A computing power center key data identification method and system
By constructing a format model and blockchain structure in the data center, the problems of diversified data sources and credible evidence storage are solved, enabling accurate quantification and traceability of the value of computing power environment, and improving the accuracy and credibility of green environmental rights of data centers.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- STATE GRID DIGITAL TECHNOLOGY HOLDING CO LTD
- Filing Date
- 2026-05-07
- Publication Date
- 2026-07-21
AI Technical Summary
Data centers face challenges in identifying key computing power data, including diverse data sources, inconsistent formats, difficulty in integration and analysis, complex relationships between computing power and energy consumption, difficulty in accurately reflecting the environmental value of computing power due to the proportion of green electricity use, lack of flexibility and adaptability in traditional methods, imperfect data credibility and traceability mechanisms, and susceptibility to tampering with key information.
By building a format model to achieve structured processing of multi-source data, dynamically adjusting computing power conversion parameters and green electricity matching strategies, constructing a dual-chain structure of data blockchain and evidence storage blockchain, generating traceability tags and uploading them to the blockchain, and calculating and verifying hash values to ensure the accuracy and credibility of the evidence storage data.
It enables precise quantification of the value of computing power environment, meets the needs of green data center evaluation and green electricity deduction, improves the traceability efficiency and credibility of key data, and prevents key information from being tampered with.
Smart Images
Figure CN122432441A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of green environmental rights technology, and in particular to a method and system for identifying key data in computing centers. Background Technology
[0002] Currently, data centers face the following key technical challenges in identifying critical computing power data: First, data sources are diverse. Servers, storage devices, network infrastructure, power supply and distribution systems, and cooling systems each generate monitoring data in different formats. The acquisition interfaces, time granularity, and data standards of these various data types are inconsistent, making effective integration and correlation analysis difficult. Second, the relationship between computing power and energy consumption is complex. Different computing scenarios (such as artificial intelligence training, scientific computing, and cloud rendering) exhibit significantly different characteristics in their consumption of computing power, and the proportion of green electricity used directly affects the environmental value of computing power. Traditional single-metering methods cannot accurately reflect the actual composition of computing power and the matching of green electricity.
[0003] Furthermore, existing data identification methods often employ fixed rule templates, lacking the ability to flexibly adapt to dynamically changing business needs and diverse identification scenarios, resulting in low accuracy in extracting key fields. In addition, the reliable data storage and traceability mechanisms are inadequate; critical information such as computing power usage records and green electricity consumption data are easily tampered with during the transfer process, making it difficult to meet the stringent data authenticity requirements of scenarios such as carbon trading, green electricity deduction, and customer reconciliation. Summary of the Invention
[0004] The purpose of this application is to provide a method and system for identifying key data in computing centers in order to solve the above-mentioned technical problems, thereby improving the efficiency of identifying key data on green environmental rights and interests of data center computing power.
[0005] In some embodiments of this application, a format model is built to achieve structured processing of multi-source data, eliminate data silos, and dynamically adjust computing power conversion parameters and green electricity matching strategies according to the data accuracy of each data source, so as to achieve accurate quantification of the environmental value of computing power and meet the needs of refined accounting of computing power composition in scenarios such as green data center evaluation, carbon footprint tracking and green electricity deduction.
[0006] In some embodiments of this application, a dual-chain structure of data blockchain and evidence storage blockchain is constructed. Traceability tags are generated based on the associated sub-structure, and the associated subset and key fields are uploaded to the data blockchain. At the same time, the verification hash value is calculated and uploaded to the evidence storage blockchain along with the timestamp. This improves the traceability efficiency of key data, prevents key information from being tampered with, and enhances the accuracy and credibility of data center green environmental rights and value data.
[0007] In some embodiments of this application, a method for identifying key data in a computing center is provided, including: Obtain the raw data and preprocess it based on a preset format model; A rights dataset is generated based on the preprocessing results, and a key field set and related architecture are generated based on the preset identification rule model and the rights dataset. The upload strategy for key field sets is set according to the associated architecture, and the consumer identifier is determined based on the smart contract model.
[0008] In some embodiments of this application, the generation of the rights dataset includes: Establish a data source sequence W, W=(w1, w2…wi…wr), where wi is the i-th data source; r is the number of data sources; Construct the transformation sub-models for each data source in sequence; A preprocessing model is established based on all transformation sub-models; Based on all data source sequences W, wi is sequentially set as the target data source; Obtain the raw data from the target data source; Based on the preprocessing model, the raw data is transformed into multiple standard data fields, and a standard data subset of the target data source is generated; Obtain standard data subsets from each data source, and generate an equity dataset based on all standard data subsets.
[0009] In some embodiments of this application, a preset recognition rule model is included, including: Multiple recognition scenarios can be set based on all data sources; Establish a sequence A of recognition scenarios, A=(a1, a2, ..., ai, ..., an), where ai is the i-th recognition scenario and m is the number of recognition scenarios; Based on the sequence of scene identification A, ai is sequentially set as the target scene; A sub-model for identifying the target scenario is constructed based on historical identification records. The sub-model includes: a computing power conversion strategy and a green electricity matching strategy. Sequentially set up recognition sub-models for each recognition scenario; Construct a recognition rule model based on all recognition sub-models.
[0010] In some embodiments of this application, the generation of the key field set and associated schema includes: Multiple recognition requirements are set based on the raw data; Establish a sequence of identification requirements B, B=(b1,b2…bi…bm), where bi is the i-th identification requirement and m is the number of identification requirements; Based on the identification requirement sequence B, bi is sequentially set as the target requirement; Generate a related subset of the target requirements based on the equity dataset, wherein the related subset includes: multiple standard data fields; Generate key fields for the target requirements based on the associated subsets and the identification rule model; Generate key fields for each recognition requirement in sequence. Generate a set of key fields based on all key fields; Define the associated sub-architecture based on the associated subset to meet the target requirements; Each identification requirement is associated with a sub-architecture in turn, and an associated architecture is generated based on all associated sub-architectures.
[0011] In some embodiments of this application, the key field set for generating the target requirement includes: The primary scenario for constructing target requirements is defined based on the associated subset; Generate similarity values between the primary scene and each recognized scene; The sub-model for the recognition scenario corresponding to the maximum value among all similar values is set as the execution recognition model; The associated subset is input into the recognition model, and the key fields of the target requirement are output.
[0012] In some embodiments of this application, the step of setting the upload strategy for the key field set according to the association architecture includes: Build a data blockchain and a notarization blockchain; Based on the identification requirement sequence B, bi is sequentially set as the requirement to be processed; Configure the upload strategy for pending requests; Configure the upload strategy for each recognition requirement in sequence; The upload strategies for pending requests include: Generate traceability tags based on the associated sub-architecture of the requirements to be processed; Based on the traceability tags, the associated subset and key fields of the requests to be processed are uploaded to the data blockchain; Generate a verification hash value based on the associated subset and key fields; Get the verification timestamp; Upload the verification hash value and verification timestamp to the evidence storage blockchain.
[0013] In some embodiments of this application, a key data identification system for computing centers is provided, including: The sensing unit is used to acquire raw data; The central control unit includes: The first processing module is used to preprocess the raw data according to the preset format model; The first processing module is further configured to generate an equity dataset based on the preprocessing results; The second processing module is used to generate a set of key fields and a related structure based on a preset identification rule model and a rights and interests dataset. The third processing module is used to set the upload strategy for key field sets according to the associated architecture, and to determine whether to issue a consumption identifier based on the smart contract model.
[0014] In some embodiments of this application, the first processing module is further configured to: Establish a data source sequence W, W=(w1, w2…wi…wr), where wi is the i-th data source; r is the data source sequence; Construct the transformation sub-models for each data source in sequence; A preprocessing model is established based on all transformation sub-models; Based on all data source sequences W, wi is sequentially set as the target data source; Obtain the raw data from the target data source; Based on the preprocessing model, the raw data is transformed into multiple standard data fields, and a standard data subset of the target data source is generated; Obtain standard data subsets from each data source, and generate an equity dataset based on all standard data subsets.
[0015] In some embodiments of this application, the second processing module is further configured to: Multiple recognition scenarios can be set based on all data sources; Establish a sequence A of recognition scenarios, A=(a1, a2, ..., ai, ..., an), where ai is the i-th recognition scenario and m is the number of recognition scenarios; Based on the sequence of scene identification A, ai is sequentially set as the target scene; A sub-model for identifying the target scenario is constructed based on historical identification records. The sub-model includes: a computing power conversion strategy and a green electricity matching strategy. Sequentially set up recognition sub-models for each recognition scenario; Construct a recognition rule model based on all recognition sub-models.
[0016] In some embodiments of this application, the second processing module is further configured to: Multiple recognition requirements are set based on the raw data; Establish a sequence of identification requirements B, B=(b1,b2…bi…bm), where bi is the i-th identification requirement and m is the number of identification requirements; Based on the identification requirement sequence B, bi is sequentially set as the target requirement; Generate a related subset of the target requirements based on the equity dataset, wherein the related subset includes: multiple standard data fields; Generate key fields for the target requirements based on the associated subsets and the identification rule model; Generate key fields for each recognition requirement in sequence. Generate a set of key fields based on all key fields; Define the associated sub-architecture based on the associated subset to meet the target requirements; Sequentially set up the associated sub-architectures for each identification requirement, and generate the associated architecture based on all associated sub-architectures; The key field set for generating the target requirements includes: The primary scenario for constructing target requirements is defined based on the associated subset; Generate similarity values between the primary scene and each recognized scene; The sub-model for the recognition scenario corresponding to the maximum value among all similar values is set as the execution recognition model; The associated subset is input into the recognition model, and the key fields of the target requirement are output.
[0017] Compared with the prior art, the key data identification method and system for computing centers described in this application have the following advantages: By building a format model, structured processing of multi-source data is achieved, eliminating data silos. Based on the data accuracy of each data source, computing power conversion parameters and green electricity matching strategies are dynamically adjusted to achieve accurate quantification of the environmental value of computing power. This meets the needs of scenarios such as green data center evaluation, carbon footprint tracking, and green electricity deduction for refined accounting of computing power composition.
[0018] By constructing a dual-chain structure of data blockchain and evidence storage blockchain, traceability tags are generated based on the associated sub-architecture. The associated subsets and key fields are uploaded to the data blockchain, and the verification hash value is calculated and uploaded to the evidence storage blockchain along with the timestamp. This improves the traceability efficiency of key data, prevents key information from being tampered with, and enhances the accuracy and credibility of data on the value of green environmental rights in data centers. Attached Figure Description
[0019] Figure 1 This is a flowchart illustrating a method for identifying key data in a computing center, as described in a preferred embodiment of this application. Detailed Implementation
[0020] The specific embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate this application, but are not intended to limit the scope of this application.
[0021] In the description of this application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.
[0022] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.
[0023] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.
[0024] like Figure 1 As shown in the preferred embodiment of this application, a method for identifying key data in a computing center includes: S101: Obtain raw data and preprocess the raw data based on a preset format model; S102: Generate a rights dataset based on the preprocessing results, and generate a key field set and related architecture based on the preset identification rule model and the rights dataset; S103: Set the upload strategy for key field sets according to the associated architecture, and determine whether to issue a consumption identifier based on the smart contract model.
[0025] Specifically, raw data refers to the raw environmental rights data collected related to the computing power of data centers, including but not limited to: green electricity trading data, carbon emission monitoring data, online energy consumption data, and renewable energy power generation data.
[0026] Specifically, generating the equity dataset includes: Establish a data source sequence W, W=(w1, w2…wi…wr), where wi is the i-th data source; r is the data source sequence; Construct the transformation sub-models for each data source in sequence; A preprocessing model is established based on all transformation sub-models; Based on all data source sequences W, wi is sequentially set as the target data source; Obtain the raw data from the target data source; Based on the preprocessing model, the raw data is transformed into multiple standard data fields, and a standard data subset of the target data source is generated; Obtain standard data subsets from each data source, and generate an equity dataset based on all standard data subsets.
[0027] Specifically, data sources include, but are not limited to: continuous emission monitoring systems, green electricity trading platforms, green certificate trading platforms, renewable energy power consumption guarantee systems of power trading centers, data center power environment monitoring systems (DCIM), smart meters, energy management systems (EMS), distribution cabinet monitoring equipment, and other data platforms that can collect raw data on environmental rights related to data center computing power. A data source sequence is established based on all selected data platforms, where a single data source represents a data platform.
[0028] Specifically, a common metadata model is constructed based on the structural characteristics of each data source, thereby setting the transformation sub-model (i.e., the corresponding mapping transformation rules) for each data source.
[0029] Specifically, the preprocessing process includes: data format parsing and cleaning, removal of outliers and duplicate data, and conversion of the cleaned data into a standard data format according to the mapping transformation rules in the preset transformation sub-model. This generates multiple standard data fields.
[0030] Specifically, a single standard data field includes: timestamp, data source identifier, data type encoding, data subject identifier, and environmental rights value field.
[0031] It is understandable that in the above embodiments, the structured processing of multi-source data is achieved by building a format model, eliminating data silos and laying a reliable foundation for the subsequent identification of key data on green environmental rights and interests of data center computing power.
[0032] In a preferred embodiment of this application, the preset recognition rule model includes: Multiple recognition scenarios can be set based on all data sources; Establish a sequence A of recognition scenarios, A=(a1, a2, ..., ai, ..., an), where ai is the i-th recognition scenario and m is the number of recognition scenarios; Based on the sequence of scene identification A, ai is sequentially set as the target scene; A sub-model for identifying the target scenario is constructed based on historical identification records. The sub-model includes: a computing power conversion strategy and a green electricity matching strategy. Sequentially set up recognition sub-models for each recognition scenario; Construct a recognition rule model based on all recognition sub-models.
[0033] Specifically, by analyzing all data sources, multiple precision levels are generated, and basic algorithms are set for each precision level. For high-precision data sources of computing power (which can obtain real-time performance counter data of computing power components, such as CPU utilization, GPU core frequency, memory bandwidth, storage IOPS, etc.), a continuous integration algorithm is used; for medium-precision data sources of computing power (which can only obtain average utilization at the system log level), a normalized coefficient lookup table method is used; and for low-precision data sources of computing power (which can only obtain power consumption data of IT equipment), a power consumption back-calculation method is used.
[0034] Specifically, the accuracy of high-precision green electricity data sources is determined based on the granularity and time attributes of green electricity transaction data. Hourly hard matching is used for high-precision green electricity data sources, monthly weighted allocation is used for medium-precision green electricity data sources, and priority downgrading is applied to low-precision green electricity data sources, which are only used for non-physical consumption statistics and are not included in the direct deduction of computing power carbon efficiency.
[0035] Specifically, based on the different proportions of each precision data source, multiple recognition scenarios are constructed. Each recognition scenario contains data of multiple precision levels simultaneously, and the proportions of each precision level in any two recognition scenarios are not exactly the same.
[0036] Specifically, the mechanism is dynamically adjusted according to different identification scenarios to construct the optimal computing power conversion strategy and green electricity matching strategy.
[0037] It is understandable that in the above embodiments, the computing power conversion parameters and green electricity matching strategies are dynamically adjusted according to the data accuracy of each data source, so as to achieve accurate quantification of the environmental value of computing power and meet the needs of green data center evaluation, carbon footprint tracking and green electricity deduction scenarios for refined accounting of computing power composition.
[0038] In a preferred embodiment of this application, the generation of a key field set and associated architecture includes: Multiple recognition requirements are set based on the raw data; Establish a sequence of identification requirements B, B=(b1,b2…bi…bm), where bi is the i-th identification requirement and m is the number of identification requirements; Based on the identification requirement sequence B, bi is sequentially set as the target requirement; Generate a related subset based on the equity dataset to meet the target requirements. The related subset includes multiple standard data fields. Generate key fields for the target requirements based on the associated subsets and the identification rule model; Generate key fields for each recognition requirement in sequence. Generate a set of key fields based on all key fields; Define the associated sub-architecture based on the associated subset to meet the target requirements; Each identification requirement is associated with a sub-architecture in turn, and an associated architecture is generated based on all associated sub-architectures.
[0039] Specifically, multiple data subjects (i.e., data centers) and data time periods (i.e., the time span of data in the original data) are generated by analyzing the raw data. Multiple identification requirements are then generated by randomly combining the data subjects and data time periods, and the parameters corresponding to each identification requirement do not overlap. For example, an identification requirement could be key data for data center DC-001 in June.
[0040] Specifically, the system filters all standard data fields in the equity dataset to determine whether each standard data field falls within the filtering range corresponding to the target requirement. If so, the data is aggregated into the associated subset corresponding to the target requirement.
[0041] Specifically, a corresponding associated sub-architecture is constructed based on the mapping relationship of each standard data field in the associated subset corresponding to the key field.
[0042] Specifically, the associated fields include the green electricity consumption corresponding to computing power consumption, carbon emission intensity, renewable energy usage ratio, and computing power carbon efficiency index.
[0043] Specifically, the key field set for generating the target requirements includes: The primary scenario for constructing target requirements is defined based on the associated subset; Generate similarity values between the primary scene and each recognized scene; The sub-model for the recognition scenario corresponding to the maximum value among all similar values is set as the execution recognition model; The associated subset is input into the recognition model, and the key fields of the target requirement are output.
[0044] Specifically, based on the data source corresponding to each standard data field in the associated subset, the real-time data proportion of each accuracy level in the associated subset is generated. The corresponding similarity value is set according to the difference between the real-time data proportion and the data proportion corresponding to the current recognition scenario. The greater the difference, the smaller the corresponding similarity value. The mapping relationship between the two can be set according to historical parameters.
[0045] In a preferred embodiment of this application, the upload strategy for key field sets is set according to the associated architecture, including: Build a data blockchain and a notarization blockchain; Based on the identification requirement sequence B, bi is sequentially set as the requirement to be processed; Configure the upload strategy for pending requests; Configure the upload strategy for each recognition requirement in sequence; The upload strategies for pending requests include: Generate traceability tags based on the associated sub-architecture of the requirements to be processed; Based on the traceability tags, the associated subset and key fields of the requests to be processed are uploaded to the data blockchain; Generate a verification hash value based on the associated subset and key fields; Get the verification timestamp; Upload the verification hash value and verification timestamp to the evidence storage blockchain.
[0046] Specifically, traceability tags are added to each standard data field, and each standard data field with traceability tags is uploaded to the data blockchain.
[0047] Specifically, the corresponding verification hash value is generated based on the hash processing results of all standard data fields and key fields in the associated subset, and the verification timestamp and verification hash value are uploaded to the existing data volume. The cross-chain verification mechanism ensures the data consistency between the evidence storage chain and the data chain.
[0048] Specifically, determining whether to issue a consumption identifier based on a smart contract model includes: deploying a smart contract on the blockchain; when the green electricity consumption in the key field reaches a preset threshold, the smart contract automatically triggers the issuance logic, generates a green electricity consumption identifier, and records it on the blockchain; associating the issued green electricity consumption identifier with the corresponding data center's main account, and generating an issuance certificate that can be queried.
[0049] It is understood that in the above embodiments, by constructing a dual-chain structure of data blockchain and evidence storage blockchain, traceability tags are generated according to the associated sub-structure, and the associated subset and key fields are uploaded to the data blockchain. At the same time, the verification hash value is calculated and uploaded to the evidence storage blockchain along with the timestamp, thereby improving the traceability efficiency of key data, preventing key information from being tampered with, and improving the accuracy and credibility of data center green environmental rights and interests value data.
[0050] In another preferred embodiment of the key data identification method for computing centers based on any of the above preferred embodiments, this preferred embodiment provides a key data identification system for computing centers, comprising: The sensing unit is used to acquire raw data; The central control unit includes: The first processing module is used to preprocess the raw data according to the preset format model; The first processing module is also used to generate an equity dataset based on the preprocessing results; The second processing module is used to generate a set of key fields and a related structure based on a preset identification rule model and a rights and interests dataset. The third processing module is used to set the upload strategy for key field sets according to the associated architecture, and to determine whether to issue a consumption identifier based on the smart contract model.
[0051] In a preferred embodiment of this application, the first processing module is further configured to: Establish a data source sequence W, W=(w1, w2…wi…wr), where wi is the i-th data source; r is the data source sequence; Construct the transformation sub-models for each data source in sequence; A preprocessing model is established based on all transformation sub-models; Based on all data source sequences W, wi is sequentially set as the target data source; Obtain the raw data from the target data source; Based on the preprocessing model, the raw data is transformed into multiple standard data fields, and a standard data subset of the target data source is generated; Obtain standard data subsets from each data source, and generate an equity dataset based on all standard data subsets.
[0052] In a preferred embodiment of this application, the second processing module is further configured to: Multiple recognition scenarios can be set based on all data sources; Establish a sequence A of recognition scenarios, A=(a1, a2, ..., ai, ..., an), where ai is the i-th recognition scenario and m is the number of recognition scenarios; Based on the sequence of scene identification A, ai is sequentially set as the target scene; A sub-model for identifying the target scenario is constructed based on historical identification records. The sub-model includes: a computing power conversion strategy and a green electricity matching strategy. Sequentially set up recognition sub-models for each recognition scenario; Construct a recognition rule model based on all recognition sub-models.
[0053] In a preferred embodiment of this application, the second processing module is further configured to: Multiple recognition requirements are set based on the raw data; Establish a sequence of identification requirements B, B=(b1,b2…bi…bm), where bi is the i-th identification requirement and m is the number of identification requirements; Based on the identification requirement sequence B, bi is sequentially set as the target requirement; Generate a related subset based on the equity dataset to meet the target requirements. The related subset includes multiple standard data fields. Generate key fields for the target requirements based on the associated subsets and the identification rule model; Generate key fields for each recognition requirement in sequence. Generate a set of key fields based on all key fields; Define the associated sub-architecture based on the associated subset to meet the target requirements; Sequentially set up the associated sub-architectures for each identification requirement, and generate the associated architecture based on all associated sub-architectures; The key field set for generating the target requirements includes: The primary scenario for constructing target requirements is defined based on the associated subset; Generate similarity values between the primary scene and each recognized scene; The sub-model for the recognition scenario corresponding to the maximum value among all similar values is set as the execution recognition model; The associated subset is input into the recognition model, and the key fields of the target requirement are output.
[0054] Based on the first concept of this application, a format model is built to achieve structured processing of multi-source data, eliminating data silos. The computing power conversion parameters and green electricity matching strategies are dynamically adjusted according to the data accuracy of each data source, so as to achieve accurate quantification of the environmental value of computing power and meet the needs of refined accounting of computing power composition in scenarios such as green data center evaluation, carbon footprint tracking and green electricity deduction.
[0055] According to the second concept of this application, by constructing a dual-chain structure of data blockchain and evidence storage blockchain, traceability tags are generated based on the associated sub-architecture. The associated subset and key fields are uploaded to the data blockchain, and the verification hash value is calculated and uploaded to the evidence storage blockchain along with the timestamp. This improves the traceability efficiency of key data, prevents key information from being tampered with, and enhances the accuracy and credibility of data on the value of green environmental rights in data centers.
[0056] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and substitutions can be made without departing from the technical principles of this application, and these improvements and substitutions should also be considered within the scope of protection of this application.
Claims
1. A method for identifying key data in a computing center, characterized in that, include: Obtain the raw data and preprocess it based on a preset format model; A rights dataset is generated based on the preprocessing results, and a key field set and related architecture are generated based on the preset identification rule model and the rights dataset. The upload strategy for key field sets is set according to the associated architecture, and the consumer identifier is determined based on the smart contract model.
2. The method for identifying key data in a computing center as described in claim 1, characterized in that, The generated equity dataset includes: Establish a data source sequence W, W=(w1, w2…wi…wr), where wi is the i-th data source; r is the data source sequence; Construct the transformation sub-models for each data source in sequence; A preprocessing model is established based on all transformation sub-models; Based on all data source sequences W, wi is sequentially set as the target data source; Obtain the raw data from the target data source; Based on the preprocessing model, the raw data is transformed into multiple standard data fields, and a standard data subset of the target data source is generated; Obtain standard data subsets from each data source, and generate an equity dataset based on all standard data subsets.
3. The method for identifying key data in a computing center as described in claim 2, characterized in that, The preset recognition rule model includes: Multiple recognition scenarios can be set based on all data sources; Establish a sequence A of recognition scenarios, A=(a1, a2, ..., ai, ..., an), where ai is the i-th recognition scenario and m is the number of recognition scenarios; Based on the sequence of scene identification A, ai is sequentially set as the target scene; A sub-model for identifying the target scenario is constructed based on historical identification records. The sub-model includes: a computing power conversion strategy and a green electricity matching strategy. Sequentially set up recognition sub-models for each recognition scenario; Construct a recognition rule model based on all recognition sub-models.
4. The method for identifying key data in a computing center as described in claim 3, characterized in that, The generation of the key field set and associated schema includes: Multiple recognition requirements are set based on the raw data; Establish a sequence of identification requirements B, B=(b1,b2…bi…bm), where bi is the i-th identification requirement and m is the number of identification requirements; Based on the identification requirement sequence B, bi is sequentially set as the target requirement; Generate a related subset of the target requirements based on the equity dataset, wherein the related subset includes: multiple standard data fields; Generate key fields for the target requirements based on the associated subsets and the identification rule model; Generate key fields for each recognition requirement in sequence. Generate a set of key fields based on all key fields; Define the associated sub-architecture based on the associated subset to meet the target requirements; Each identification requirement is associated with a sub-architecture in turn, and an associated architecture is generated based on all associated sub-architectures.
5. The method for identifying key data in a computing center as described in claim 4, characterized in that, The key field set for generating the target requirements includes: The primary scenario for constructing target requirements is defined based on the associated subset; Generate similarity values between the primary scene and each recognized scene; The sub-model for the recognition scenario corresponding to the maximum value among all similar values is set as the execution recognition model; The associated subset is input into the recognition model, and the key fields of the target requirement are output.
6. The method for identifying key data in a computing center as described in claim 4, characterized in that, The upload strategy for setting key field sets based on the associated architecture includes: Build a data blockchain and a notarization blockchain; Based on the identification requirement sequence B, bi is sequentially set as the requirement to be processed; Configure the upload strategy for pending requests; Configure the upload strategy for each recognition requirement in sequence; The upload strategies for pending requests include: Generate traceability tags based on the associated sub-architecture of the requirements to be processed; Based on the traceability tags, the associated subset and key fields of the requests to be processed are uploaded to the data blockchain; Generate a verification hash value based on the associated subset and key fields; Get the verification timestamp; Upload the verification hash value and verification timestamp to the evidence storage blockchain.
7. A key data identification system for computing power centers, employing the key data identification method for computing power centers as described in any one of claims 1-6, characterized in that, include: The sensing unit is used to acquire raw data; The central control unit includes: The first processing module is used to preprocess the raw data according to the preset format model; The first processing module is also used to generate an equity dataset based on the preprocessing results; The second processing module is used to generate a set of key fields and a related structure based on a preset identification rule model and a rights and interests dataset. The third processing module is used to set the upload strategy for key field sets according to the associated architecture, and to determine whether to issue a consumption identifier based on the smart contract model.
8. The key data identification system for computing centers as described in claim 7, characterized in that, The first processing module is also used for: Establish a data source sequence W, W=(w1, w2…wi…wr), where wi is the i-th data source; r is the data source sequence; Construct the transformation sub-models for each data source in sequence; A preprocessing model is established based on all transformation sub-models; Based on all data source sequences W, wi is sequentially set as the target data source; Obtain the raw data from the target data source; Based on the preprocessing model, the raw data is transformed into multiple standard data fields, and a standard data subset of the target data source is generated; Obtain standard data subsets from each data source, and generate an equity dataset based on all standard data subsets.
9. The key data identification system for computing centers as described in claim 8, characterized in that, The second processing module is also used for: Multiple recognition scenarios can be set based on all data sources; Establish a sequence A of recognition scenarios, A=(a1, a2…ai…an), where ai is the i-th recognition scenario; m is the number of recognition scenarios; Based on the sequence of scene identification A, ai is sequentially set as the target scene; A sub-model for identifying the target scenario is constructed based on historical identification records. The sub-model includes: a computing power conversion strategy and a green electricity matching strategy. Sequentially set up recognition sub-models for each recognition scenario; Construct a recognition rule model based on all recognition sub-models.
10. The key data identification system for computing centers as described in claim 9, characterized in that, The second processing module is also used for: Multiple recognition requirements are set based on the raw data; Establish a sequence of identification requirements B, B=(b1,b2…bi…bm), where bi is the i-th identification requirement and m is the number of identification requirements; Based on the identification requirement sequence B, bi is sequentially set as the target requirement; Generate a related subset of the target requirements based on the equity dataset, wherein the related subset includes: multiple standard data fields; Generate key fields for the target requirements based on the associated subsets and the identification rule model; Generate key fields for each recognition requirement in sequence. Generate a set of key fields based on all key fields; Define the associated sub-architecture based on the associated subset to meet the target requirements; Sequentially set up the associated sub-architectures for each identification requirement, and generate the associated architecture based on all associated sub-architectures; The key field set for generating the target requirements includes: The primary scenario for constructing target requirements is defined based on the associated subset; Generate similarity values between the primary scene and each recognized scene; The sub-model for the recognition scenario corresponding to the maximum value among all similar values is set as the execution recognition model; The associated subset is input into the recognition model, and the key fields of the target requirement are output.