A carbon evidence detection system and method based on carbon evidence anchor point technology

Abstract

The application discloses a carbon evidence detection system and method based on carbon evidence anchor point technology, relates to the technical field of carbon emission monitoring, and reflects a time differential set Tdif of carbon data generation time characteristics and a modification track set Trac capable of reflecting data modification propagation relations through time sequence analysis and modification tracking of an original data set Cdat, and is built as a fingerprint set Fing, so that the carbon data generation process has unique behavior fingerprint characteristics; the behavior fingerprint set Fing is compared with a benchmark fingerprint set Base for consistency, a fingerprint deviation result Devi is generated, and an anchor point evidence set Anch is formed accordingly, so as to identify whether the carbon data generation behavior deviates from the normal generation mode; finally, the anchor point evidence set Anch is bound with corresponding carbon data records and written into a carbon evidence index structure, thereby effectively solving the problem that the prior art can only record carbon emission values and cannot prove the authenticity of the data generation process.

Description

Technical Field

[0001] This invention relates to the field of carbon emission monitoring technology, specifically to a carbon evidence detection system and method based on carbon evidence anchoring technology. Background Technology

[0002] Against the backdrop of deepening carbon emission regulation, carbon trading management, and green supply chain evaluation, carbon emission data is gradually becoming an important information carrier reflecting the environmental impact of corporate production activities. With the widespread application of digital management systems in energy management, production scheduling, and emission statistics, various production systems continuously generate a large amount of carbon emission-related data records, which are then aggregated, calculated, and reported through information platforms. In this application environment, carbon emission data not only serves the function of emission statistics but is also increasingly used in scenarios such as carbon asset accounting, carbon trading compliance, and regulatory auditing. Therefore, it is necessary to use more reliable technical means to form a traceable and verifiable evidence structure for carbon emission data.

[0003] In current carbon emission data management practices, most systems primarily retain data by recording carbon emission values, timestamps, and source identifiers. For example, in carbon emission accounting systems, energy management systems, or carbon trading platforms, the system typically only stores the carbon emission calculation results or records them via data chains. While this approach preserves the carbon emission data results, the system focuses mainly on the carbon emission values ​​themselves, without recording the behavioral characteristics of the carbon emission data during its generation. For instance, during carbon emission data calculation, the system typically does not record the data calculation path, the temporal differential structure of the data generation, or the behavioral trajectory of the data during generation or updating. Therefore, once the carbon emission data is modified or regenerated in subsequent stages, the system often struggles to identify whether the data has been manipulated or recalculated based solely on the data results, making it difficult to form an evidentiary structure that reflects the true data generation process. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a carbon evidence detection system and method based on carbon evidence anchoring technology, which solves the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a carbon evidence detection method based on carbon evidence anchoring technology, comprising the following steps:

[0006] S1. Collect the raw record information generated during the carbon data generation process, and organize the collected information to form the raw data set Cdat. At the same time, construct the generation path set Path based on the data generation and calling relationship.

[0007] S2. Perform time-series parsing and change tracking on the original dataset Cdat to extract time behavior information and modification behavior information during the carbon data generation process, forming the time differential set Tdif and the modification trajectory set Trac, respectively.

[0008] S3. Based on the generated path set Path, the time differential set Tdif, and the modified trajectory set Trac, perform behavior mapping processing to generate the behavior feature set Behv, and construct the behavior fingerprint set Fing based on the behavior feature set Behv.

[0009] S4. Perform a consistency comparison between the behavioral fingerprint set Fing and the pre-established benchmark fingerprint set Base to obtain the fingerprint deviation result Devi, and generate the anchor evidence set Anch based on the fingerprint deviation result Devi.

[0010] S5. Bind the anchor evidence set Anch with the corresponding carbon data record and write it into the carbon evidence index structure, and output the carbon evidence result Cert.

[0011] Preferably, S1 includes S11;

[0012] S11. Extract raw record information related to the carbon data generation process from carbon emission accounting systems, energy management systems, and log recording systems, and convert raw record information from different sources into a unified standard record structure.

[0013] Each original record unit includes a task identifier, processing node, record type, record time, call source, and call target;

[0014] Then, using the task identifier as the basis for aggregation, multiple original record units belonging to the same carbon data generation task are merged and organized, and each original record unit is arranged in order according to the recording time, thereby forming the original data set Cdat used to describe the basic record structure of the carbon data generation process.

[0015] Preferably, S1 further includes S12;

[0016] S12. Filter multiple raw record units with the same task identifier in the raw data set Cdat to limit the processing scope of the same carbon data generation task; then read the calling source and calling target of each raw record unit to determine the calling connection relationship between different processing nodes.

[0017] Based on this, the order of the call connection relationship is determined by combining the recording time corresponding to each original recording unit, and multiple processing nodes participating in the call connection relationship are connected in chronological order to form a generation path corresponding to the same task identifier, thus forming the call correspondence relationship and the chronological order relationship.

[0018] Preferably, S2 includes S21;

[0019] S21. Read multiple original record units with the same task identifier from the original data set Cdat to limit the time resolution range of the same carbon data generation task; then, extract the processing nodes and recording times from the multiple original record units, and arrange the multiple original record units in order according to the recording times to obtain the time sequence arrangement result of each processing node under the same task identifier.

[0020] After obtaining the timing arrangement result, the time difference between adjacent original recording units is calculated sequentially to obtain the time interval; at the same time, the node connection relationship corresponding to each time interval is determined by combining the processing nodes corresponding to the adjacent original recording units to obtain the node timing pair.

[0021] Then, multiple time intervals under the same task identifier are collected and organized with the corresponding multiple node time series pairs, so that each time interval has a unique correspondence with the corresponding node time series pair, thereby forming a time differential set Tdif used to describe the time behavior information of the carbon data generation process.

[0022] Preferably, S2 further includes S22;

[0023] S22. By reading multiple original record units with the same task identifier from the original dataset Cdat, and filtering out the original record units whose record type indicates data update, data supplementation, data overwrite, or data write-back, as the record units to be tracked; then, reading the processing node, recording time, call source, and call target corresponding to each record unit to be tracked, and arranging the multiple record units to be tracked in order according to the recording time, the change sequence result under the same task identifier is obtained;

[0024] After obtaining the change timing result, based on the call source and call target of each record unit to be tracked, the transmission direction of the change behavior between different processing nodes is determined to obtain the change direction; at the same time, combined with the position of the record unit to be tracked in the change timing result, the sequential connection relationship between consecutive change behaviors is extracted to obtain the change sequence chain.

[0025] Then, multiple change directions and multiple change sequence chains under the same task identifier are collected and organized to form a continuous correspondence between the occurrence location, transmission direction and sequence of each change behavior, thereby forming a set of modification trajectories Trac used to describe the modification behavior information in the carbon data generation process.

[0026] Preferably, S3 includes S31;

[0027] S31. Read each generated path corresponding to the same task identifier from the generated path set Path, and extract the connection order between adjacent processing nodes in each generated path to obtain a path segment; then, read multiple time intervals and corresponding node time sequence pairs corresponding to the task identifier from the time differential set Tdif, and extract the combination relationship between each time interval and the corresponding node time sequence pair to obtain a time segment; then, read each modification behavior information corresponding to the task identifier from the modified trajectory set Trac, and extract the combination relationship between each change direction and the corresponding change sequence chain to obtain a trajectory segment;

[0028] After obtaining multiple path segments, multiple time segments, and multiple trajectory segments, each path segment is aligned with the corresponding time segment in time sequence to determine the time behavior corresponding to the connection relationship of the same processing node; then the aligned result is associated with the trajectory segment corresponding to the process position to establish a unified mapping relationship between the path behavior, time behavior, and modification behavior under the same carbon data generation task.

[0029] The established unified mapping relationship is then organized into multiple behavioral units, each of which represents a composite behavioral relationship in the carbon data generation process. Finally, the multiple behavioral units obtained under the same task identifier are aggregated and organized to form a behavioral feature set Behv that describes the behavioral structure of carbon data generation.

[0030] Preferably, S3 further includes S32;

[0031] S32. Read multiple behavioral units corresponding to the same task identifier from the behavioral feature set Behv, and arrange them in order according to their positions in the carbon data generation process to obtain a feature sequence chain; then, perform correlation analysis on adjacent behavioral units in the feature sequence chain to extract the connection and continuation relationships between adjacent behavioral units and form a correlation fragment.

[0032] After obtaining multiple associated fragments, they are continuously spliced ​​according to the order of their arrangement in the feature sequence chain, while keeping the process position corresponding to each associated fragment unchanged, so that multiple behavioral units under the same task identifier form a complete feature chain structure; then, the feature chain structure is uniformly structure-encoded to generate a fingerprint description result that uniquely represents the behavioral structure of the carbon data generation task, and the fingerprint description result is the fingerprint fragment.

[0033] Finally, one or more fingerprint fragments obtained under the same task identifier are collected and organized to form a behavioral fingerprint set Fing used to describe the overall characteristics of carbon data generation behavior.

[0034] Preferably, S4 includes S41;

[0035] S41. Read one or more fingerprint fragments corresponding to the same task identifier from the behavior fingerprint set Fing, and read one or more reference fingerprint records corresponding to the task identifier from the pre-established reference fingerprint set Base, so that the fingerprint fragments and the reference fingerprint records form a comparable fingerprint pair;

[0036] Subsequently, a consistency comparison process is performed on each set of fingerprint pairs. Specifically, by comparing the structural arrangement and segment order relationship between the fingerprint fragment and the reference fingerprint record segment by segment, the matching and non-matching parts between the two are identified. During the comparison process, when the segment order, number of segments, or segment connection relationship in the fingerprint fragment is inconsistent with the reference fingerprint record, the corresponding inconsistency information is recorded as deviation information, and a fingerprint deviation result Devi is formed based on the occurrence position and number of each deviation information.

[0037] After obtaining the fingerprint deviation result Devi, the fingerprint portions with stable structures and the fingerprint portions that have deviated during the carbon data generation process are extracted based on the fingerprint deviation result Devi. The information of stable structures and deviation information are then uniformly organized so that each piece of deviation information corresponds to a specific fingerprint segment position and a corresponding task identifier. Subsequently, the organized stable structure information and deviation information are grouped according to the same task identifier to generate the anchor evidence set Anch, which describes the evidence structure of carbon data generation behavior.

[0038] Preferably, S5 includes S51;

[0039] S51. Read the evidence structure information corresponding to the same task identifier from the anchor evidence set Anch, and locate the corresponding carbon data record based on the task identifier and fingerprint fragment position recorded in each piece of evidence structure information, so that each piece of evidence structure information is associated with the corresponding carbon data record.

[0040] Subsequently, each piece of evidence structure information is bound to the corresponding carbon data record, so that the corresponding carbon data record has associated behavioral evidence content; wherein, the binding process includes writing the task identifier, fingerprint fragment position and deviation information into the associated field of the corresponding carbon data record, or establishing an index mapping relationship between the corresponding carbon data record and the evidence structure information;

[0041] After the binding process is completed, the bound evidence structure information is collected according to the task identifier and written into the carbon evidence index structure, so that the carbon data records under the same task identifier and the corresponding evidence structure information form a searchable evidence index relationship. Then, based on the index result written into the carbon evidence index structure, an output result describing the carbon data record and its corresponding behavioral evidence relationship is generated. The output result is defined as the carbon evidence result Cert.

[0042] A carbon evidence detection system based on carbon evidence anchoring technology includes a carbon data acquisition module, a data parsing and extraction module, a feature mapping module, a consistency comparison module, and a carbon data decision module.

[0043] The carbon data acquisition module collects the raw record information generated during the carbon data generation process, and associates and organizes the collected information to form a raw data set Cdat. At the same time, it constructs a generation path set Path based on the data generation and calling relationship.

[0044] The data parsing and extraction module performs time-series parsing and change tracking on the original data set Cdat, extracting time behavior information and modification behavior information during the carbon data generation process, forming the time differential set Tdif and the modification trajectory set Trac, respectively.

[0045] The feature mapping module performs behavior mapping processing based on the generated path set Path, the time differential set Tdif, and the modified trajectory set Trac to generate a behavior feature set Behv, and constructs a behavior fingerprint set Fing based on the behavior feature set Behv.

[0046] The consistency comparison module performs a consistency comparison between the behavioral fingerprint set Fing and the pre-established benchmark fingerprint set Base to obtain the fingerprint deviation result Devi, and generates the anchor evidence set Anch based on the fingerprint deviation result Devi.

[0047] The carbon data decision module binds the anchor evidence set Anch with the corresponding carbon data record and writes it into the carbon evidence index structure, outputting the carbon evidence result Cert.

[0048] This invention provides a carbon evidence detection system and method based on carbon evidence anchoring technology, which has the following beneficial effects:

[0049] (1) By performing time-series analysis and change tracking on the original dataset Cdat, a time differential set Tdif reflecting the time characteristics of carbon data generation and a modification trajectory set Trac reflecting the data modification propagation relationship are formed. Based on this, the generation path set Path, the time differential set Tdif, and the modification trajectory set Trac are uniformly mapped to behavior, and a behavior fingerprint set Fing is further constructed to give the carbon data generation process a unique behavior fingerprint feature. Then, the behavior fingerprint set Fing is compared with the baseline fingerprint set Base to generate a fingerprint deviation result Devi, and an anchor evidence set Anch is formed accordingly to identify whether the carbon data generation behavior deviates from the normal generation mode. Finally, by binding the anchor evidence set Anch with the corresponding carbon data record and writing it into the carbon evidence index structure, a carbon evidence result Cert that can be directly used for verification is output. This effectively solves the problem that existing technologies only record carbon emission values ​​but cannot prove the authenticity of the data generation process.

[0050] (2) By extracting the connection order between adjacent processing nodes from the generated path set Path to form path segments, and combining the time interval and node time sequence pairs in the time differential set Tdif to form time segments, and then forming trajectory segments by combining the change direction and change sequence chain in the modified trajectory set Trac, on this basis, by constructing feature sequence chains for multiple behavioral units in the behavioral feature set Behv according to the process order and identifying the continuity relationship between adjacent behaviors to form association segments, and then generating fingerprint segments through structural encoding, finally forming the behavioral fingerprint set Fing, so that each carbon data generation task has a unique behavioral structure identifier. This improves the behavioral uniqueness identification capability of the carbon data generation process and provides a stable and reliable behavioral feature foundation for subsequent consistency comparison and evidence anchoring.

[0051] (3) By comparing the fingerprint fragments in the behavioral fingerprint set Fing with the benchmark fingerprint records in the benchmark fingerprint set Base, it is possible to identify whether there are differences in the number of fragments, the order of fragments, or the connection relationship of fragments in the carbon data generation process, and to form the fingerprint deviation result Devi based on the identified deviation information. Subsequently, based on the fingerprint deviation result Devi, the stable behavioral structure and the deviation behavioral structure in the carbon data generation process are extracted, and the anchor evidence set Anch containing task identifiers, fingerprint fragment positions, and deviation information is generated. On this basis, the evidence structure information in the anchor evidence set Anch is bound to the specific carbon data record through step S51 and written into the carbon evidence index structure, and finally the carbon evidence result Cert is generated, so that each carbon data record can be associated with the behavioral evidence in its generation process. The carbon data result and its generation behavioral structure can be integrated into an integrated evidence record, so that the carbon data has directly verifiable evidence in the carbon verification, carbon audit, and carbon trading compliance process. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the steps of a carbon evidence detection method based on carbon evidence anchoring technology according to the present invention.

[0053] Figure 2 This is a schematic diagram of a carbon evidence detection system based on carbon evidence anchoring technology according to the present invention.

[0054] Figure 3 This is a schematic diagram of the method flow for step S1 of the present invention. Detailed Implementation

[0055] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. Example

[0056] This invention provides a carbon evidence detection method based on carbon evidence anchoring technology. Please refer to [link / reference]. Figure 1 This includes the following steps:

[0057] S1. Collect the raw record information generated during the carbon data generation process, and organize the collected information to form the raw data set Cdat. At the same time, construct the generation path set Path based on the data generation and calling relationship.

[0058] S2. Perform time-series parsing and change tracking on the original dataset Cdat to extract time behavior information and modification behavior information during the carbon data generation process, forming the time differential set Tdif and the modification trajectory set Trac, respectively.

[0059] S3. Based on the generated path set Path, the time differential set Tdif, and the modified trajectory set Trac, perform behavior mapping processing to generate the behavior feature set Behv, and construct the behavior fingerprint set Fing based on the behavior feature set Behv.

[0060] S4. Perform a consistency comparison between the behavioral fingerprint set Fing and the pre-established benchmark fingerprint set Base to obtain the fingerprint deviation result Devi, and generate the anchor evidence set Anch based on the fingerprint deviation result Devi.

[0061] S5. Bind the anchor evidence set Anch with the corresponding carbon data record and write it into the carbon evidence index structure, and output the carbon evidence result Cert.

[0062] In this embodiment, through the coordinated execution of steps S1 to S5, this solution, based on traditional carbon data recording methods, transforms behavioral information during the carbon data generation process into verifiable evidence structures, achieving full-process anchoring of carbon data generation behavior. Specifically, firstly, step S1 collects and organizes the raw recording information generated during the carbon data generation process, forming a raw data set Cdat and constructing a generation path set Path, thereby completely recording the generation path of carbon data between processing nodes; then, in step S2, the raw data set Cdat undergoes time-series analysis and change tracking, forming a time differential set Tdif that reflects the time characteristics of carbon data generation and a modification trajectory set Trac that reflects the data modification propagation relationship; based on this, step S3 further integrates the generation path set Path, the time differential set Tdif, and the modification trajectory set Trac... A unified behavior mapping is used to obtain the behavior feature set Behv, which describes the structure of carbon data generation behavior, and further constructs a behavior fingerprint set Fing, giving the carbon data generation process unique behavior fingerprint features. Then, in step S4, the behavior fingerprint set Fing is compared with the baseline fingerprint set Base to generate a fingerprint deviation result Devi, which forms an anchor evidence set Anch to identify whether the carbon data generation behavior deviates from the normal generation pattern. Finally, in step S5, the anchor evidence set Anch is bound to the corresponding carbon data record and written into the carbon evidence index structure, outputting a carbon evidence result Cert that can be directly used for verification. Through this method, this solution can not only record carbon data results, but also the path structure, temporal behavior, and modification propagation relationships during the carbon data generation process, thus effectively solving the problem that existing technologies only record carbon emission values ​​but cannot prove the authenticity of the data generation process. For example, in a corporate carbon emission accounting system, when a carbon emission record is manually modified after submission, traditional systems can usually only see the modified value and cannot determine whether the modification actually occurred. However, with the carbon evidence result Cert generated by this solution, regulators can identify whether the behavioral structure of the record has deviated abnormally between processing nodes based on the corresponding behavioral fingerprint set Fing and fingerprint deviation result Devi. This allows for timely detection of carbon data being supplemented or tampered with, thereby improving the data credibility and regulatory efficiency in carbon data verification, carbon trading compliance, and carbon auditing processes. Example

[0063] Please see Figure 3 Specifically: S1 includes S11;

[0064] S11. Extract raw record information related to the carbon data generation process from carbon emission accounting systems, energy management systems, and log recording systems, and convert raw record information from different sources into a unified standard record structure.

[0065] Each original record unit includes a task identifier, processing node, record type, record time, call source, and call target;

[0066] Then, using the task identifier as the basis for aggregation, multiple original record units belonging to the same carbon data generation task are merged and organized, and each original record unit is arranged in order according to the recording time, thereby forming the original data set Cdat used to describe the basic record structure of the carbon data generation process.

[0067] S1 further includes S12;

[0068] S12. Filter multiple raw record units with the same task identifier in the raw data set Cdat to limit the processing scope of the same carbon data generation task; then read the calling source and calling target of each raw record unit to determine the calling connection relationship between different processing nodes.

[0069] Based on this, the order of calling connections is determined by combining the recording time corresponding to each original recording unit, and multiple processing nodes involved in the calling connection are connected in chronological order to form a generation path corresponding to the same task identifier, thus forming a calling correspondence and a chronological sequence.

[0070] It should be noted that:

[0071] To enable implementers to construct the path set Path based on the above information, the following implementation instructions are provided for this step;

[0072] First, all raw record units are read from the raw dataset Cdat, and the raw record units are grouped according to the task identifier, so that multiple raw record units with the same task identifier form a task record group, and each task record group corresponds to a complete carbon data generation task.

[0073] Subsequently, the processing node, recording time, call source and call target of each original recording unit are extracted from each task record group, and all original recording units are sorted by time according to the recording time to obtain the time order of each processing behavior in the same carbon data generation task.

[0074] After completing the time sorting, the call connection relationship between processing nodes is determined based on the call source and call target of each original record unit: when the call source of a certain original record unit is consistent with the processing node of another original record unit, it is determined that there is a call connection relationship between the original record units; at the same time, the execution order of the call connection relationship is determined by combining the order of the recording time.

[0075] After determining the call connection relationships and their time sequence among the processing nodes, the multiple processing nodes involved in the call relationship are connected sequentially according to the call connection order, thereby forming a data processing path that describes the carbon data generation process.

[0076] In this embodiment, through the coordinated execution of steps S11 and S12, this solution can perform structured recording and path reconstruction of the data generation process in the initial stage of carbon data generation. This makes the carbon data generation process no longer just a single data result, but a complete data processing flow with a clear processing order and calling relationship. Specifically, by uniformly structuring and organizing the multi-source record information from the carbon emission accounting system, energy management system, and log recording system, a standard record structure containing task identifier, processing node, record type, record time, calling source, and calling target is formed. This structure is further aggregated into an original data set Cdat, which can unify and integrate the discrete records generated by the same carbon data generation task in different systems into a traceable data record set. On this basis, by grouping each original record unit in the original data set Cdat according to the task identifier at the task level, and constructing a generation path set Path based on the correspondence between the record time and the calling source and calling target, the calling connection relationship between each processing node in the carbon data generation process can be linked together according to the actual execution order, thereby forming a complete data processing path structure. In this way, the system can not only identify the final generated carbon data, but also clearly define the flow and order of carbon data across processing nodes. For example, in actual enterprise carbon emission accounting scenarios, a carbon emission record often needs to go through multiple processing steps, such as energy data collection, emission factor calculation, emission aggregation, and result writing. The generation path set Path formed by this solution can clearly record the flow order of the carbon emission record across each processing step. When a processing node is abnormally inserted or skipped, regulators can quickly identify whether an anomaly has occurred in the data processing flow by comparing the generation path set Path, thereby improving the traceability and integrity of the carbon data generation process and providing a reliable data foundation for subsequent carbon data behavior identification and evidence anchoring. Example

[0077] Specifically: S2 includes S21;

[0078] S21. Read multiple original record units with the same task identifier from the original data set Cdat to limit the time resolution range of the same carbon data generation task; then, extract the processing nodes and recording times from the multiple original record units, and arrange the multiple original record units in order according to the recording times to obtain the time sequence arrangement result of each processing node under the same task identifier.

[0079] After obtaining the timing arrangement result, the time difference between adjacent original recording units is calculated sequentially to obtain the time interval; at the same time, the node connection relationship corresponding to each time interval is determined by combining the processing nodes corresponding to the adjacent original recording units to obtain the node timing pair.

[0080] Then, the multiple time intervals under the same task identifier are collected and organized with the corresponding multiple node time series pairs, so that each time interval is uniquely associated with the corresponding node time series pair, thereby forming a time differential set Tdif used to describe the time behavior information of the carbon data generation process.

[0081] A node time sequence pair represents the sequential combination of two processing nodes corresponding to the same time interval, used to identify which two processing nodes the time interval occurs between; the purpose of setting node time sequence pairs is to provide node correspondence for the specific processing position to which the time behavior information belongs in subsequent analysis;

[0082] It should be noted that:

[0083] To enable implementers to generate time intervals and node timing pairs based on the above information, the processing method is further explained as follows;

[0084] First, select multiple original record units with the same task identifier from the original dataset Cdat, and sort the multiple original record units in ascending order according to the recording time of each original record unit, so that the original record units under the same task identifier form a record sequence arranged in chronological order.

[0085] Subsequently, two adjacent raw recording units are read sequentially from the recording sequence, and the recording time corresponding to each of these two raw recording units is extracted. By calculating the difference between the recording time of the latter raw recording unit and the recording time of the former raw recording unit, a time interval is obtained; the time interval represents the length of time consumed between two adjacent processing actions in the carbon data generation process corresponding to the same task identifier.

[0086] While obtaining the time interval, the processing nodes corresponding to the two original recording units are read, and the processing nodes of the previous original recording unit and the processing nodes of the next original recording unit are combined in chronological order to form a node timing pair; wherein, the node timing pair is used to represent the sequential connection relationship between two adjacent processing nodes in the carbon data generation process.

[0087] Then, each time interval is bound to a corresponding node time pair, so that each time interval corresponds to a specific node time pair, thereby forming a data recording unit for describing the time variation characteristics of the carbon data generation process; finally, multiple data recording units under the same task identifier are collected and organized to obtain a time differential set Tdif for describing the time behavior information of the carbon data generation process.

[0088] S2 further includes S22;

[0089] S22. By reading multiple original record units with the same task identifier from the original dataset Cdat, and filtering out the original record units whose record type indicates data update, data supplementation, data overwrite, or data write-back, as the record units to be tracked; then, reading the processing node, recording time, call source, and call target corresponding to each record unit to be tracked, and arranging the multiple record units to be tracked in order according to the recording time, the change sequence result under the same task identifier is obtained;

[0090] After obtaining the change timing result, based on the call source and call target of each record unit to be tracked, the transmission direction of the change behavior between different processing nodes is determined to obtain the change direction; at the same time, combined with the position of the record unit to be tracked in the change timing result, the sequential connection relationship between consecutive change behaviors is extracted to obtain the change sequence chain.

[0091] Then, multiple change directions and multiple change sequence chains under the same task identifier are collected and organized to form a continuous correspondence between the occurrence location, transmission direction and sequence of each change behavior, thereby forming a set of modification trajectories Trac used to describe the modification behavior information in the carbon data generation process.

[0092] It should be noted that:

[0093] To enable implementers to achieve the corresponding functions based on the above change sequence chain, the processing procedure is further explained as follows;

[0094] First, select the record units to be tracked from the original dataset Cdat that have the same task identifier, and sort them in ascending order according to the recording time of each record unit to be tracked, so that the record units to be tracked under the same task identifier form a change record sequence arranged in chronological order, thereby obtaining the change time sequence result;

[0095] Subsequently, for each record unit to be tracked in the change record sequence, its corresponding call source and call target are read. When the processing node corresponding to the call source is consistent with the processing node of another record unit to be tracked, it is determined that the change behavior of the record unit to be tracked was initiated by the processing node and points to the processing node corresponding to the call target. By taking the processing node corresponding to the call source as the starting node and the processing node corresponding to the call target as the target node, the transmission direction of the change behavior in the carbon data generation process can be obtained, and the transmission direction is the change direction.

[0096] After obtaining multiple change directions, the system continues to determine whether there is a continuous change relationship between adjacent record units based on the sequential positional relationship of each record unit to be tracked in the change record sequence. When the calling source of the subsequent record unit to be tracked corresponds to the calling target of the previous record unit to be tracked, it is determined that there is a continuous change connection relationship between the two, and this continuous relationship is recorded as a sequential connection unit;

[0097] Subsequently, all sequential connection units with sequential connection relationships under the same task identifier are connected in series according to the recording time to form a sequential structure describing the continuous propagation process of multiple change behaviors. The sequential structure is the change sequence chain.

[0098] Finally, the multiple change directions obtained under the same task identifier are collected and organized with their corresponding change sequence chains, so that the occurrence location, propagation direction and time sequence of each change behavior form a continuous correspondence, thereby forming a set of modification trajectories Trac used to describe the propagation path of modification behavior in the carbon data generation process.

[0099] In this embodiment, by jointly executing steps S21 and S22, this solution can simultaneously establish a temporal behavior structure and a modification propagation structure during the carbon data generation process. This allows the carbon data generation process to not only be reconstructed by the path but also to be finely characterized by temporal features and modification behaviors. Specifically, by sorting the original record units with the same task identifier in the original dataset Cdat according to their recording time, and calculating the time interval between adjacent original record units and the corresponding node time sequence pairs, a temporal differential set Tdif is formed. This quantifies the time span consumed when carbon data flows between different processing nodes, thereby reflecting the temporal rhythm changes in the carbon data generation process. Simultaneously, by selecting original record units from the original dataset Cdat whose record types are data update, data supplementation, data overwrite, or data write-back, and determining the change direction based on the call source and call target, and then constructing a continuous change sequence chain in conjunction with the recording time, a modification trajectory set Trac is finally formed. This can completely describe the propagation path of modification behaviors of carbon data between different processing nodes. By simultaneously obtaining the temporal differential set Tdif and the modification trajectory set Trac, the system can identify temporal anomalies and modification propagation anomalies in the carbon data generation process. For example, in corporate carbon emission accounting scenarios, a carbon emission record is usually calculated and written to the database very quickly after energy data collection. When a carbon data record is generated but data is added or overwritten a long time later, analyzing the corresponding time differential set Tdif can reveal an abnormal delay between adjacent processing nodes. At the same time, by modifying the change direction and change sequence chain in the trajectory set Trac, it can be determined which processing node initiated the addition and which node it propagated to. This helps regulators quickly identify whether carbon data has been added or modified after the fact, thereby improving the monitorability and anomaly identification capabilities of the carbon data generation process. Example

[0100] Specifically: S3 includes S31;

[0101] S31. Read each generated path corresponding to the same task identifier from the generated path set Path, and extract the connection order between adjacent processing nodes in each generated path to obtain a path segment; then, read multiple time intervals and corresponding node time sequence pairs corresponding to the task identifier from the time differential set Tdif, and extract the combination relationship between each time interval and the corresponding node time sequence pair to obtain a time segment; then, read each modification behavior information corresponding to the task identifier from the modified trajectory set Trac, and extract the combination relationship between each change direction and the corresponding change sequence chain to obtain a trajectory segment;

[0102] After obtaining multiple path segments, multiple time segments, and multiple trajectory segments, each path segment is aligned with the corresponding time segment in time sequence to determine the time behavior corresponding to the connection relationship of the same processing node; then the aligned result is associated with the trajectory segment corresponding to the process position to establish a unified mapping relationship between the path behavior, time behavior, and modification behavior under the same carbon data generation task.

[0103] The established unified mapping relationship is then organized into multiple behavioral units, each of which represents a composite behavioral relationship in the carbon data generation process. Finally, the multiple behavioral units obtained under the same task identifier are aggregated and organized to form a set of behavioral features Behv that describes the behavioral structure of carbon data generation.

[0104] Here, path segments represent the ordered connection relationships between adjacent processing nodes extracted from the generated path set Path, which are used to reflect the node flow structure in the carbon data generation process; the purpose of setting path segments is to provide a path basis for behavior mapping processing.

[0105] A time segment represents the corresponding combination of time intervals extracted from the time differential set Tdif and node time series pairs, used to reflect the temporal characteristics between adjacent processing behaviors during carbon data generation; the purpose of setting time segments is to provide a temporal basis for behavior mapping processing.

[0106] Trajectory segments represent the corresponding combinations of change directions and change order chains extracted from the modified trajectory set Trac, used to reflect the modification propagation characteristics during carbon data generation; the purpose of setting trajectory segments is to provide a modification basis for behavior mapping processing;

[0107] A behavior unit represents a composite behavior record formed by mapping path segments, time segments, and trajectory segments. It is used to uniformly describe the path behavior, time behavior, and modification behavior that occur at the same process position during carbon data generation. The purpose of setting behavior units is to provide a standardized feature carrier for the subsequent construction of the behavior fingerprint set Fing.

[0108] S3 further includes S32;

[0109] S32. Read multiple behavioral units corresponding to the same task identifier from the behavioral feature set Behv, and arrange them in order according to their positions in the carbon data generation process to obtain a feature sequence chain; then, perform correlation analysis on adjacent behavioral units in the feature sequence chain to extract the connection and continuation relationships between adjacent behavioral units and form a correlation fragment.

[0110] After obtaining multiple associated fragments, they are continuously spliced ​​according to the order of their arrangement in the feature sequence chain, while keeping the process position corresponding to each associated fragment unchanged, so that multiple behavioral units under the same task identifier form a complete feature chain structure; then, the feature chain structure is uniformly structure-encoded to generate a fingerprint description result that uniquely represents the behavioral structure of the carbon data generation task, and the fingerprint description result is the fingerprint fragment.

[0111] Finally, one or more fingerprint fragments obtained under the same task identifier are collected and organized to form a behavioral fingerprint set Fing used to describe the overall characteristics of carbon data generation behavior;

[0112] Among them, the feature sequence chain represents the ordered arrangement structure of multiple behavioral units under the same task identifier according to their sequential positions in the process, which is used to reflect the overall temporal organization of carbon data generation behavior; the purpose of setting the feature sequence chain is to provide a sequential basis for the subsequent extraction of continuous behavioral relationships.

[0113] The associated fragment represents the connection relationship record between adjacent behavioral units extracted from the feature sequence chain, which is used to reflect the succession relationship between adjacent composite behaviors; the purpose of setting the associated fragment is to provide an association basis for the subsequent formation of the overall behavioral fingerprint;

[0114] A fingerprint fragment represents the feature description result obtained after structural encoding of multiple consecutively arranged related fragments. It is used to characterize the local behavioral fingerprint of the carbon data generation process under the same task identifier. The purpose of setting fingerprint fragments is to provide an encoding basis for subsequent aggregation to form the behavioral fingerprint set Fing.

[0115] It should be noted that:

[0116] To enable implementers to carry out the associated segment processing process based on the above content, the implementation method is further explained as follows.

[0117] First, multiple behavioral units corresponding to the same task identifier are read from the Behv behavioral feature set, and these behavioral units are arranged into a feature sequence chain according to the order of the carbon data generation process. In the feature sequence chain, each behavioral unit contains a corresponding path segment, time segment, and trajectory segment.

[0118] Subsequently, two adjacent behavioral units are read sequentially in the feature sequence chain, and the path segment, time segment, and trajectory segment from these two behavioral units are extracted respectively. Implementers can determine whether there is a behavioral continuity relationship between adjacent behavioral units in the following manner:

[0119] First, the path segments in the two behavioral units are compared. When the starting processing node in the path segment of the latter behavioral unit matches the ending processing node in the path segment of the former behavioral unit, it is determined that there is a node connection relationship between the two in the carbon data generation process.

[0120] Secondly, the time segments in the two behavioral units are compared. When the starting processing node in the node timing pair corresponding to the time segment of the later behavioral unit is consistent with the ending processing node in the node timing pair of the earlier behavioral unit, and the time interval corresponding to the later behavioral unit is immediately following the time interval corresponding to the earlier behavioral unit, it is determined that there is a continuity relationship between the two in terms of time behavior.

[0121] Next, the trajectory segments in the two behavior units are compared. When the source of the direction change call in the trajectory segment of the later behavior unit is consistent with the target of the direction change call in the trajectory segment of the earlier behavior unit, it is determined that there is a continuity relationship between the two in the propagation path of the modified behavior.

[0122] When at least one of the above three relationships exists, the connection relationship between the two behavioral units is recorded, and a behavioral connection record is generated. The behavioral connection record is defined as an associated fragment.

[0123] Finally, multiple associated fragments identified under the same task identifier are grouped and organized according to the order in the feature sequence chain, so that multiple associated fragments form an associated structure that can reflect the continuous relationship of carbon data generation behavior, thereby providing a basic structure for the subsequent generation of fingerprint fragments and behavioral fingerprint set Fing.

[0124] In this embodiment, through the coordinated processing of steps S31 and S32, this solution can uniformly map and structurally encode the path structure, temporal behavior, and modification propagation relationships in the carbon data generation process. This transforms the behavioral information, originally scattered across different systems, into identifiable behavioral structural features, and further forms a unique behavioral fingerprint structure. Specifically, by extracting the connection order between adjacent processing nodes from the generation path set Path to form path segments, and combining the time interval and node time sequence pairs in the time differential set Tdif to form time segments, and then forming trajectory segments by combining the change direction and change sequence chain in the modification trajectory set Trac, the system can simultaneously capture the path behavior, temporal behavior, and modification behavior of carbon data generation at the same process location, thereby constructing standardized behavioral units and aggregating them into a behavioral feature set Behv. On this basis, by constructing feature sequence chains for multiple behavioral units in the behavioral feature set Behv according to the process order and identifying the continuity relationship between adjacent behaviors to form associated segments, and then generating fingerprint segments through structural encoding, a behavioral fingerprint set Fing is finally formed, so that each carbon data generation task has a unique behavioral structural identifier. For example, in real-world enterprise carbon emission accounting scenarios, a carbon emission record typically passes through an energy data acquisition and processing node, an emission calculation and processing node, and a result writing and processing node in sequence. When an additional data correction step is manually inserted during the generation of a carbon data record, although the final carbon emission value may remain consistent, this additional step will alter the path segments, corresponding time segments, and the combined structure of the trajectory segments formed between adjacent processing nodes, thereby causing a change in the generated behavioral fingerprint set (Fing). In this way, the system can identify whether the carbon data generation behavior has been manually adjusted while the carbon data result remains unchanged, thus improving the ability to identify the uniqueness of the behavior in the carbon data generation process and providing a stable and reliable behavioral feature foundation for subsequent consistency comparison and evidence anchoring. Example

[0125] Specifically: S4 includes S41;

[0126] S41. Read one or more fingerprint fragments corresponding to the same task identifier from the behavior fingerprint set Fing, and read one or more reference fingerprint records corresponding to the task identifier from the pre-established reference fingerprint set Base, so that the fingerprint fragments and the reference fingerprint records form a comparable fingerprint pair;

[0127] Subsequently, a consistency comparison process is performed on each set of fingerprint pairs. Specifically, by comparing the structural arrangement and segment order relationship between the fingerprint fragment and the reference fingerprint record segment by segment, the matching and non-matching parts between the two are identified. During the comparison process, when the segment order, number of segments, or segment connection relationship in the fingerprint fragment is inconsistent with the reference fingerprint record, the corresponding inconsistency information is recorded as deviation information, and a fingerprint deviation result Devi is formed based on the occurrence position and number of each deviation information.

[0128] After obtaining the fingerprint deviation result Devi, the fingerprint parts with stable structure and the fingerprint parts that have deviated during the carbon data generation process are extracted based on the fingerprint deviation result Devi. The information of stable structure and deviation information are then uniformly organized so that each piece of deviation information corresponds to a specific fingerprint segment position and the corresponding task identifier. Subsequently, the organized stable structure information and deviation information are grouped according to the same task identifier to generate the anchor evidence set Anch, which describes the evidence structure of carbon data generation behavior.

[0129] The fingerprint deviation result Devi is used to indicate the differences between fingerprint fragments in the behavioral fingerprint set Fing and the baseline fingerprint records in the baseline fingerprint set Base. The differences include at least the differences in the number of fragments, the differences in the order of fragments, and the differences in the connection relationship of fragments. The purpose of setting the fingerprint deviation result Devi is to identify whether the carbon data generation behavior deviates from the predetermined generation pattern.

[0130] Anchor evidence set Anch is used to record evidence structure information related to carbon data generation behavior. Each piece of evidence structure information includes at least the corresponding task identifier, the corresponding fingerprint fragment position, and the corresponding deviation information. The purpose of setting anchor evidence set Anch is to store key structural information in carbon data generation behavior as evidence, thereby providing a basis for subsequent carbon data authenticity verification.

[0131] It should be noted that:

[0132] To enable implementers to establish and use the baseline fingerprint set Base according to this solution, the construction method of the baseline fingerprint set Base is further explained as follows.

[0133] First, during the initial system operation phase, multiple carbon data generation tasks that have been manually verified or system-verified to have no abnormal behavior are selected from historical carbon data generation tasks. Then, a corresponding behavioral fingerprint set (Fing) is read from each of these carbon data generation tasks. Within the behavioral fingerprint set (Fing), each carbon data generation task corresponds to one or more fingerprint fragments, which describe the complete behavioral structure of that carbon data generation task.

[0134] Subsequently, the fingerprint fragments corresponding to multiple confirmed and normal carbon data generation tasks were classified and organized according to their corresponding task identifiers, so that fingerprint fragments of the same type of carbon data generation task formed the same fingerprint template structure. Then, multiple fingerprint fragments of the same type of carbon data generation task underwent structural alignment processing to ensure that fingerprint fragments with the same process structure maintained consistency in fragment order, and the recurring stable fragment structures were extracted to generate the corresponding baseline fingerprint template.

[0135] After obtaining multiple benchmark fingerprint templates, each benchmark fingerprint template is associated with its corresponding task identifier type and stored to form a benchmark fingerprint set Base for describing the normal carbon data generation behavior pattern.

[0136] During system operation, when a new carbon data generation task is completed and passes the consistency check, the fingerprint fragment in the behavioral fingerprint set Fing corresponding to the task can be matched with the existing benchmark fingerprint set Base. When the structural consistency between the fingerprint fragment and the existing benchmark fingerprint template meets the preset stability conditions, the fingerprint fragment can be added to the benchmark fingerprint set Base as a new sample, thereby dynamically updating the benchmark fingerprint set Base so that the benchmark fingerprint set Base can continuously reflect the stable behavioral pattern of the real carbon data generation process.

[0137] The baseline fingerprint set Base established in the above manner actually records the behavioral structural features common to multiple normal carbon data generation tasks. This allows the system to perform a consistency comparison between the fingerprint fragments in the behavioral fingerprint set Fing and the baseline fingerprint templates in the baseline fingerprint set Base when executing step S41. This helps identify whether the carbon data generation behavior has deviated abnormally and generate the corresponding fingerprint deviation result Devi.

[0138] S5 includes S51;

[0139] S51. Read the evidence structure information corresponding to the same task identifier from the anchor evidence set Anch, and locate the corresponding carbon data record based on the task identifier and fingerprint fragment position recorded in each piece of evidence structure information, so that each piece of evidence structure information is associated with the corresponding carbon data record.

[0140] Subsequently, each piece of evidence structure information is bound to the corresponding carbon data record, so that the corresponding carbon data record has associated behavioral evidence content; wherein, the binding process includes writing the task identifier, fingerprint fragment position and deviation information into the associated field of the corresponding carbon data record, or establishing an index mapping relationship between the corresponding carbon data record and the evidence structure information;

[0141] After the binding process is completed, the bound evidence structure information is collected according to the task identifier and written into the carbon evidence index structure, so that the carbon data records under the same task identifier and the corresponding evidence structure information form a searchable evidence index relationship. Then, based on the index result written into the carbon evidence index structure, an output result describing the carbon data record and its corresponding behavioral evidence relationship is generated. The output result is defined as the carbon evidence result Cert.

[0142] The carbon evidence result Cert is used to represent the evidentiary result information bound to the carbon data record. The carbon evidence result Cert includes at least the corresponding task identifier, the corresponding carbon data record identifier, and the corresponding evidence index information. The purpose of setting the carbon evidence result Cert is to output the carbon data record and its behavioral evidence content in a unified manner, thereby providing directly callable evidentiary results for subsequent carbon data authenticity verification, carbon verification, and carbon audit.

[0143] It should be noted that:

[0144] The implementation method for establishing a one-to-one correspondence between each piece of evidence structure information and the corresponding carbon data record is as follows:

[0145] In the carbon data storage system, a record index is established for each carbon data record. This record index includes at least the task identifier, processing node, recording time, and carbon data record identifier. Using this record index, the carbon data records in the carbon data generation task can be grouped together according to their task identifiers.

[0146] Subsequently, the system reads the structural information of a specific piece of evidence from the anchor evidence set Anch, and extracts the task identifier and fingerprint fragment location from this structural information. The system first filters the carbon data record set belonging to the same carbon data generation task in the carbon data storage system based on the task identifier, thereby limiting the search scope.

[0147] After limiting the search scope, the specific processing position in the carbon data generation process is retrieved based on the behavioral structure information corresponding to the fingerprint fragment position. Implementers can map this position to the carbon data record at the same sequential position in the carbon data generation process by reading the sequential relationship of the processing nodes corresponding to the fingerprint fragment and its sequence number in the behavioral structure.

[0148] When the processing node of a carbon data record is consistent with the processing position corresponding to the fingerprint fragment, and the recording time of the carbon data record is within the behavior range corresponding to the fingerprint fragment, it is determined that the carbon data record matches the evidence structure information, and the carbon data record identifier is bound to the evidence structure information, thereby forming a correspondence between the carbon data record and the evidence structure information.

[0149] Subsequently, the above matching process is repeated for each piece of evidence structure information in the anchor evidence set Anch, ensuring that each piece of evidence structure information can be located to its corresponding carbon data record and form a unique binding relationship. This method establishes a one-to-one correspondence between the evidence structure information in the anchor evidence set Anch and the carbon data records, thus providing the data foundation for subsequently writing the carbon evidence index structure and generating the carbon evidence result Cert.

[0150] In this embodiment, through the consecutive execution of steps S41 and S51, this solution can directly bind the behavioral structure in the carbon data generation process to specific carbon data records, enabling carbon data to not only possess numerical results but also verifiable behavioral evidence structures, thereby forming traceable and verifiable carbon evidence results. Specifically, by comparing the fingerprint fragments in the behavioral fingerprint set Fing with the benchmark fingerprint records in the benchmark fingerprint set Base, it is possible to identify whether there are differences in the number of fragments, the order of fragments, or the connection relationship of fragments in the carbon data generation process, and to form the identified deviation information into a fingerprint deviation result Devi; subsequently, based on the fingerprint deviation result Devi, stable behavioral structures and deviation behavioral structures in the carbon data generation process are extracted, and an anchor evidence set Anch containing task identifiers, fingerprint fragment positions, and deviation information is generated. On this basis, through step S51, the evidence structure information in the anchor evidence set Anch is bound to specific carbon data records and written into the carbon evidence index structure, finally generating a carbon evidence result Cert, so that each carbon data record can correspond to behavioral evidence in its generation process. For example, in a real-world carbon emission accounting system, when a company adjusts a carbon emission record after calculating energy consumption data, although the carbon emission value in the system database may have been updated, regulators can use the carbon evidence result Cert generated by this solution to determine whether the generation behavior of the record is consistent with the normal behavior structure in the baseline fingerprint set Base, based on the recorded task identifier, the corresponding fingerprint fragment position, and the fingerprint deviation result Devi. Furthermore, the evidence index can be used to locate the behavior position of the carbon data record in the specific processing node, thereby identifying whether the carbon data has been modified afterward or abnormally generated. In this way, this solution can integrate the carbon data result with its generation behavior structure into an integrated evidence record, providing directly verifiable evidence for carbon verification, carbon auditing, and carbon trading compliance. Example

[0151] A carbon evidence detection system based on carbon evidence anchoring technology, please refer to... Figure 2 Specifically, it includes a carbon data acquisition module, a data parsing and extraction module, a feature mapping module, a consistency comparison module, and a carbon data decision module;

[0152] The carbon data acquisition module collects the raw record information generated during the carbon data generation process, and associates and organizes the collected information to form a raw data set Cdat. At the same time, it constructs a generation path set Path based on the data generation and calling relationship.

[0153] The data parsing and extraction module performs time-series parsing and change tracking on the original data set Cdat, extracting time behavior information and modification behavior information during the carbon data generation process, forming the time differential set Tdif and the modification trajectory set Trac, respectively.

[0154] The feature mapping module performs behavior mapping processing based on the generated path set Path, the time differential set Tdif, and the modified trajectory set Trac to generate a behavior feature set Behv, and constructs a behavior fingerprint set Fing based on the behavior feature set Behv.

[0155] The consistency comparison module performs a consistency comparison between the behavioral fingerprint set Fing and the pre-established benchmark fingerprint set Base to obtain the fingerprint deviation result Devi, and generates the anchor evidence set Anch based on the fingerprint deviation result Devi.

[0156] The carbon data decision module binds the anchor evidence set Anch with the corresponding carbon data record and writes it into the carbon evidence index structure, outputting the carbon evidence result Cert.

[0157] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A carbon evidence detection method based on carbon evidence anchoring technology, characterized in that: Includes the following steps: S1. Collect the raw record information generated during the carbon data generation process, and organize the collected information to form the raw data set Cdat. At the same time, construct the generation path set Path based on the data generation and calling relationship. S2. Perform time-series parsing and change tracking on the original dataset Cdat to extract time behavior information and modification behavior information during the carbon data generation process, forming the time differential set Tdif and the modification trajectory set Trac, respectively. Specifically, by binding each time interval to a corresponding node time sequence pair, each time interval corresponds to a specific node time sequence pair, thereby forming a data recording unit for describing the time change characteristics of the carbon data generation process; finally, multiple data recording units under the same task identifier are collected and organized to obtain a time differential set Tdif for describing the time behavior information of the carbon data generation process. S3. Based on the generated path set Path, the time differential set Tdif, and the modified trajectory set Trac, perform behavior mapping processing to generate the behavior feature set Behv, and construct the behavior fingerprint set Fing based on the behavior feature set Behv. S4. Perform a consistency comparison between the behavioral fingerprint set Fing and the pre-established benchmark fingerprint set Base to obtain the fingerprint deviation result Devi, and generate the anchor evidence set Anch based on the fingerprint deviation result Devi. S5. Bind the anchor evidence set Anch with the corresponding carbon data record and write it into the carbon evidence index structure, and output the carbon evidence result Cert.

2. The carbon evidence detection method based on carbon evidence anchoring technology according to claim 1, characterized in that: S1 includes S11; S11. Extract raw record information related to the carbon data generation process from the carbon emission accounting system, energy management system and log recording system, and convert the raw record information from different sources into a standard record structure. Then, using the task identifier as the basis for aggregation, multiple original record units belonging to the same carbon data generation task are merged and organized, and each original record unit is arranged in order according to the recording time, thereby forming the original data set Cdat used to describe the basic record structure of the carbon data generation process.

3. The carbon evidence detection method based on carbon evidence anchoring technology according to claim 2, characterized in that: S1 further includes S12; S12. Filter multiple raw record units with the same task identifier in the raw data set Cdat to limit the processing scope of the same carbon data generation task; then read the calling source and calling target of each raw record unit to determine the calling connection relationship between different processing nodes. Based on this, the order of the call connection relationship is determined by combining the recording time corresponding to each original recording unit, and multiple processing nodes participating in the call connection relationship are connected in chronological order to form a generation path corresponding to the same task identifier, thus forming the call correspondence relationship and the chronological order relationship.

4. A carbon evidence detection method based on carbon evidence anchoring technology according to claim 3, characterized in that: S2 includes S21; S21. Read multiple original record units with the same task identifier from the original data set Cdat to limit the time resolution range of the same carbon data generation task; then, extract the processing nodes and recording times from the multiple original record units, and arrange the multiple original record units in order according to the recording times to obtain the time sequence arrangement result of each processing node under the same task identifier. After obtaining the timing arrangement result, the time difference between adjacent original recording units is calculated sequentially to obtain the time interval; at the same time, the node connection relationship corresponding to each time interval is determined by combining the processing nodes corresponding to the adjacent original recording units to obtain the node timing pair. Then, the multiple time intervals under the same task identifier and the corresponding multiple node time series pairs are collected and organized to form the time differential set Tdif.

5. A carbon evidence detection method based on carbon evidence anchoring technology according to claim 4, characterized in that: S2 further includes S22; S22. Read multiple original record units with the same task identifier from the original data set Cdat, and filter the original record units whose record type indicates data update, data supplementation, data overwrite or data write-back as the record units to be tracked. Subsequently, the processing node, recording time, call source and call target corresponding to each record unit to be tracked are read, and multiple record units to be tracked are arranged in order according to the recording time to obtain the change time sequence result under the same task identifier; After obtaining the change timing result, based on the call source and call target of each record unit to be tracked, the transmission direction of the change behavior between different processing nodes is determined to obtain the change direction; at the same time, combined with the position of the record unit to be tracked in the change timing result, the sequential connection relationship between consecutive change behaviors is extracted to obtain the change sequence chain. Then, multiple change directions and multiple change sequence chains under the same task identifier are collected and organized to form the modification trajectory set Trac.

6. A carbon evidence detection method based on carbon evidence anchoring technology according to claim 5, characterized in that: S3 includes S31; S31. Read each generated path corresponding to the same task identifier from the generated path set Path, and extract the connection order between adjacent processing nodes in each generated path to obtain a path segment; then, read multiple time intervals and corresponding node time sequence pairs corresponding to the task identifier from the time differential set Tdif, and extract the combination relationship between each time interval and the corresponding node time sequence pair to obtain a time segment; then, read each modification behavior information corresponding to the task identifier from the modified trajectory set Trac, and extract the combination relationship between each change direction and the corresponding change sequence chain to obtain a trajectory segment; After obtaining multiple path segments, multiple time segments, and multiple trajectory segments, each path segment is aligned with the corresponding time segment in time order to determine the temporal behavior corresponding to the connection relationship of the same processing node. Then, the aligned results are associated with the trajectory segments corresponding to the process positions, so that a unified mapping relationship is established for the path behavior, time behavior and modification behavior under the same carbon data generation task. The established unified mapping relationship is then organized into multiple behavioral units, each of which represents a composite behavioral relationship in the carbon data generation process. Finally, the multiple behavioral units obtained under the same task identifier are aggregated and organized to form the behavioral feature set Behv.

7. A carbon evidence detection method based on carbon evidence anchoring technology according to claim 6, characterized in that: S3 further includes S32; S32. Read multiple behavioral units corresponding to the same task identifier from the behavioral feature set Behv, and arrange them in order according to their positions in the carbon data generation process to obtain the feature sequence chain. Subsequently, correlation analysis is performed on adjacent behavioral units in the feature sequence chain to extract the connection and continuation relationships between adjacent behavioral units and form associated fragments; After obtaining multiple associated segments, they are continuously spliced ​​together according to their arrangement in the feature sequence chain, while keeping the process position corresponding to each associated segment unchanged, so that multiple behavioral units under the same task identifier form a complete feature chain structure. Subsequently, the feature concatenation structure is uniformly encoded to generate a fingerprint description result that uniquely represents the behavioral structure of the carbon data generation task. The fingerprint description result is the fingerprint fragment, which is then collected and organized to form the behavioral fingerprint set Fing.

8. A carbon evidence detection method based on carbon evidence anchoring technology according to claim 7, characterized in that: S4 includes S41; S41. Read one or more fingerprint fragments corresponding to the same task identifier from the behavior fingerprint set Fing, and read one or more reference fingerprint records corresponding to the task identifier from the pre-established reference fingerprint set Base, so that the fingerprint fragments and the reference fingerprint records form a comparable fingerprint pair; Subsequently, a consistency comparison process is performed on each set of fingerprint pairs. Specifically, by comparing the structural arrangement and segment order relationship between the fingerprint fragment and the reference fingerprint record segment by segment, the matching and non-matching parts between the two are identified. During the comparison process, when the segment order, number of segments, or segment connection relationship in the fingerprint fragment is inconsistent with the reference fingerprint record, the corresponding inconsistency information is recorded as deviation information, and a fingerprint deviation result Devi is formed based on the occurrence position and number of each deviation information. After obtaining the fingerprint deviation result Devi, the fingerprint portion with a stable structure and the fingerprint portion that deviated during the carbon data generation process are extracted based on the fingerprint deviation result Devi. The information of the stable structure and the deviation information are then uniformly organized to generate the anchor evidence set Anch, which describes the evidence structure of the carbon data generation behavior.

9. A carbon evidence detection method based on carbon evidence anchoring technology according to claim 8, characterized in that: S5 includes S51; S51. Read the evidence structure information corresponding to the same task identifier from the anchor evidence set Anch, and locate the corresponding carbon data record based on the task identifier and fingerprint fragment position recorded in each piece of evidence structure information, so that each piece of evidence structure information is associated with the corresponding carbon data record. Subsequently, each piece of evidence structure information is bound to the corresponding carbon data record, so that the corresponding carbon data record has associated behavioral evidence content; wherein, the binding process includes writing the task identifier, fingerprint fragment position and deviation information into the associated field of the corresponding carbon data record; After the binding process is completed, the bound evidence structure information is collected according to the task identifier and written into the carbon evidence index structure, so that the carbon data records under the same task identifier and the corresponding evidence structure information form a searchable evidence index relationship. Then, based on the index result written into the carbon evidence index structure, an output result describing the carbon data record and its corresponding behavioral evidence relationship is generated. The output result is defined as the carbon evidence result Cert.

10. A carbon evidence detection system based on carbon evidence anchoring technology, applied to the carbon evidence detection method based on carbon evidence anchoring technology as described in any one of claims 1 to 9, characterized in that: It includes a carbon data acquisition module, a data parsing and extraction module, a feature mapping module, a consistency comparison module, and a carbon data decision module; The carbon data acquisition module collects the raw record information generated during the carbon data generation process, and associates and organizes the collected information to form a raw data set Cdat. At the same time, it constructs a generation path set Path based on the data generation and calling relationship. The data parsing and extraction module performs time-series parsing and change tracking on the original data set Cdat, extracting time behavior information and modification behavior information during the carbon data generation process, forming the time differential set Tdif and the modification trajectory set Trac, respectively. The feature mapping module performs behavior mapping processing based on the generated path set Path, the time differential set Tdif, and the modified trajectory set Trac to generate a behavior feature set Behv, and constructs a behavior fingerprint set Fing based on the behavior feature set Behv. The consistency comparison module performs a consistency comparison between the behavioral fingerprint set Fing and the pre-established benchmark fingerprint set Base to obtain the fingerprint deviation result Devi, and generates the anchor evidence set Anch based on the fingerprint deviation result Devi. The carbon data decision module binds the anchor evidence set Anch with the corresponding carbon data record and writes it into the carbon evidence index structure, outputting the carbon evidence result Cert.

Images