plant, gas holders, and steam pipelines. Internally, the enterprise has deployed a production DCS system, an energy management system, a metering instrument management system, a laboratory testing system, a carbon ledger system, and a financial settlement system. During its daily operations, the enterprise generates a large amount of production data, metering data, energy consumption data, laboratory data, ledger data, and invoice data; however, not all of this data directly constitutes the subject of carbon measurement audits.

[0257] refer to Figure 2 When implementing the method proposed in this application, the relevance of enterprise multi-source data is first determined. Specifically, based on enterprise boundaries, process boundaries, emission source catalogs, medium catalogs, product catalogs, by-product catalogs, and the requirements for review evidence, each piece of original multi-source data is mapped and determined. Only data related to boundaries, processes, emission sources, media, value methods, and evidence requirements are identified as carbon data objects and further reconstructed into reviewable carbon data governance units.

[0258] For example, blast furnace gas production, blast furnace gas consumption in self-owned power plants, and blast furnace gas emissions are identified as metering-based carbon data; blast furnace ironmaking processes, self-owned power plant boiler emission sources, and gas holder inventory changes are identified as basic attribute carbon data; blast furnace gas component analysis values, carbon content, and inventory changes are identified as accounting support carbon data; flow meter calibration certificates, gas component inspection reports, responsible positions, generation time, and modification traces are identified as evidence-based carbon data; and whether it is traceable, whether there are source conflicts, and whether it has been reviewed are identified as governance status carbon data.

[0259] After identification and classification, data from different sources but pointing to the same process, emission source, medium, and time window are merged to generate carbon data governance units. A carbon data governance unit is not a single numerical value, but rather the smallest governance object composed of process, emission source, medium, time window, primary source, secondary source, evidence chain, responsible party, and governance status, used to carry out subsequent merging, verification, diagnosis, rectification, and review. For example, refer to... Figure 2 and Figure 3 For the scenario of blast furnace gas being supplied to a self-owned power plant, a carbon data governance unit U1 is generated within the time window of 08:00–09:00 on March 15, 2026. This unit includes a process identifier, emission source identifier, medium identifier, time window identifier, primary source value, secondary source value, value retrieval method identifier, evidence chain identifier, responsible entity identifier, and governance status identifier. The carbon data governance unit is then preprocessed to obtain the preprocessed carbon data governance unit U1.

[0260] Furthermore, the preprocessed carbon data governance unit is input into the governance status diagnosis model to perform carbon data governance status diagnosis. (Reference) Figure 4 The governance status diagnostic model revolves around the comprehensive quality score at the unit level, source consistency verification, and process constraint residual analysis. It outputs status labels such as governance completeness, governance gap, source conflict, missing evidence, unclear attribution, and need for review. It's important to note that the governance status diagnosis focuses on whether carbon data governance units have been correctly classified, whether there are source conflicts, whether the evidence is complete, and whether they are suitable for review. The following sections further explain the calculation of the comprehensive quality score for carbon data governance units, the calculation of the relative deviation index for carbon data governance units, the calculation of process constraint residuals, and the anomaly classification and grading.

[0261] First, the relative deviation index of the preprocessed carbon data governance unit is calculated to verify source consistency. The primary source value is used. , The cumulative flow rate of the gas flow meter within the time window t is taken as the value from the secondary source. , The converted flow rate value for the corresponding time window t in the energy management system ledger is calculated, and the relative deviation index between sources is calculated. If the calculated D_ij is greater than the preset threshold of 0.03 for three consecutive time windows, then the preprocessed carbon data governance unit is marked as a source conflict anomaly.

[0262] Secondly, a comprehensive quality score is calculated for the preprocessed carbon data governance unit. Specifically, if the preprocessed carbon data governance unit lacks a current verification certificate, the evidence completeness sub-score will be lower. If the score is below the set threshold; if the record exists but the responsible position information is missing, the traceability score will be lowered. The value decreases. Ultimately, this preprocessed carbon data governance unit can be marked as a unit with missing evidence.

[0263] Next, the process constraint residuals are calculated for the preprocessed carbon data governance unit to perform process constraint residual analysis. Taking the blast furnace gas system as an example, the process constraint equations are as follows:

[0264] ;

[0265] in, This represents the blast furnace gas output within the time window t. This indicates the amount consumed by the boiler. This indicates the amount of electricity consumed by sending power to the self-owned power plant. This indicates the amount of blast furnace gas emitted. This represents the change in gas holder inventory. The process constraint residuals within this time window t are shown in the following formula:

[0266] .

[0267] when When the value is significantly greater than zero, the primary suspicion should be either that downstream consumption is underreported or that inventory change records are missing; when When the value is significantly less than zero, underestimation of output, measurement anomalies, or double counting of downstream consumption should be the primary suspicions. Furthermore, by comparing the direction and magnitude of the residual changes after substitution of each variable, the set of suspected variables most likely to cause residual anomalies is determined.

[0268] Finally, based on the source consistency verification results, comprehensive quality score results, and process constraint residual analysis results, the preprocessed carbon data governance units are classified as anomalies and the governance status labels of source conflict, missing evidence, abnormal process logic, and need for review are output.

[0269] To facilitate understanding, the source consistency verification and process constraint residual analysis are illustrated with specific numerical values, specifically using the pre-processed carbon data management unit U1 corresponding to the time window of 08:00-09:00 when blast furnace gas is sent to the self-owned power plant.

[0270] The numerical derivation for the source consistency check is as follows:

[0271] Main source value (gas flow meter reading) Secondary source value (value converted from energy management system ledger) Preset positive correction term The relative deviation index between sources is calculated as follows:

[0272] .

[0273] The deviation index threshold is set at 0.03, based on the calculated... And the monitoring of the time window 07:00-08:00 Time window 08:00-09:00 Time window 09:00-10:00 The preprocessed carbon data governance unit U1 was marked as having a source conflict anomaly because D_ij=0.038 and exceeded the deviation index threshold of 0.03 for three consecutive time windows.

[0274] The residual analysis and deviation location derivation based on process constraints are as follows:

[0275] Blast furnace gas production Boiler consumption The main source value for preprocessed carbon data governance unit U1; power generation consumption sent to self-owned power plants. Blast furnace gas emission Changes in gas holder inventory Where a positive value indicates an increase in counter inventory, i.e., net inventory inflow. The process constraint residual r_t is calculated as follows:

[0276] .

[0277] Therefore, the process constraint residual within this time window This indicates that the blast furnace gas output exceeds the sum of downstream consumption, emissions, and inventory increases. Based on the process constraint residual analysis rules proposed in this application, when... When the value is significantly greater than zero, the following situations should be suspected first: 1. There is a lack of recording or under-recording of a certain downstream consumption destination; 2. The recorded increase in gas holder inventory is too low; 3. The recorded output of upstream is too high.

[0278] In the source consistency check mentioned above, there were source conflict anomalies in three consecutive time windows. Therefore, the secondary source value of boiler consumption, the record value of gas holder inventory change, and the primary source value of blast furnace gas output were further included in the set of suspicious variables, instead of directly determining a variable as the main source of the problem based on a single residual direction.

[0279] Subsequently, the method of recalculating alternative sources and quantifying the contribution rate of deviations was invoked to replace the above-mentioned suspicious variables one by one and recalculate the residuals. The decrease in residuals before and after the replacement of each variable was compared to determine the priority review targets.

[0280] As can be seen from the above calculation and analysis process of source consistency verification and process constraint residual analysis combined with specific numerical values, this invention does not draw conclusions directly based on a single deviation value or a single residual value. Instead, it first narrows down the range of suspicious variables based on the residual direction, and then completes further positioning by combining the multi-source substitution recalculation results.

[0281] Furthermore, in this embodiment, for the preprocessed carbon data governance unit U1 that still requires verification after governance status diagnosis, a measurement support verification and an accounting support verification are performed for review purposes. It should be noted that the support verification is used to verify whether the carbon data governance unit is authentic, consistent, correctly attributed, and usable for carbon measurement review.

[0282] First, the preprocessed carbon data management unit U1 undergoes metrological support verification to determine whether the corresponding gas flow meter is equipped, whether it is within its calibration validity period, whether the range is suitable, and whether the traceability chain is complete. If the flow meter has exceeded its calibration validity period, a metrological support deficiency mark is added to the preprocessed carbon data management unit U1.

[0283] Secondly, refer to Figure 5 A process control network for long-process steel plants is established, which abstracts coking, sintering / pelletizing, blast furnace, converter, gas holder, self-owned power plant, and steam pipeline network as process control nodes. Coke oven gas, blast furnace gas, converter gas, steam, electricity, coke, and molten iron are used as flow edges between nodes. Carbon element flow rate, carbon balance residual, and carbon balance residual rate are calculated, and priority suspicious variables are identified through recalculation of alternative sources and quantification of deviation contribution rate.

[0284] For the control volume containing the preprocessed carbon data governance unit U1, calculate the carbon element flow rate of the control volume within the time window t. .

[0285] Then, the carbon balance residual is calculated within the time window t. and carbon balance residual rate .

[0286] If the carbon balance residual rate of the control body containing the preprocessed carbon data governance unit U1 is... If the residual rate continuously exceeds the preset threshold, it indicates that the preprocessed carbon data governance unit U1 has a risk of non-compliance in its accounting process. In this case, secondary source data from the same preprocessed carbon data governance unit U1 is used as a substitute source value, including at least laboratory sources, ledger sources, settlement sources, inventory estimates, or estimated values ​​derived by inverse calculation based on the conservation relationships of adjacent processes. For the k-th suspicious variable, the control volume residual is recalculated after replacing the primary source value with the substitute source value, resulting in the substitute residual. According to the substitution residual. Calculate the variable correction benefit value Then adjust the return value according to the variables. Calculate the contribution rate of deviation If the composition analysis values ​​of blast furnace gas... If the value is the largest, it indicates that the variable is a priority suspect variable, with the highest contribution rate to the bias, and should be reviewed first.

[0287] Finally, the following usability review conclusions are output: The preprocessed carbon data governance unit U1 has source conflicts, an incomplete chain of evidence, and the residual rate exceeds the threshold in the control volume carbon balance verification. The priority suspicious variable is the value of blast furnace gas components. The recommended corrective actions include supplementing the test report, reviewing the test values, reconstructing the secondary source mapping relationship, and re-performing the carbon balance verification.

[0288] Furthermore, a list of issues for carbon measurement review is generated. The list includes: incorrect carbon data classification, conflicting carbon data sources, missing carbon data evidence, incomplete carbon data governance status, insufficient measurement support, abnormal process logic, omission of accounting boundaries, misjudgment of emission sources, incorrect activity level values, and mismatch in emission factor selection.

[0289] For the pre-processed carbon data governance unit U1, the output issues include: the process is blast furnace—gas holder—self-owned power plant; the emission source is the gas-fired boiler emission source of the self-owned power plant; the medium is blast furnace gas; the time period is 08:00-09:00 on March 15, 2026; the issue categories are source conflict, missing evidence, abnormal process logic, and insufficient accounting support; the impact level is moderate to severe; the evidence gaps are missing verification certificates and missing component test reports; the possible responsible parties are the energy metering position, the inspection position, and the self-owned power plant operation position; the rectification path is to supplement evidence, correct the classification relationship, replace the source, re-verify, and recalculate.

[0290] After rectification, the system re-executes classification verification, evidence binding verification, source consistency verification in governance status diagnosis, process constraint residual analysis, and accounting support verification on the pre-processed carbon data governance unit U1. If the governance status is restored to classified and auditable after rectification, and the residual rate falls back to within the residual rate threshold, the status of the pre-processed carbon data governance unit U1 is updated to restored auditable status, realizing a closed-loop process from problem list output, rectification action execution, governance unit re-verification to restoration of auditable status, such as... Figure 6 As shown. Therefore, the carbon data processing method proposed in this application for the carbon measurement review scenario in the steel industry can establish a complete closed loop of identification, classification and integration, governance, diagnosis, rectification and review, thereby significantly improving the efficiency and pertinence of enterprises in carrying out self-inspection, pre-review and rectification improvement of carbon measurement review.

[0291] The steps S1-S8 above do not necessarily represent the order of the steps, but are represented by step symbols. The order of the steps can be adjusted.

[0292] Further reference Figure 7 As an implementation of the methods shown in the above figures, this application provides an embodiment of a carbon data processing device for a carbon metering review scenario in the steel industry. This device embodiment is similar to... Figure 1 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0293] This application provides a carbon data processing device for carbon metering review scenarios in the steel industry, including:

[0294] The identification and classification module 1 is configured to acquire raw multi-source data, perform data object identification and classification on the raw multi-source data, and obtain classified carbon data.

[0295] Carbon data governance unit reconstruction module 2 is configured to merge classified carbon data from different sources but pointing to the same process, the same emission source or the same medium to generate carbon data governance units;

[0296] Preprocessing module 3 is configured to preprocess the carbon data governance unit according to the process type, value retrieval method, time granularity, source quantity and inventory changes of the carbon data governance unit to obtain the preprocessed carbon data governance unit.

[0297] The governance status diagnosis module 4 is configured to build a governance status diagnosis model. The governance status diagnosis model is used to mark carbon data governance units according to the set diagnosis content, obtain the marking results, and input the preprocessed carbon data governance units into the governance status diagnosis model to obtain the first marking result.

[0298] The support verification module 5 is configured to perform measurement support verification and accounting support verification on the preprocessed carbon data governance unit to obtain the first verification result.

[0299] The problem list and rectification path generation module 6 is configured to generate a problem list and rectification path based on the first marking result and the first verification result;

[0300] The rectification module 7 is configured to rectify the preprocessed carbon data governance unit according to the problem list and rectification path, so as to obtain the rectified carbon data governance unit.

[0301] The review module 8 is configured to input the governance status diagnosis model into the rectified carbon data governance unit and review and verify it to obtain the second marking result and the second verification result. If the second marking result indicates that the governance is complete and the second verification result indicates that all are passed, the rectified carbon data governance unit is restored to the reviewable state. Otherwise, a problem list and rectification path are generated based on the second marking result and the second verification result, and the steps of the rectification module are repeated until the rectified carbon data governance unit is restored to the reviewable state.

[0302] Figure 8 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present invention. For example... Figure 8 As shown, the electronic device of this embodiment includes a processor 801 and a memory 802; wherein the memory 802 is used to store computer execution instructions; and the processor 801 is used to execute the computer execution instructions stored in the memory to implement the various steps performed by the electronic device in the above embodiment. For details, please refer to the relevant descriptions in the foregoing method embodiments.

[0303] Alternatively, the memory 802 can be either standalone or integrated with the processor 801.

[0304] When the memory 802 is set up independently, the electronic device also includes a bus 803 for connecting the memory 802 and the processor 801.

[0305] This invention also provides a computer storage medium storing computer execution instructions, which, when executed by processor 801, implement the above method.

[0306] This invention also provides a computer program product, including a computer program that, when executed by a processor 801, implements the above-described method.

[0307] In the embodiments provided by this invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or modules, and may be electrical, mechanical, or other forms.

[0308] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to implement the solution of this embodiment according to actual needs.

[0309] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The unit composed of the above modules can be implemented in hardware or in the form of hardware plus software functional units.

[0310] The integrated modules implemented as software functional modules described above can be stored in a computer-readable storage medium. These software functional modules, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor 801 to execute some steps of the methods of the various embodiments of this application.

[0311] It should be understood that the processor 801 described above can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor, or the processor 801 can be any conventional processor 801. The steps of the method disclosed in this invention can be directly manifested as the hardware processor 801 executing the steps, or as a combination of hardware and software modules within the processor 801 executing the steps.

[0312] The memory 802 may include high-speed RAM memory, and may also include non-volatile memory NVM, such as at least one disk storage device, and may also be a USB flash drive, portable hard drive, read-only memory, disk or optical disc, etc.

[0313] Bus 803 can be an Industry Standard Architecture (ISA), a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Bus 803 can be divided into address bus, data bus, control bus, etc. For ease of illustration, the bus 803 in the accompanying drawings of this application is not limited to only one bus 803 or one type of bus 803.

[0314] The aforementioned storage medium can be implemented from any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The storage medium can be any available medium accessible to general-purpose or special-purpose computers.

[0315] An exemplary storage medium is coupled to a processor 801, enabling the processor 801 to read information from and write information to the storage medium. Alternatively, the storage medium can be an integral part of the processor 801. The processor 801 and the storage medium can reside in an application-specific integrated circuit (ASIC). Alternatively, the processor 801 and the storage medium can exist as discrete components in an electronic device or a host device.

[0316] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0317] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A carbon data processing method for carbon metering review scenarios in the steel industry, characterized in that, Includes the following steps: Obtain raw multi-source data, perform data object identification and classification on the raw multi-source data, and obtain classified carbon data; The classified carbon data from different sources but pointing to the same process, the same emission source or the same medium are merged to generate a carbon data governance unit; The carbon data governance unit is preprocessed according to the process type, value retrieval method, time granularity, source quantity and inventory changes of the carbon data governance unit to obtain the preprocessed carbon data governance unit. A governance status diagnostic model is constructed. The governance status diagnostic model is used to mark carbon data governance units according to the set diagnostic content to obtain the marking results. The preprocessed carbon data governance units are input into the governance status diagnostic model to obtain the first marking result. The preprocessed carbon data governance unit is subjected to metrological support verification and accounting support verification to obtain the first verification result; A problem list and rectification path are generated based on the first marking result and the first verification result; The preprocessed carbon data governance unit is rectified according to the problem list and the rectification path to obtain the rectified carbon data governance unit. The rectified carbon data governance unit is input into the governance status diagnosis model and reviewed and verified to obtain a second marking result and a second verification result. If the second marking result indicates that the governance is complete and the second verification result indicates that all are passed, the rectified carbon data governance unit is restored to the reviewable state. Otherwise, a problem list and rectification path are generated based on the second marking result and the second verification result, and the previous step is repeated until the rectified carbon data governance unit is restored to the reviewable state.

2. The carbon data processing method for the carbon metering review scenario in the steel industry according to claim 1, characterized in that, The step of performing data object identification and classification on the original multi-source data to obtain classified carbon data specifically includes: Based on the steel enterprise boundaries, process boundaries, emission source catalog, medium catalog, product catalog, and by-product catalog, identify the data objects belonging to the carbon metering review scenario in the original multi-source data to obtain the identified carbon data. A carbon data object classification system is constructed, comprising: basic attribute carbon data, measurement and value carbon data, accounting support carbon data, evidence and traceability carbon data, and governance status carbon data. The basic attribute carbon data includes enterprise boundaries, process boundaries, emission sources, facilities, media, products, by-products, and solid waste destinations. The measurement and value carbon data includes flow rate, weight, electricity, steam volume, composition, temperature, pressure, laboratory values, and inventory values. The accounting support carbon data includes activity levels, emission factors, carbon content, inventory changes, material destinations, and product destinations. The evidence and traceability carbon data includes original records, instrument certificates, inspection reports, system documents, responsible departments, responsible positions, generation time, and modification records. The governance status carbon data includes whether there are source conflicts, whether the evidence is complete, whether responsibilities are clearly defined, and whether the rectification loop has been completed. The identified carbon data is classified according to the carbon data object classification system to obtain classified carbon data.

3. The carbon data processing method for the carbon measurement and review scenario in the steel industry according to claim 2, characterized in that, The process of merging the classified carbon data from different sources but pointing to the same process, the same emission source, or the same medium to generate a carbon data governance unit specifically includes: Among the basic attribute carbon data, measurement value carbon data, accounting support carbon data, evidence traceability carbon data, and governance status carbon data, the classified carbon data that point to the same process, the same emission source, or the same medium are merged to obtain several groups of merged carbon data. Each set of merged carbon data constitutes a carbon data governance unit, and each carbon data governance unit includes: process identifier, emission source identifier, medium identifier, time window identifier, data source identifier, value retrieval method identifier, evidence chain identifier, responsible entity identifier, and governance status identifier.

4. The carbon data processing method for carbon metering review scenarios in the steel industry according to claim 1, characterized in that, The preprocessing of the carbon data governance unit based on its process type, value retrieval method, time granularity, source quantity, and inventory changes to obtain a preprocessed carbon data governance unit specifically includes: The preprocessing includes basic preprocessing and condition-triggered special preprocessing. The basic preprocessing steps include: time synchronization, unit unification, format normalization, and source merging. Basic preprocessing is performed on all the carbon data governance units to obtain intermediate carbon data governance units. The condition-triggered special preprocessing specifically includes: In response to determining that the intermediate carbon data governance unit corresponds to a continuous process and that asynchronous sampling exists, process event anchor point alignment is performed to obtain the preprocessed carbon data governance unit. In response to the determination that the intermediate carbon data governance unit involves both continuous process data and batch process data, a dual index mapping is performed to obtain the preprocessed carbon data governance unit. In response to the determination that there are two or more sources for the same intermediate carbon data governance unit, source priority fusion and multi-source three-way cross-alignment processing are performed to obtain the preprocessed carbon data governance unit. In response to determining that the intermediate carbon data governance unit includes inventory change data, wherein the inventory change data includes gas holders, raw material and fuel warehouses, and lime warehouses, inventory constraint completion is performed to obtain a preprocessed carbon data governance unit. In response to determining the start-up or shutdown of the equipment, load fluctuation, or process switching process corresponding to the intermediate carbon data governance unit, the operating condition segmentation and operating condition segment adaptive threshold processing are performed to obtain the preprocessed carbon data governance unit. Specifically, the alignment of process event anchor points includes: Get the set of process event timestamps The process event timestamps include shift switching time, furnace start time, furnace end time, blast furnace tapping time, converter blowing start time, converter blowing end time, electric furnace energization start time, electric furnace tapping time, gas holder switching time, and equipment start / stop time, where n represents the number of process event timestamps. For continuous time series data, the sampling time t is mapped to the process event anchor point with the smallest time distance and within a preset tolerance interval. , ; For batch or furnace type data, the batch or furnace type data is directly assigned to the process event interval with the highest overlap with the start and end time of the batch. For other carbon data in the intermediate carbon data management unit, when the time distance between the sampling time t and any process event anchor point exceeds the preset tolerance range, the other carbon data will be retained in the original time coordinate and marked as an unanchored unit. The dual-index mapping specifically includes: A dual-index structure is constructed, which includes a time index T and a process batch index B. The time index T is established in units of fixed time windows, and the process batch index B is established in units of furnace number, casting number, ladle number or batch number. The intermediate carbon data management unit first determines the time window to which the carbon data belongs based on the time index, and then determines the process batch to which the carbon data belongs based on the process batch index. For continuous data that cannot be directly assigned to a batch, it is mapped to the corresponding batch through process event anchors and production allocation coefficients. The source priority fusion and multi-source three-party cross-alignment processing specifically includes: For the same intermediate carbon data governance unit, the primary source is selected according to the priority chain of direct measurement, production records, laboratory testing, manual ledgers, and financial settlement, while the lower priority sources are retained as verification copies. For the same intermediate carbon data governance unit, the main source data, ledger source data, and settlement or voucher source data are aligned three-way, and the relative deviation index between the sources is calculated as follows: ； in, The index representing the relative deviation between the i-th source value and the j-th source value within the same intermediate carbon data governance unit; This represents the value of the i-th source within the same time window; This represents the value of the j-th source within the same time window; This represents the absolute difference between two source values; Indicates taking , The maximum value in the positive correction term ε is used as the normalized denominator; ε represents the preset positive correction term to prevent the denominator from being zero; The inventory constraint completion specifically includes: When the primary source of the inventory change data is missing within a certain time window, the missing amount is calculated as shown in the following formula: ； in, Indicates the amount of missing data. Indicates beginning inventory. Indicates upstream output. Indicates the amount purchased. Represents ending inventory. This indicates downstream consumption. Indicates export volume; If the missing amount is not within the preset reasonable range, then no completion will be performed, and the inventory change data will be marked as an unrepairable anomaly; otherwise, completion will be performed based on the missing amount. The working condition segmentation and working condition segment adaptive threshold processing specifically include: Based on equipment start-up and shutdown, production load, furnace status, gas holder inventory changes, and steam pipeline switching status, carbon data is divided into stable operating conditions, switching operating conditions, shutdown operating conditions, and abnormal operating conditions. The stable operating conditions refer to the period when equipment operation is continuously stable, key process parameters fluctuate within a preset stable range, and the rate of change in output and energy consumption is below a stable threshold. The switching operating conditions refer to the period during shift changes, furnace changes, raw material changes, gas holder changes, steam changes, or rapid equipment load adjustments. The shutdown operating conditions refer to the period when major equipment is shut down, there is no effective output, and the flow rate of key media remains below the shutdown threshold. The abnormal operating conditions refer to the period when equipment failures, sensor malfunctions, prolonged data interruptions, parameter mutations, or residual abnormalities exceeding thresholds occur. Different integrity thresholds, continuity deviation thresholds, and source deviation thresholds are configured for different operating conditions. Data is then supplemented based on the operating condition in which the carbon data is located and the corresponding integrity threshold, continuity deviation threshold, and source deviation threshold.

5. The carbon data processing method for the carbon metering review scenario in the steel industry according to claim 4, characterized in that, The processing procedure of the governance status diagnosis model is as follows: The overall quality score of the carbon data governance unit is calculated using the following formula: ； in, This represents the overall quality score of the u-th carbon data governance unit; The integrity sub-score is calculated based on the ratio of the expected number of samples to the actual number of samples. The range reasonableness score is calculated by combining the process allowable range, the instrument range range, and the historical stable operation quantile range. The continuous sub-score is calculated based on the deviation of the time interval between adjacent records and the number of breakpoints within the sliding window. The traceability score is calculated based on the completeness of the original record, generation time, responsible party, and modification traces. The sub-score for evidence completeness is calculated based on the binding of certificates, ledgers, inspection reports, and institutional materials. The governance consistency sub-score is calculated based on the degree of consistency in process attribution, emission source attribution, medium attribution, and time attribution. , , , , and To preset weights, ; Calculate the relative deviation index of the carbon data governance unit. ; The calculation formula for the process constraint residuals of the activity level data of the carbon data governance unit is shown below: ； in, This represents the residual vector within the time window t; A represents a column vector consisting of all activity level data within a time window t; A represents the process constraint matrix, and A represents the coefficients in the material balance equation or energy balance equation. Each row corresponds to a balance equation, with the coefficient of input or output terms being +1 and the coefficient of consumption or output terms being -1. Represents a vector of constant terms; The diagnostic criteria include whether the classification is correct, whether the integration is complete, whether the sources are conflicting, whether the evidence is complete, whether the responsible party is clear, whether the time attribution is correct, whether the process attribution is correct, and whether the verification and rectification loop has been completed. Anomaly judgment thresholds are set for the completeness sub-score, scope reasonableness sub-score, continuity sub-score, traceability sub-score, evidence completeness sub-score, governance consistency sub-score, relative deviation index, and process constraint residual. The integrity sub-score, scope reasonableness sub-score, continuity sub-score, traceability sub-score, evidence integrity sub-score, governance consistency sub-score, relative deviation index, and process constraint residual are compared with their corresponding anomaly judgment thresholds to obtain the comparison results. Based on the set diagnostic content and the comparison results, it is determined whether the carbon data governance unit has any abnormalities, and the carbon data governance unit is marked to obtain the marking results. The markings include missing anomalies, range anomalies, jump anomalies, source conflict anomalies, link interruption anomalies, evidence missing anomalies, process logic anomalies, and governance integrity.

6. The carbon data processing method for the carbon metering review scenario in the steel industry according to claim 5, characterized in that, The step of performing metrological support verification and accounting support verification on the preprocessed carbon data governance unit to obtain a first verification result specifically includes: The preprocessed carbon data management unit determines whether the relevant measuring instruments are equipped, whether they are within the validity period of verification or calibration, whether the range and accuracy are suitable, and whether the traceability chain of measurement value is complete. If the relevant measuring instruments are equipped, are within the validity period of verification or calibration, the range and accuracy are suitable, and the traceability chain of measurement value is complete, then the metrological support verification passes; otherwise, the metrological support verification fails. The accounting support verification specifically includes: Based on the actual production boundaries of steel enterprises, coking, sintering or pelletizing, blast furnace, converter or electric furnace, lime kiln, self-owned power plant, gas holder, steam pipeline network and external purchase and sales nodes are abstracted as process control body nodes, and a directed network is formed by the medium or material flow path to obtain the process control body network. Each process control body node is established with a one-to-one mapping relationship with the pre-processed carbon data governance unit. The carbon element flow rate of the control volume is calculated based on the process control volume network, as shown in the following formula: ； in, This represents the carbon flow rate of the carrying medium or material m within the time window t; This indicates the quantity of medium or material m within the time window t; This indicates the carbon content coefficient, carbon mass fraction, or equivalent carbon factor corresponding to the medium or material; For each process control node v, the carbon balance residual is calculated within the time window t, as shown in the following formula: ； in, Indicates the carbon balance residual; This represents the total amount of carbon elements flowing into control node v; This represents the carbon content converted from inventory changes, which is the difference between the carbon content of the inventory at the beginning of the time window and the carbon content of the inventory at the end of the time window. This represents the total amount of carbon elements flowing out of the control volume node v; Indicates the amount of carbon emitted into the atmosphere; Indicates the amount of carbon entering the product, by-product, or solid waste destination; The carbon balance residual rate of control node v within time window t is calculated based on the absolute value of the carbon balance residual and the total amount of carbon elements flowing into control node v, as shown in the following formula: ； in, Indicates the carbon balance residual rate. To prevent the use of a pre-defined positive correction term when the denominator is zero; like If the residual rate exceeds a preset threshold, it is determined that the preprocessed carbon data governance unit corresponding to the process control unit node has a risk of non-compliance in its accounting process, the accounting support verification fails, and the preprocessed carbon data governance unit undergoes alternative source recalculation and deviation contribution rate quantification; otherwise, the accounting support verification passes, wherein the process of alternative source recalculation and deviation contribution rate quantification specifically includes: For the activity level data involved in the accounting, the substitution source values ​​are called for recalculation to obtain the substitution residual of the kth suspicious variable, denoted as: ; The variable correction benefit is calculated as shown in the following formula: ； in, Indicates the variable-adjusted return value. This represents the residual obtained by recalculating the activity level data without invoking alternative source values; Normalizing the returns of each variable yields the deviation contribution rate, as shown in the following formula: ； in, The bias contribution rate of the kth suspicious variable within the time window t. This represents the sum of the variable-corrected returns of all questionable variables within the time window t; In response to the determination that the measurement support verification and accounting support verification of the preprocessed carbon data governance unit have passed, the first verification result is that the preprocessed carbon data governance unit can be used for carbon measurement review; otherwise, the first verification result is that the preprocessed carbon data governance unit cannot be used for carbon measurement review.

7. The carbon data processing method for the carbon measurement and review scenario in the steel industry according to claim 6, characterized in that, The accounting support verification also includes cross-validation of material balance and energy balance, and identification of boundary omissions and emission source misjudgments. The process of cross-validating material and energy balance specifically includes: When the carbon balance residual exceeds a preset carbon balance residual threshold or the carbon balance residual rate exceeds a preset residual rate threshold, the material balance residual and energy balance residual of the corresponding process of the preprocessed carbon data governance unit corresponding to the carbon balance residual are calculated. In response to the determination that the carbon balance residual exceeds a preset carbon balance residual threshold and the material balance residual exceeds a preset material balance residual threshold, it is determined that the activity level value is incorrect or the accounting boundary is omitted. In response to determining that the carbon balance residual exceeds a preset carbon balance residual threshold and the energy balance residual exceeds a preset first energy balance threshold, the input gas quantity, calorific value or component value is determined to be abnormal. In response to determining that the carbon balance residual exceeds a preset carbon balance residual threshold and the energy balance residual is less than a preset second energy balance threshold, the emission factor, carbon content, or carbon destination classification is determined to be incorrect. The process for handling boundary omissions and misidentification of emission sources specifically includes: Establish a process-emission source-medium-destination template to obtain a preset template; In response to the determination that the process control node has a long-term residual in the same direction and its carbon flow pattern matches the preset template, it is determined that there is a risk of boundary omission. In response to determining that a carbon stream is recorded as a product or by-product destination, and that the flow characteristics of the carbon stream conform to the fuel combustion or by-product gas emission pattern, it is determined that there is a risk of misjudgment of the emission source.

8. The carbon data processing method for the carbon metering review scenario in the steel industry according to claim 7, characterized in that, The list of issues includes incorrect carbon data classification, conflicting carbon data sources, missing carbon data evidence, incomplete carbon data governance status, insufficient measurement support, abnormal process logic, misjudgment of emission sources, incorrect activity level values, and mismatched emission factor selection. The rectification path includes supplementing evidence, correcting classification relationships, reconstructing mappings, replacing sources, and re-verifying or recalculating.

9. The carbon data processing method for carbon metering review scenarios in the steel industry according to claim 1, characterized in that, The review and verification process includes classification verification, evidence binding verification, and accounting support verification.

10. A carbon data processing device for carbon metering and auditing scenarios in the steel industry, characterized in that, include: The identification and classification module is configured to acquire raw multi-source data, perform data object identification and classification on the raw multi-source data, and obtain classified carbon data. The carbon data governance unit reconstruction module is configured to merge the classified carbon data from different sources but pointing to the same process, the same emission source or the same medium to generate a carbon data governance unit. The preprocessing module is configured to preprocess the carbon data governance unit according to the process type, value retrieval method, time granularity, source quantity and inventory changes of the carbon data governance unit to obtain the preprocessed carbon data governance unit. The governance status diagnosis module is configured to construct a governance status diagnosis model. The governance status diagnosis model is used to mark carbon data governance units according to the set diagnosis content, obtain the marking results, and input the preprocessed carbon data governance units into the governance status diagnosis model to obtain the first marking result. The support verification module is configured to perform measurement support verification and accounting support verification on the preprocessed carbon data governance unit to obtain a first verification result. The problem list and rectification path generation module is configured to generate a problem list and rectification path based on the first marking result and the first verification result; The rectification module is configured to rectify the preprocessed carbon data governance unit according to the problem list and the rectification path, so as to obtain the rectified carbon data governance unit. The review module is configured to input the governance status diagnosis model into the rectified carbon data governance unit and perform review and verification to obtain a second marking result and a second verification result. If the second marking result indicates that the governance is complete and the second verification result indicates that all are passed, the rectified carbon data governance unit is restored to the reviewable state. Otherwise, a problem list and rectification path are generated based on the second marking result and the second verification result, and the steps of the rectification module are repeated until the rectified carbon data governance unit is restored to the reviewable state.

Images