Blockchain technology-based standard system efficiency evaluation data traceability method

Abstract

This invention discloses a method for tracing the source of standard system performance evaluation data based on blockchain technology. The method includes: storing raw data in partitions according to the revision timestamps of the clauses; encrypting the storage path of the raw data in each partition; performing data change analysis to obtain the structural change time points and fluctuation change time points of the performance evaluation data to be traced; matching the standard system number with the structural change time points and decrypting to obtain the storage path of the raw data to be traced; verifying the implementation entity's fingerprint; matching the parameter update timestamps of the evaluation model information with the fluctuation change time points; tracing the parameter updates that caused the fluctuation changes in the performance evaluation data; calculating data relevance; and tracing the revisions that caused the structural changes in the performance evaluation data. This method provides a complete, automated, and refined methodology for data change attribution and tracing, and can provide high-quality data insights and decision support for the scientific formulation, continuous optimization, and implementation supervision of standards.

Description

Technical Field

[0001] This invention relates to the intersection of information technology and standardization management, and in particular to a data traceability method for evaluating the effectiveness of a standard system based on blockchain technology. Background Technology

[0002] With the acceleration of industrial digital transformation, various standard systems have become key infrastructures for ensuring product quality, regulating industry order, and improving governance capabilities. The effectiveness evaluation data of the standard system can provide decision-making basis for policy formulation, resource allocation, and continuous improvement. Ensuring the authenticity, relevance, and traceability of the effectiveness evaluation data, and accurately identifying whether changes in evaluation results stem from structural adjustments of the standard system itself, optimization of evaluation model parameters, or fluctuations in the external environment, is of great significance for evaluating the effectiveness of standard implementation and improving standards.

[0003] While existing technologies offer solutions for data storage using blockchain, they still suffer from several significant drawbacks: First, the data organization is static, failing to correlate with the dynamic revision process of the standards system and thus unable to reflect the specific standard version followed by the data, potentially leading to misaligned evaluation criteria. Second, the data has weak correlation; key elements such as original data, evaluation models, implementing entities, and standard clauses are isolated from each other, failing to form a verifiable and complete chain of evidence. Third, the granularity of traceability is coarse, only able to trace the existence of data, unable to further distinguish whether data changes stem from "structural changes" caused by standard revisions or "fluctuation changes" caused by model parameter adjustments. Fourth, data security and access control largely rely on centralized systems, posing single points of failure and tampering risks. Therefore, this invention proposes a data traceability method for standard system effectiveness evaluation based on blockchain technology. Through a dynamic encrypted path of "revision feature-timestamp-system number" triple and the introduction of a dual traceability mechanism, combined with the implementation entity fingerprint verification and relevance calculation model, it achieves for the first time penetrating traceability from macro evaluation results to micro clause revisions. This not only ensures the immutability of data but also improves the accuracy and efficiency of traceability. It provides solid technical support for the continuous optimization of the standard system and the credibility verification of evaluation results, and is of great significance for promoting the evolution of standardized management towards intelligence and transparency. Summary of the Invention

[0004] The purpose of this invention is to provide a data traceability method for evaluating the effectiveness of a standard system based on blockchain technology.

[0005] To achieve the above objectives, the present invention is implemented according to the following technical solution:

[0006] This invention includes the following steps:

[0007] Extract revision logs from each standard system to obtain clause revision timestamps and clause revision information. Divide the raw data under each standard system into sections according to clause revision timestamps. Associate the raw data of each section with the implementation entity fingerprint and package it with the evaluation model information and store it in sections.

[0008] The original data storage path for each interval is encrypted based on the revision operation characteristics, clause revision timestamp, and standard system number, and the encrypted storage path is standardized.

[0009] Data change analysis is performed on the data to be traced for effectiveness assessment to obtain the time points of structural changes and fluctuations. The closeness of the time points of structural changes to the revision timestamps of each clause is calculated to determine the standard system number. The storage path of the original data to be traced is obtained by decrypting the data according to the time period to be traced.

[0010] Extract the original data to be traced and the fingerprint assessment model information of the implementing entity, verify the fingerprint of the implementing entity, match the parameter update timestamp of the assessment model information according to the time point of fluctuation and change, and update the parameters of the performance assessment data fluctuation caused by the traceability.

[0011] Calculate the correlation between the revised content and the characteristics of changes in the original data and the characteristics of changes in the performance evaluation data, respectively, and trace the source of the revised content that caused the changes in the performance evaluation data structure;

[0012] The revised terms information is publicly available data, including revision operation characteristics and revision content; the implementing entity fingerprint is determined by the implementing entity's identity code, implementation scope, and implementation time period; the evaluation model information includes the model identifier code, parameter update timestamp, and parameter update information.

[0013] Furthermore, the method for encrypting the original data storage path for each interval includes:

[0014] A derived obfuscation key is generated based on the revision operation characteristics, clause revision timestamp, and standard system number. The expression is:

[0015] ;

[0016] in To derive the obfuscated key, Derived function for key. The structure hash function is SHA-256. For standard system numbering, For content hash function, For the timestamp of the clause revision, For the first The weight coefficients of each operational feature, For the first One operational feature, For the number of operational features, To fix the context string, The length of the derived key;

[0017] The original path is divided into a list of path components according to the directory hierarchy. Each path component is combined with a derived obfuscation key to generate an obfuscation component identifier. The obfuscation component identifiers are concatenated to form the encrypted path expression as follows:

[0018] ;

[0019] ;

[0020] in For encrypted paths, For the first An obfuscation component identifier, 32 characters long. For content hash function, For the first A path component.

[0021] Furthermore, the method for performing data change analysis includes the following steps:

[0022] After standardizing the data for performance evaluation, the discrete wavelet transform is used to decompose each type of performance evaluation data into components of different scales, as expressed in the following expression:

[0023] ;

[0024] in for Time of the first Similar performance evaluation data, As approximate components, It is a large-scale cycle. For intermediate scale details, It is a mesoscale periodic index. For mesoscale quantities;

[0025] Calculate the structural change index within a large-scale periodic window, and then calculate the Bayesian posterior probability based on the structural change index. Calculate the dynamic threshold based on the statistical characteristics of the performance evaluation data. Finally, determine the structural change points for each type of performance evaluation data based on the structural change index, the Bayesian posterior probability, and the dynamic threshold. The expression is as follows:

[0026] ;

[0027] ;

[0028] ;

[0029] in for Time of the first The structural change index of similar performance evaluation data, For the corresponding posterior probability, For dynamic thresholds, This represents the length of a large-scale periodic window. It is a function of standard deviation. The sensitivity coefficient for the structural change index. The benchmark threshold for the structural change index. It is an exponential moving average, with the time constant being the same as the window length. Dynamic threshold adjustment coefficient

[0030] Calculate the volatility change index within the mesoscale periodic window, extract the time-frequency characteristics of various performance evaluation data to calculate auxiliary feature scores, and determine the volatility change points of each type of performance evaluation data based on the volatility change index and auxiliary feature scores. The expression is as follows:

[0031] ;

[0032] ;

[0033] ;

[0034] in Time of the first A volatility index for performance evaluation data. To assist in feature scoring, These are time-frequency characteristics, including the difference between short-term and long-term means, volatility ratio, relative interquartile range, autocorrelation, skewness, and kurtosis. Time-frequency characteristics The weight, For the first Mesoscale equivalent of performance evaluation data Time-frequency characteristics Abnormal scores, for Time-frequency characteristics Values, This represents the long-term statistical mean of the time-frequency characteristics. For time-frequency characteristic sensitivity parameters, It has time-frequency characteristics;

[0035] The structural change points of all types of performance evaluation data are combined to obtain the time points of structural change in the performance evaluation data to be traced. The fluctuation change points of all types of performance evaluation data are combined to obtain the time points of fluctuation change in the performance evaluation data to be traced.

[0036] Furthermore, the method for obtaining the storage path of the original data to be traced includes:

[0037] Obtain publicly available revision information for each clause in the same field, calculate the closeness between the time point of the change in the data structure of the traceability effectiveness assessment data and the revision timestamp of each clause, and select the standard system corresponding to the revision timestamp of the clause with the highest closeness as the standard system to be traced.

[0038] Determine the time period to be traced, extract the revision operation characteristics, clause revision timestamps, and standard system numbers corresponding to the standard system to be traced, and reconstruct the corresponding derived obfuscation key; the starting point of the clause revision timestamp is the timestamp to the left of the starting point of the time period to be traced and the timestamp to the right of the ending point of the time period to be traced; the clause revision timestamp covers at least one data time partition;

[0039] The encrypted path is divided into a list of obfuscated component identifiers. Each obfuscated component identifier is decoded and hashed to obtain its hash value. The derived obfuscated key, the hash value of the obfuscated component identifier, and the standard system number are used as candidate original components. Strings that match the conditions are searched to obtain path components. The path components are then concatenated to generate the original data storage path.

[0040] Furthermore, the method for updating parameters that cause fluctuations in performance evaluation data due to the source tracing includes:

[0041] Extract the parameter update timestamps from the evaluation model information, select the time point of fluctuation change to be traced, calculate the two parameter update timestamps closest to the time point of fluctuation change to be traced, select the parameter update timestamp of the previous moment, and extract the matching evaluation model parameter update information associated with the parameter update timestamps to analyze the fluctuation changes in the effectiveness evaluation data.

[0042] Furthermore, the method for revising the content of the performance evaluation data structure caused by the source tracing includes:

[0043] The time point of the structural change to be traced is determined, and the original data and performance evaluation data before and after the corresponding time point are extracted to calculate the change characteristics of the original data and the change characteristics of the performance evaluation data. The change characteristics of the original data include mean deviation, variance change rate, distribution pattern distance, extreme value deviation rate, and autocorrelation decay rate. The change characteristics of the performance evaluation data include indicator level migration, trend direction consistency, stability change index, and achievement rate change magnitude.

[0044] Select the two clause revision timestamps closest to the time point of the structural change to be traced, select the previous clause revision timestamp, extract the clause revision content, and calculate the correlation between the clause revision content and the original data change characteristics, and the correlation between the clause revision content and the performance evaluation data change characteristics, respectively; the correlation between the clause revision content and the original data change characteristics is calculated based on mutual information; the correlation between the clause revision content and the performance evaluation data change characteristics is calculated based on the semantic correlation of cosine similarity.

[0045] The weighted average of the two types of relevance determines the specific revisions that would cause changes in the data structure of a certain type of performance evaluation.

[0046] The beneficial effects of this invention are:

[0047] This invention is a data traceability method for evaluating the performance of a standard system based on blockchain technology. Compared with existing technologies, this invention has the following technical advantages:

[0048] This invention extracts standard revision logs, uses revision timestamps as anchors to dynamically partition and store the original data, and associates the implementation entity fingerprint with the evaluation model information, so that each piece of data naturally carries the credible identifiers of its effective "standard version", "implementer" and "evaluation tool", ensuring the consistency of data with the evolution of the standard system from the source.

[0049] This invention uses revision operation features, timestamps, and system numbers to encrypt the storage path. This encryption method is strongly correlated with specific standard revision events, thereby enhancing the environmental security of data storage.

[0050] This invention performs dual detection of "structural changes" and "fluctuation changes" in performance evaluation data. By calculating the "closeness" to the standard revision timestamp and matching the evaluation model's "parameter update timestamp," it can accurately distinguish and trace the root cause of data changes—whether it is the macro-level impact of standard clause revisions or the micro-level disturbances of model parameter adjustments.

[0051] After obtaining the original data packet through the decryption path, this invention can verify the fingerprint of the implementing entity to ensure the credibility of the data source, and can accurately match the specific parameter update record that caused the fluctuation. At the same time, by calculating the correlation between the revised content and the data change characteristics, the specific clause that caused the structural change can be located. This ability to trace the source from macro changes to specific clauses / parameters greatly improves the accuracy of problem location.

[0052] The entire solution relies on the immutability and traceability of blockchain to store key information such as data relationships, access paths, and evidence of changes on the chain. Any participant can independently verify the integrity of the data and the authenticity of the traceability chain under authorization. It does not rely on the credit of a single centralized institution and builds a distributed trust system, providing a solid technical foundation for the fairness and transparency of standard performance evaluation. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating the steps of the data traceability method for evaluating the effectiveness of a standard system based on blockchain technology, as described in this invention. Detailed Implementation

[0054] The present invention will be further described below through specific embodiments. The illustrative embodiments and descriptions herein are used to explain the present invention, but are not intended to limit the present invention.

[0055] The present invention provides a data traceability method for evaluating the effectiveness of a standard system based on blockchain technology, comprising the following steps:

[0056] like Figure 1 As shown, this embodiment includes the following steps:

[0057] Extract revision logs from each standard system to obtain clause revision timestamps and clause revision information. Divide the raw data under each standard system into sections according to clause revision timestamps. Associate the raw data of each section with the implementation entity fingerprint and package it with the evaluation model information and store it in sections.

[0058] The original data storage path for each interval is encrypted based on the revision operation characteristics, clause revision timestamp, and standard system number, and the encrypted storage path is standardized.

[0059] Data change analysis is performed on the data to be traced for effectiveness assessment to obtain the time points of structural changes and fluctuations. The closeness of the time points of structural changes to the revision timestamps of each clause is calculated to determine the standard system number. The storage path of the original data to be traced is obtained by decrypting the data according to the time period to be traced.

[0060] Extract the original data to be traced and the fingerprint assessment model information of the implementing entity, verify the fingerprint of the implementing entity, match the parameter update timestamp of the assessment model information according to the time point of fluctuation and change, and update the parameters of the performance assessment data fluctuation caused by the traceability.

[0061] Calculate the correlation between the revised content and the characteristics of changes in the original data and the characteristics of changes in the performance evaluation data, respectively, and trace the source of the revised content that caused the changes in the performance evaluation data structure;

[0062] The revised terms information is publicly available data, including revision operation characteristics and revision content; the implementing entity fingerprint is determined by the implementing entity's identity code, implementation scope, and implementation time period; the evaluation model information includes the model identifier code, parameter update timestamp, and parameter update information.

[0063] In this embodiment, the method for encrypting the original data storage path of each interval includes:

[0064] A derived obfuscation key is generated based on the revision operation characteristics, clause revision timestamp, and standard system number. The expression is:

[0065] ;

[0066] in To derive the obfuscated key, Derived function for key. The structure hash function is SHA-256. For standard system numbering, For content hash function, For the timestamp of the clause revision, For the first The weight coefficients of each operational feature, For the first One operational feature, For the number of operational features, To fix the context string, The length of the derived key;

[0067] The original path is divided into a list of path components according to the directory hierarchy. Each path component is combined with a derived obfuscation key to generate an obfuscation component identifier. The obfuscation component identifiers are concatenated to form the encrypted path expression as follows:

[0068] ;

[0069] ;

[0070] in For encrypted paths, For the first An obfuscation component identifier, 32 characters long. For content hash function, For the first One path component;

[0071] In practical evaluation, taking the smart grid and distributed energy access standards in the energy field as an example, the standard system "NB / T XXXX-2023 Distributed Power Grid Connection Technical Specification" (standard system number STD-NB / T-XXXX-2023-DG) was selected. For a revision on December 1, 2023 (the revision timestamp is 2023-12-01T00:00:00Z), three clauses "5.2.1, 5.3.5, and 6.1.2" under the two categories of "Technical Parameter Adjustment" and "Communication Protocol Update" were revised (the revision operation characteristics are [2,3]). Specifically... The revisions are as follows: (1) Clause 5.2.1 changes the “operating range of power factor of grid-connected inverter” from “should be adjustable within the range of 0.95 (leading) to 0.95 (lagging)” to “should be adjustable within the range of 0.90 (leading) to 0.90 (lagging)”; (2) Clause 5.3.5 adds a test requirement for “frequency tolerance”, clarifying that it should operate continuously within the range of 49.5Hz-50.5Hz; (3) Clause 6.1.2 updates the “remote communication protocol” from “comply with DL / T 634.5104-2009” to “should support DL / T 634.5104-2019 or a compatible version”;

[0072] The timestamp of the next revision was identified as 2024-06-01T00:00:00Z. These two revision timestamps are used as a data time partition to divide the original data. Assuming a photovoltaic power station conducted a grid-connection performance test during this time period, and considering the implementing entity's identity code (Org-91370100MA3FJABC123, i.e., the unified social credit code of XX Co., Ltd.), the scope of implementation (Site-CN-37-01-PV, i.e., the grid connection point of the No. 01 photovoltaic power station in XX Province), and the implementation time period ([2023-01-15T10:00:00Z, 2025-...),... [12-15T15:30:00Z], i.e., the start and end time of the site test), generate the implementation entity fingerprint FP=Hash(Org-91370100MA3FJABC123||Site-CN-37-01-PV||[2023-01-15T10:00:00Z,2025-12-15T15:30:00Z]), and obtain the evaluation model information selected for this data time partition (power quality evaluation model / model identification code EM-PQ-2023V2, parameter update timestamp and corresponding parameter update information / multiple updates were performed within this time period);

[0073] The implementation entity indicators, evaluation model information, and raw data within the data time partition are logically bound together, and a unique evidence hash (DataHash_P1_001) is generated for the data packet. Based on the partition logic (STD-NB / T-XXXX-2023-DG,Partition_1), the implementation entity fingerprint FP, and the data packet hash, the system generates a verifiable storage path index: blockchain_anchor / STD-NB / T-XXXX-2023-DG / partitions / (T_rev1_to_T_rev2) / fp_HASH(FP) / data_package / DataHash_P1_001.meta;

[0074] The raw data includes process compliance data (records of the actual adoption and implementation of standards in specific business operations, products, or services), output data (measurable results in terms of quality, safety, efficiency, and compatibility after the implementation of standards), economic benefit data (cost savings, increased revenue, or return on investment brought about by the application of standards), social and environmental benefit data (the impact of standards on public interests, such as data on improvements in consumer satisfaction and environmental protection indicators), and system health data (data reflecting the dynamic adaptability of the standard system itself).

[0075] Taking a revision of the STD-NB / T-XXXX-2023-DG standard system as an example (the revision start timestamp is [2023-12-01T00:00:00Z, 2024-06-01T00:00:00Z], the revision operation feature is [2,3], and the weight coefficient of the operation feature is 0.3 / 0.7), the derived key length is 256, the key derivation function is HKDF-SHA256, and the derived obfuscation key (32-byte binary) "8e3a7f2c5b9d4016e2c8a4f7b1d5e3a94c7b2f6e1a8d3c5f9b2e7a4d1c6f8e3" is generated;

[0076] The original storage path index "blockchain_anchor / STD-NB / T-XXXX-2023-DG / partitions / (T_rev1_to_T_rev2) / fp_HASH(FP) / data_package / DataHash_P1_001.meta" is decomposed into a path component list consisting of 7 path components;

[0077] Taking the encryption of the path component "blockchain_anchor" as an example, the obfuscation component identifier (64 hexadecimal characters, the first 32 characters) "e5a7c3f9b1d8e2a4c6f8b0d2e6a8c4" is generated by combining the standard system number, the derived obfuscation key and the path component index. The seven obfuscation component identifiers are then concatenated to obtain the final encrypted path.

[0078] Standard storage paths are generated according to the standard system number, implementing entity, and time period, and then encrypted.

[0079] In this embodiment, the method for performing data change analysis includes the following steps:

[0080] After standardizing the data for performance evaluation, the discrete wavelet transform is used to decompose each type of performance evaluation data into components of different scales, as expressed in the following expression:

[0081] ;

[0082] in for Time of the first Similar performance evaluation data, As approximate components, It is a large-scale cycle. For intermediate scale details, It is a mesoscale periodic index. For mesoscale quantities;

[0083] Calculate the structural change index within a large-scale periodic window, and then calculate the Bayesian posterior probability based on the structural change index. Calculate the dynamic threshold based on the statistical characteristics of the performance evaluation data. Finally, determine the structural change points for each type of performance evaluation data based on the structural change index, the Bayesian posterior probability, and the dynamic threshold. The expression is as follows:

[0084] ;

[0085] ;

[0086] ;

[0087] in for Time of the first The structural change index of similar performance evaluation data, For the corresponding posterior probability, For dynamic thresholds, This represents the length of a large-scale periodic window. It is a function of standard deviation. The sensitivity coefficient for the structural change index. The benchmark threshold for the structural change index. It is an exponential moving average, with the time constant being the same as the window length. Dynamic threshold adjustment coefficient

[0088] Calculate the volatility change index within the mesoscale periodic window, extract the time-frequency characteristics of various performance evaluation data to calculate auxiliary feature scores, and determine the volatility change points of each type of performance evaluation data based on the volatility change index and auxiliary feature scores. The expression is as follows:

[0089] ;

[0090] ;

[0091] ;

[0092] in Time of the first A volatility index for performance evaluation data. To assist in feature scoring, These are time-frequency characteristics, including the difference between short-term and long-term means, volatility ratio, relative interquartile range, autocorrelation, skewness, and kurtosis. Time-frequency characteristics The weight, For the first Mesoscale equivalent of performance evaluation data Time-frequency characteristics Abnormal scores, for Time-frequency characteristics Values, This represents the long-term statistical mean of the time-frequency characteristics. For time-frequency characteristic sensitivity parameters, It has time-frequency characteristics;

[0093] The structural change points of all types of performance evaluation data are combined to obtain the time points of structural change in the performance evaluation data to be traced; the fluctuation change points of all types of performance evaluation data are combined to obtain the time points of fluctuation change in the performance evaluation data to be traced.

[0094] In actual assessment, the acquired performance evaluation data sequence is Z-score standardized, and different discrete wavelet transforms are used for different types of performance evaluation data. The specific correspondences are as follows: technical compliance data - Daubechies 4, operational efficiency data - Symlet 6, economic data - Coiflet 3, reliability data - biorthogonal 3.5, and environmental impact data - Haar. Among them, the large-scale approximate component corresponds to a period of 30 days (reflecting long-term trends and structural changes), and the mesoscale detailed component uniformly adopts three scales, including 15-30 days, 8-15 days, and 4-8 days.

[0095] Taking the determination of structural change points in technical compliance data as an example, the sensitivity coefficient of the structural change index is taken. Structural change index benchmark threshold EMA time constant 90 days, dynamic threshold adjustment factor The structural change index at time X is calculated to be 2.5, the Bayesian posterior probability is 0.73, and the dynamic threshold is 2.6. Since 2.5*0.73=1.828<2.6, time X is not identified as a structural change point. The same method is used to scan the technical compliance data for the entire time period to determine the technical compliance structural change points.

[0096] Taking the determination of volatility change points in technical compliance data as an example, the time-frequency feature sensitivity parameter is set to [2, 1.5, 2.5, 3, 4, 4], the feature weight is set to [0.25, 0.3, 0.15, 0.2, 0.05, 0.05], the time-frequency feature sensitivity parameter is set to 2, the volatility change index threshold is set to 1.5, and the auxiliary feature score threshold is set to 0.5. Based on the three mesoscale detail components, the volatility change index at time X is calculated to be 1.143, and the auxiliary feature score is set to 0.72. Since 0.72 > 0.5 but 1.143 < 1.5 (the volatility change index and the auxiliary feature score must both be greater than the corresponding thresholds), time X is not determined to be a volatility change point. The same method is used to scan the technical compliance data for the entire time period to determine the volatility change points of technical compliance.

[0097] The intersection of structural change points across all performance evaluation data categories yields the time points of structural changes in the performance evaluation data to be traced (revisions to clauses do not necessarily trigger structural changes in all categories of performance evaluation data; taking the union provides a comprehensive view of the revision time points for each clause). Similarly, the intersection of volatility change points across all performance evaluation data categories yields the time points of volatility changes in the performance evaluation data to be traced (updates to evaluation model parameters do not necessarily trigger volatility changes in all categories of performance evaluation data; taking the union provides a comprehensive view of the update time points for all evaluation model parameters). Considering that the effective time of structural / volatile changes triggered by clause revisions / evaluation model parameter updates at the same time may differ, when taking the intersection, if the time distance between multiple structural / volatile change points is less than 10% of the corresponding time window length, it is determined that clause revisions / evaluation model parameter updates occurred at the same time, and the median value of the corresponding structural / volatile change points is taken.

[0098] In this embodiment, the method for obtaining the storage path of the original data to be traced includes:

[0099] Obtain publicly available revision information for each clause in the same field, calculate the closeness between the time point of the change in the data structure of the traceability effectiveness assessment data and the revision timestamp of each clause, and select the standard system corresponding to the revision timestamp of the clause with the highest closeness as the standard system to be traced.

[0100] Determine the time period to be traced, extract the revision operation characteristics, clause revision timestamps, and standard system numbers corresponding to the standard system to be traced, and reconstruct the corresponding derived obfuscation key; the starting point of the clause revision timestamp is the timestamp to the left of the starting point of the time period to be traced and the timestamp to the right of the ending point of the time period to be traced; the clause revision timestamp covers at least one data time partition;

[0101] The encrypted path is divided into a list of obfuscated component identifiers. Each obfuscated component identifier is decoded and hashed to obtain the hash value of the obfuscated component identifier. The derived obfuscated key, the hash value of the obfuscated component identifier, and the standard system number are used as candidate original components. Strings that match the conditions are searched to obtain path components. The path components are concatenated to generate the original data storage path.

[0102] In actual assessment, we obtain the revision information of each clause publicly available in the same field, calculate the closeness between the time point of change of the data structure to be traced and the revision timestamp of each clause, and take the standard system (STD-NB / T-XXXX-2023-DG) associated with the revision timestamp of the clause with the largest closeness of 0.89 as the standard system to be traced.

[0103] Based on the time period to be traced [2024-03-01T00:00:00Z, 2024-04-01T00:00:00Z], from the revision timestamp sequence: [2023-01-01, 2023-12-01, 2024-06-01, 2024-12-01], select the timestamp to the left of the start time and the timestamp to the right of the end time to be traced [2023-12-01T00:00:00Z, 2024-06-01T00:00:00Z]. Combined with the standard system number and the corresponding revision operation characteristics, reconstruct the derived obfuscation key as "8e3a7f2c5b9d4016e2c8a4f7b1d5e3a9". 4c7b2f6e1a8d3c5f9b2e7a4d1c6f8e3”;

[0104] The encrypted path associated with the performance evaluation data to be traced is obtained. It is divided into 7 segments of obfuscated component identifiers according to 32 hexadecimal characters. Taking the decoding of the obfuscated component identifier "e5a7c3f9b1d8e2a4c6f8b0d2e6a8c4" as an example, the 32 bytes of hexadecimal characters are restored to the 16 bytes of binary data "e5a7c3f9b1d8e2a4c6f8b0d2e6a8c4".

[0105] The derived obfuscation key, the hash value of the obfuscation component identifier, and the standard system number are used as candidate original components. Based on the data storage specification of the standard system STD-NB / T-XXXX-2023-DG, the original component set is generated using the path component mode to calculate the candidate obfuscation component identifier, match the corresponding path component, and concatenate the successfully matched path components in order to obtain the original data storage path "blockchain_anchor / STD-NB / T-XXXX-2023-DG / partitions / [T_rev1_to_T_rev2) / fp_HASH(FP) / data_package / DataHash_P1_001.meta".

[0106] In this embodiment, the method for updating parameters that cause fluctuations in performance evaluation data due to source tracing includes:

[0107] Extract the parameter update timestamps from the evaluation model information, select the time point of fluctuation change to be traced, calculate the two parameter update timestamps closest to the time point of fluctuation change to be traced, select the parameter update timestamp of the previous moment, and extract the matching evaluation model parameter update information associated with the parameter update timestamps to analyze the fluctuation changes in the effectiveness evaluation data.

[0108] In the actual evaluation, assuming the original data storage path has been obtained, the corresponding evaluation model parameter update information is obtained (each timestamp is associated with detailed update data records): 2024-01-15T10:00:00Z - First monthly calibration, 2024-02-01T14:30:00Z - Special adjustment for Spring Festival, 2024-03-01T09:00:00Z - Spring standard adaptation, 2024-03-15T16:45:00Z - Emergency security patch, 2024-04-01T10:00:00Z - Second quarter baseline update, 2024-05-01T10:00:00Z - Optimization before Labor Day, 2024-05-20T11:20:00Z - Standard revision response update, 2024-06-01T10:00:00Z - Summer operation mode enabled;

[0109] The time point of the fluctuation change to be traced is determined as 2024-03-29T00:00:00Z. The parameter update timestamp sequence is searched for the two timestamps closest to the time point of the fluctuation change to be traced: 2024-03-15T16:45:00Z (the previous time) and 2024-04-01T10:00:00Z (the next time). The corresponding parameter update information is extracted from the previous timestamp, and the fluctuation change correlation analysis is performed.

[0110] In addition, another example of tracing the source is provided: for performance evaluation data at a certain moment, explore the evaluation model parameter version and evaluation model parameter information at the time when the data was generated, select the time to be traced, find the most recent timestamp before the time to be traced in the parameter update timestamp sequence, and extract the corresponding evaluation model information.

[0111] In this embodiment, the method for revising the content of the performance evaluation data structure caused by the source tracing includes:

[0112] The time point of the structural change to be traced is determined, and the original data and performance evaluation data before and after the corresponding time point are extracted to calculate the change characteristics of the original data and the change characteristics of the performance evaluation data. The change characteristics of the original data include mean deviation, variance change rate, distribution pattern distance, extreme value deviation rate, and autocorrelation decay rate. The change characteristics of the performance evaluation data include indicator level migration, trend direction consistency, stability change index, and achievement rate change magnitude.

[0113] Select the two clause revision timestamps closest to the time point of the structural change to be traced, select the previous clause revision timestamp, extract the clause revision content, and calculate the correlation between the clause revision content and the original data change characteristics, and the correlation between the clause revision content and the performance evaluation data change characteristics, respectively; the correlation between the clause revision content and the original data change characteristics is calculated based on mutual information; the correlation between the clause revision content and the performance evaluation data change characteristics is calculated based on the semantic correlation of cosine similarity.

[0114] The two types of correlation are weighted to determine the specific revisions that cause changes in the data structure of a certain type of performance evaluation;

[0115] In the actual assessment, given that the original data storage path (standard system STD-NB / T-XXXX-2023-DG) has been obtained, the structural change time point of the technical compliance data under the performance assessment data, 2024-03-18, is traced back. The publicly available revision log of the standard system "STD-NB / T-XXXX-2023-DG Distributed Power Grid Connection Technical Specification" is extracted, the most recent clause revision timestamp 2024-03-15T10:00:00Z is determined, and the corresponding revision content is extracted.

[0116] Based on the original data and performance evaluation data before and after March 18, 2024, calculate the characteristics of change in the original data and the characteristics of change in the performance evaluation data;

[0117] The revised content is vectorized into text, and the original data change features are discretized. The mutual information between each parameter of the revised content vector and the original data change features of all technical compliance is calculated and normalized to obtain the correlation between each revised content and the original data change features of technical compliance.

[0118] The BERT model is used to encode each revision, construct and encode the change characteristics of the technical compliance effectiveness assessment data, calculate the cosine similarity between the code of each revision and the code of all technical compliance effectiveness assessment data change characteristics, and obtain the correlation between each revision and the change characteristics of the technical compliance effectiveness assessment data.

[0119] The two types of correlation were weighted at 0.4 and 0.6, respectively. The revision corresponding to the highest weighted correlation of 0.616 (the revision of Clause 5.2.1 from "the operating range of the power factor of the grid-connected inverter should be adjustable from 0.95 (leading) to 0.95 (lagging)" to "it should be adjustable from 0.90 (leading) to 0.90 (lagging)") was taken as the source factor causing the structural change in the performance evaluation data at the time point 2024-03-18.

[0120] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A data traceability method for evaluating the effectiveness of a standard system based on blockchain technology, characterized in that: Includes the following steps: S1. Extract the revision logs of each standard system to obtain the clause revision timestamps and clause revision information. Divide the original data under each standard system into sections according to the clause revision timestamps. Associate the original data of each section with the implementation entity fingerprint and package it with the evaluation model information and store it in sections. S2. Encrypt the original data storage path of each interval according to the revision operation characteristics, clause revision timestamp and standard system number, and standardize the encrypted storage path. S3. Perform data change analysis on the data to be traced for effectiveness assessment, obtain the time points of structural changes and fluctuations, calculate the closeness between the time points of structural changes and the revision timestamps of each clause to determine the standard system number, and decrypt the original data storage path to be traced based on the time period to be traced. S4. Extract the original data to be traced, the fingerprint of the implementing entity, and the evaluation model information; verify the fingerprint of the implementing entity; match the parameter update timestamp of the evaluation model information according to the time point of fluctuation and change; and update the parameters of the performance evaluation data fluctuation caused by the traceability. S5. Calculate the correlation between the revised content and the characteristics of changes in the original data and the characteristics of changes in the performance evaluation data, respectively, and trace the source of the revised content that caused the changes in the performance evaluation data structure; The revised terms information is publicly available data, including revision operation characteristics and revision content; the implementing entity fingerprint is determined by the implementing entity's identity code, implementation scope, and implementation time period; the evaluation model information includes the model identifier code, parameter update timestamp, and parameter update information.

2. The data traceability method for evaluating the effectiveness of a standard system based on blockchain technology according to claim 1, characterized in that, The method for encrypting the original data storage path of each interval includes: A derived obfuscation key is generated based on the revision operation characteristics, clause revision timestamp, and standard system number. The expression is: ； in To derive the obfuscated key, Derived function for key. The structure hash function is SHA-256. For standard system numbering, For content hash function, For the timestamp of the clause revision, For the first The weight coefficients of each operational feature, For the first One operational feature, For the number of operational features, To fix the context string, The length of the derived key; The original path is divided into a list of path components according to the directory hierarchy. Each path component is combined with a derived obfuscation key to generate an obfuscation component identifier. The obfuscation component identifiers are concatenated to form the encrypted path expression as follows: ； ； in For encrypted paths, For the first An obfuscation component identifier, 32 characters long. For content hash function, For the first A path component.

3. The data traceability method for evaluating the effectiveness of a standard system based on blockchain technology according to claim 1, characterized in that, The method for performing data change analysis includes the following steps: After standardizing the data for performance evaluation, the discrete wavelet transform is used to decompose each type of performance evaluation data into components of different scales, as expressed in the following expression: ； in for Time of the first Similar performance evaluation data, As approximate components, It is a large-scale cycle. For intermediate scale details, It is a mesoscale periodic index. For mesoscale quantities; Calculate the structural change index within a large-scale periodic window, and then calculate the Bayesian posterior probability based on the structural change index. Calculate the dynamic threshold based on the statistical characteristics of the performance evaluation data. Finally, determine the structural change points for each type of performance evaluation data based on the structural change index, the Bayesian posterior probability, and the dynamic threshold. The expression is as follows: ； ； ； in for Time of the first The structural change index of similar performance evaluation data, For the corresponding posterior probability, For dynamic thresholds, This represents the length of a large-scale periodic window. It is a function of standard deviation. The sensitivity coefficient for the structural change index. The benchmark threshold for the structural change index. It is an exponential moving average, with the time constant being the same as the window length. Dynamic threshold adjustment coefficient Calculate the volatility change index within the mesoscale periodic window, extract the time-frequency characteristics of various performance evaluation data to calculate auxiliary feature scores, and determine the volatility change points of each type of performance evaluation data based on the volatility change index and auxiliary feature scores. The expression is as follows: ； ； ； in Time of the first A volatility index for performance evaluation data. To assist in feature scoring, These are time-frequency characteristics, including the difference between short-term and long-term means, volatility ratio, relative interquartile range, autocorrelation, skewness, and kurtosis. Time-frequency characteristics The weight, For the first Mesoscale equivalent of performance evaluation data Time-frequency characteristics Abnormal scores, for Time-frequency characteristics Values, This represents the long-term statistical mean of the time-frequency characteristics. For time-frequency characteristic sensitivity parameters, It has time-frequency characteristics; The structural change points of all types of performance evaluation data are combined to obtain the time points of structural change in the performance evaluation data to be traced. The fluctuation change points of all types of performance evaluation data are combined to obtain the time points of fluctuation change in the performance evaluation data to be traced.

4. The data traceability method for evaluating the effectiveness of a standard system based on blockchain technology according to claim 1, characterized in that, The method for obtaining the storage path of the original data to be traced includes: Obtain publicly available revision information for each clause in the same field, calculate the closeness between the time point of the change in the data structure of the traceability effectiveness assessment data and the revision timestamp of each clause, and select the standard system corresponding to the revision timestamp of the clause with the highest closeness as the standard system to be traced. Determine the time period to be traced, extract the revision operation characteristics, clause revision timestamps, and standard system numbers corresponding to the standard system to be traced, and reconstruct the corresponding derived obfuscation key; the starting point of the clause revision timestamp is the timestamp to the left of the starting point of the time period to be traced and the timestamp to the right of the ending point of the time period to be traced; the clause revision timestamp covers at least one data time partition; The encrypted path is divided into a list of obfuscated component identifiers. Each obfuscated component identifier is decoded and hashed to obtain its hash value. The derived obfuscated key, the hash value of the obfuscated component identifier, and the standard system number are used as candidate original components. Strings that match the conditions are searched to obtain path components. The path components are then concatenated to generate the original data storage path.

5. The data traceability method for evaluating the effectiveness of a standard system based on blockchain technology according to claim 1, characterized in that, The method for updating parameters that cause fluctuations in performance evaluation data due to source tracing includes: Extract the parameter update timestamps from the evaluation model information, select the time point of fluctuation change to be traced, calculate the two parameter update timestamps closest to the time point of fluctuation change to be traced, select the parameter update timestamp of the previous moment, and extract the matching evaluation model parameter update information associated with the parameter update timestamps to analyze the fluctuation changes in the effectiveness evaluation data.

6. The data traceability method for evaluating the effectiveness of a standard system based on blockchain technology according to claim 1, characterized in that, The method for revising the content of the performance evaluation data structure caused by the source tracing includes: The time point of the structural change to be traced is determined, and the original data and performance evaluation data before and after the corresponding time point are extracted to calculate the change characteristics of the original data and the change characteristics of the performance evaluation data. The change characteristics of the original data include mean deviation, variance change rate, distribution pattern distance, extreme value deviation rate, and autocorrelation decay rate. The change characteristics of the performance evaluation data include indicator level migration, trend direction consistency, stability change index, and achievement rate change magnitude. Select the two clause revision timestamps closest to the time point of the structural change to be traced, select the previous clause revision timestamp, extract the clause revision content, and calculate the correlation between the clause revision content and the original data change characteristics, and the correlation between the clause revision content and the performance evaluation data change characteristics, respectively; the correlation between the clause revision content and the original data change characteristics is calculated based on mutual information; the correlation between the clause revision content and the performance evaluation data change characteristics is calculated based on the semantic correlation of cosine similarity. The weighted average of the two types of relevance determines the specific revisions that would cause changes in the data structure of a certain type of performance evaluation.

Images