Energy and power data asset evaluation method based on artificial intelligence

Abstract

The present invention relates to an artificial intelligence-based energy and power data asset evaluation method, comprising the following steps: S1: acquiring multi-source heterogeneous data from the energy and power industry; S2: constructing a database based on a time series database and distributed storage to store the multi-source heterogeneous data; S3: constructing an energy knowledge graph based on the multi-source heterogeneous data; S4: constructing a multidimensional value assessment model to quantify the economic benefits, quality indicators, usage, and potential mining capabilities of data assets; and S5: establishing a dynamic monitoring system to conduct real-time evaluation of the value of data assets throughout their life cycle based on the multidimensional value assessment model and the energy knowledge graph. This invention effectively improves the utilization efficiency of energy data assets.

Description

Technical Field

[0001] The present invention relates to the field of data management, and in particular to an energy and power data asset evaluation method based on artificial intelligence. Background Art

[0002] With the rapid development of the Energy Internet and smart grids, the energy and power industry is facing the challenges of rapidly growing and diverse data demands. Energy systems involve a variety of data sources, including field equipment operating data, SCADA (Supervisory Control and Data Acquisition) data, historical records, market transaction data, and external environmental data (such as meteorological information and policy updates). These data types are characterized by multi-source heterogeneity, temporal dynamics, and complex correlations, making traditional data management and value assessment technologies difficult to meet current needs.

[0003] On the one hand, the integration and processing of multi-source, heterogeneous data is difficult, and data suffer from semantic inconsistencies, different formats, and quality differences, resulting in low data utilization. On the other hand, traditional data evaluation methods typically focus on a single dimension (such as efficiency or quality), failing to fully quantify the data's comprehensive value in terms of economic benefits, quality indicators, scenario adaptation, and potential mining. Furthermore, because data in energy and power systems is highly time-sensitive and dynamically changing, relying solely on static analysis methods makes it difficult to timely grasp the full lifecycle performance of data assets, let alone identify data issues or optimize potentially high-value data in real time. Summary of the Invention

[0004] In order to solve the above problems, the purpose of the present invention is to provide an energy and power data asset evaluation method based on artificial intelligence, which can not only perform real-time evaluation of the full life cycle value of data, but also provide automated problem repair and optimization suggestions, thereby significantly improving the utilization efficiency of energy data assets.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] An energy and power data asset assessment method based on artificial intelligence includes the following steps:

[0007] S1: Acquire multi-source heterogeneous data in the energy and power industry, including field equipment data, SCADA system data, historical records, market and external data;

[0008] S2: Build a database based on a time series database and distributed storage to store multi-source heterogeneous data;

[0009] S3: Build an energy knowledge graph based on multi-source heterogeneous data, modeling equipment associations, energy process chains, and market change information into a knowledge graph to provide context-enhanced information;

[0010] S4: Build a multi-dimensional value assessment model to quantify the economic benefits, quality indicators, usage, and potential mining capabilities of data assets;

[0011] S5: Establish a dynamic monitoring system to conduct real-time evaluation of the value of data assets throughout their life cycle based on a multi-dimensional value assessment model and energy knowledge graph.

[0012] Furthermore, field equipment data includes sensor data, smart meters, and relay protection device data; SCADA system data includes grid dispatch, equipment operating status, and accident alarm data; market and external data includes electricity market prices, load forecast data, policy documents, climate and environmental data; historical record data includes electricity usage history and equipment operation and maintenance logs.

[0013] Furthermore, the database adopts a tiered storage design for hot and cold data. The time series database is used to store high-frequency, real-time time series data; distributed storage is used to store large-scale, low-frequency access historical data;

[0014] In a time series database, data is partitioned based on time and device ID;

[0015] In distributed storage, data is partitioned and stored by time and type;

[0016] Use the built-in compression algorithm of the time series database for time series data to save storage space;

[0017] Regularly migrate hot data, i.e., data from the time series database, to distributed storage, and automate the migration process through Apache Nifi.

[0018] Furthermore, the energy knowledge graph is as follows:

[0019] Let nodes V={v1,v2,…,v i ,...,v n}, where v i It is the i-th entity, the number of entities is n, and the node types include: device node v device 、Market node v market 、Process node v process , external environment node v env ;

[0020] Edges represent the relationships between entities, including the connection relationship between devices, the sequence relationship of process chains, and the impact relationship between market dynamics and equipment operations. The edge set is defined as E:

[0021] ;

[0022] where e m =(va ,v b ,r) represents the connection entity v a and v b The relationship r;

[0023] Edge relationship classification: connection relationship between devices connect ;Process chain sequence relationship r process ; Market and equipment relationship market ;Relationship between environment and process env ;

[0024] The attributes of relation r include: transmission capacity C, time attribute T and directionality attribute D;

[0025] The energy knowledge graph is modeled through a directed attribute graph G:

[0026] ;

[0027] Where V is the set of nodes in the graph; E is the set of edges in the graph; A V A collection of attributes for a node;

[0028] A E The attribute set of the edge;

[0029] A directed attribute graph G, including the following subgraphs:

[0030] Device association subgraph G device , corresponding to the physical connection and logical relationship between devices:

[0031] ;

[0032] Among them, V device Indicates the device node, E connect Indicates the physical connection relationship of the device; A device Indicates the device node attributes, A connect Attributes representing the physical connection of the device;

[0033] Energy process chain diagram G process , links that model energy flows:

[0034] ;

[0035] Among them, E process Indicates the order of the process, A flow Indicates the power value of the participating nodes in the link; A process Indicates the attributes of the process node; V process Indicates a process node;

[0036] Market change subgraph G market, which represents the relationship between market rules, price fluctuations and physical systems:

[0037] ;

[0038] Among them, V market represents the market node; E market Indicates the correlation between market changes and equipment operation; A price is the characteristic factor of market price change; A policy policy attribute set;

[0039] Weather impact subgraph G env , indicating the relationship between weather and equipment operation:

[0040] ;

[0041] Among them, V env is the external environment node; E impact Indicates the impact of meteorological conditions on equipment or processes; A weather A is a collection of Qi attributes; env Environmental impact attribute set;

[0042] The energy process chain path is inferred from the path in the figure:

[0043] ;

[0044] Where P is the path; v j Represents node j;v k Represents node k;e ij Represents node v i With v j edge;e jk Represents node v j With v k the edge;

[0045] Path weight W p Expressed as the rated power loss of each link in the process:

[0046] ;

[0047] Among them, f loss Defined as a function used to calculate power loss between devices.

[0048] Furthermore, the multidimensional value assessment model G P , specifically as follows:

[0049] ;

[0050] Among them: G ecois the economic benefit sub-model; G qual is the quality index sub-model; G util is the usage sub-model; G pot It is the potential mining capability sub-model; is a weighted model;

[0051] The economic benefit sub-model is specifically:

[0052] ;

[0053]

[0054] Among them, E rev 、E cost 、E fit They are revenue improvement, cost savings and key scenario matching; R t represents the benefits brought by data support in scenario t; 、 Respectively The cost of individual data before and after optimization; 、 are the scene contribution data volume and the total data volume respectively; w1, w2, w3 are weight coefficients; T is the total number of scenes; is the total data volume;

[0055] The quality indicator sub-model is specifically:

[0056]

[0057] Among them, Q acc , Q comp , Q cons They are data accuracy, data completeness and data consistency; For the Measured values; is the predicted value; is the amount of missing data; is the amount of data that is consistent in time or logic; w4, w5, and w6 are weight coefficients;

[0058] The usage sub-model is specifically:

[0059]

[0060] Among them, U freq 、U scen 、U depth They are access frequency, scenario coverage and usage depth; is the total number of visits; For the evaluation cycle; is the number of usage scenarios; is the total number of supported scenarios; The number of data fields to be utilized; is the number of available data fields; w7, w8, and w9 are weight coefficients;

[0061] The potential mining capability sub-model,

[0062] ;

[0063] Among them, P ext 、P reuse 、P market They are scalability, reuse value and reuse value respectively; w 10 、w 11 、w 12 is the weight coefficient.

[0064] Furthermore, the dynamic monitoring system includes a data stream acquisition and storage module, a model evaluation and calculation module, a knowledge graph collaboration module, and a feedback optimization module;

[0065] The data stream collection and storage module accesses data sources through Apache Kafka or Flink, including equipment operating status, market fluctuation information, energy process chain monitoring information, and meteorological data; and performs preliminary cleaning and verification;

[0066] The model evaluation calculation module uses a multi-dimensional value evaluation model to evaluate the value of data, and performs value calculation and analysis through periodic and event-triggered mechanisms; it dynamically calls the knowledge graph to assist in the inference calculation of value;

[0067] The knowledge graph collaboration module deduces the intrinsic value or problems of the data through the association relationship between entity nodes, obtains the nodes that need to be repaired or optimized, and corrects or optimizes the data.

[0068] Furthermore, through the association relationship between entity nodes, the intrinsic value or problem of the data is derived, as follows:

[0069] Through the association relationship and attribute weights between nodes, the intrinsic value or problem of the node is derived. For a target node v i , its intrinsic value V value (v i ) is derived through the influence of neighboring nodes:

[0070] ;

[0071] Among them, w(v i, v j ) is the node v i and v j The weight of the relationship between is the neighbor node v j Attribute value of Indicates v i Neighbor nodes of

[0072] If the target node v i There is an abnormal problem, and the target node v is deduced through the association relationship of neighbor nodes i The problem :

[0073] ;

[0074] in, is the attribute difference function, used to determine the node v i and v j unusual differences between attributes;

[0075] According to time attributes and transmission capacity C to dynamically adjust node association weights :

[0076] ;

[0077] in, is the initial weight; is the time-dependent weight function; is the transmission capacity related weight function.

[0078] Furthermore, the nodes that need to be repaired or optimized are obtained, and the data is corrected or optimized, as follows:

[0079] For missing data or abnormal problems of nodes, use neighboring nodes to infer and repair data:

[0080] ;

[0081] in, A fixed (v i ) is the attribute value after repair;

[0082] Transmission path optimization, based on energy process chain diagram G process , find the optimal path to minimize power loss :

[0083] ;

[0084] Among them, P loss (e) is the power loss value on edge e;

[0085] Node load balancing, redistributing power flow to nodes with transmission loads above a threshold:

[0086] ;

[0087] Among them, P surplus (v j ) is the neighbor node v j Can be reassigned to v i The remaining power P redistribute (v i ) is the node v i The power flow that can be regained through neighboring nodes.

[0088] The present invention has the following beneficial effects:

[0089] 1. This invention not only enables real-time evaluation of the full lifecycle value of data, but also provides automated problem repair and optimization suggestions, thereby significantly improving the utilization efficiency of energy data assets;

[0090] 2. This invention constructs an energy knowledge graph based on multi-source heterogeneous data. Through the device association subgraph, energy process chain subgraph, market change subgraph, and weather impact subgraph, it can comprehensively model the complex relationships among device connections, energy process chains, market dynamics, and environmental impacts in the energy system, providing support for energy system optimization and analysis.

[0091] 3. The multidimensional value assessment model of the present invention provides a comprehensive value quantification method for data assets in the energy and power industry by combining the four dimensions of economic benefits, data quality, usage and potential mining capabilities; and by combining the multidimensional value assessment model with the energy knowledge graph, a dynamic monitoring mechanism realizes real-time value assessment of the entire life cycle of data assets, and uses the reasoning ability of the knowledge graph to assist in assessment and repair, which will further improve the accuracy, real-time nature and potential value mining capabilities of the data. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0093] The following is combined with Figure 1 The present invention is further described in detail with specific examples:

[0094] An energy and power data asset assessment method based on artificial intelligence includes the following steps:

[0095] S1: Acquire multi-source heterogeneous data in the energy and power industry, including field equipment data, SCADA system data, historical records, market and external data;

[0096] S2: Build a database based on a time series database and distributed storage to store multi-source heterogeneous data;

[0097] S3: Build an energy knowledge graph based on multi-source heterogeneous data, modeling equipment associations, energy process chains, and market change information into a knowledge graph to provide context-enhanced information;

[0098] S4: Build a multi-dimensional value assessment model to quantify the economic benefits, quality indicators, usage, and potential mining capabilities of data assets;

[0099] S5: Establish a dynamic monitoring system to conduct real-time evaluation of the value of data assets throughout their life cycle based on a multi-dimensional value assessment model and energy knowledge graph.

[0100] In this embodiment, field equipment data includes sensor data (voltage, current, frequency, etc.), smart meters, and relay protection device data; SCADA system data includes grid dispatch, equipment operating status, and accident alarm data; market and external data includes electricity market prices, load forecast data, policy documents, climate and environmental data; and historical record data includes electricity usage history and equipment operation and maintenance logs.

[0101] In this embodiment, the database adopts a tiered storage design for hot and cold data. The time series database is used to store high-frequency, real-time time series data (such as sensor data and load curves); distributed storage is used to store large-scale, low-frequency access historical data (such as equipment operation logs and policy documents);

[0102] In a time series database, data is partitioned based on time and device ID;

[0103] In distributed storage, data is partitioned and stored by time and type (e.g., directories are divided by year or device type);

[0104] Use the built-in compression algorithm of the time series database for time series data to save storage space;

[0105] Regularly migrate hot data, i.e., data from the time series database, to distributed storage, and automate the migration process through Apache Nifi.

[0106] In this embodiment, the energy knowledge graph is as follows:

[0107] Let nodes V={v1,v2,…,v i ,...,v n}, where v i is the i-th entity, the number of entities is n, and the node types include:

[0108] Device node v device: Power generation equipment, power transmission equipment, power equipment, etc.;

[0109] Market Node v market : Electricity market prices, policy rules, and load forecast results;

[0110] Process Node v process : Indicates the process link where energy flows;

[0111] External environment node v env : Meteorological data (such as temperature, wind speed) and other environmental factors;

[0112] Edges represent the relationships between entities, including the connection relationship between devices, the sequence relationship of process chains, and the impact relationship between market dynamics and equipment operations. The edge set is defined as E:

[0113] ;

[0114] where e m =(v a ,v b ,r) represents the connection entity v a and v b Relationship r; edge relationship classification:

[0115] Connection relationship between devices connect :Such as "Relay protection device connected to transformer";

[0116] Process chain sequence relationship process : Such as "the transmission equipment belongs to a specific grid area";

[0117] Market and equipment relationship market : Such as "market price fluctuations affect equipment load";

[0118] Environment and process relationship env : Such as "temperature changes affect transmission power";

[0119] The attributes of a relation r include: transmission capacity C, such as the power transmission capacity between devices; time attribute T, which indicates the time when the relation occurs (such as seasonality or real-time impact); directionality attribute D, for example, the direction of energy flow;

[0120] The energy knowledge graph is modeled through a directed attribute graph G:

[0121] ;

[0122] Where V is the set of nodes in the graph; E is the set of edges in the graph; A V A collection of attributes for a node;

[0123] AE The attribute set of the edge;

[0124] A directed attribute graph G, including the following subgraphs:

[0125] Device association subgraph G device , corresponding to the physical connection and logical relationship between devices:

[0126] ;

[0127] Among them, V device Indicates the device node, E connect Indicates the physical connection relationship of the device; A device Indicates the device node attributes, A connect Attributes representing the physical connection of the device;

[0128] Energy process chain diagram G process , links that model energy flows:

[0129] ;

[0130] Among them, E process Indicates the order of the process, A flow Indicates the power value of the participating nodes in the link; A process Indicates the attributes of the process node; V process Indicates a process node;

[0131] Market change subgraph G market , which represents the relationship between market rules, price fluctuations and physical systems:

[0132] ;

[0133] Among them, V market represents the market node; E market Indicates the correlation between market changes and equipment operation; A price is the characteristic factor of market price change; A policy policy attribute set;

[0134] Weather impact subgraph G env , indicating the relationship between weather and equipment operation:

[0135] ;

[0136] Among them, V env is the external environment node; E impact Indicates the impact of meteorological conditions on equipment or processes; A weather A is a collection of Qi attributes; env Environmental impact attribute set;

[0137] The energy process chain path is inferred from the path in the figure:

[0138] ;

[0139] Where P is the path; v j Represents node j;v k Represents node k;e ij Represents node v i With v j edge;e jk Represents node v j With v k the edge;

[0140] Path weight W p Expressed as the rated power loss of each link in the process:

[0141] ;

[0142] Among them, f loss Defined as a function used to calculate power loss between devices.

[0143] In this embodiment, the multi-dimensional value assessment model G P , specifically as follows:

[0144] ;

[0145] Among them: G eco is the economic benefit sub-model; G qual is the quality index sub-model; G util is the usage sub-model; G pot It is the potential mining capability sub-model; is a weighted model;

[0146] The economic benefit sub-model is specifically:

[0147]

[0148] Among them, E rev 、E cost 、E fit They are revenue improvement, cost savings and key scenario matching; R t represents the benefits brought by data support in scenario t; 、 Respectively The cost of individual data before and after optimization; 、 are the scene contribution data volume and the total data volume respectively; w1, w2, w3 are weight coefficients; T is the total number of scenes; is the total data volume;

[0149] The quality indicator sub-model is specifically:

[0150]

[0151] Among them, Q acc , Q comp , Q cons They are data accuracy, data completeness and data consistency; For the Measured values; is the predicted value; is the amount of missing data; is the amount of data that is consistent in time or logic; w4, w5, and w6 are weight coefficients;

[0152] The usage sub-model is specifically:

[0153]

[0154]

[0155] Among them, U freq 、U scen 、U depth They are access frequency, scenario coverage and usage depth; is the total number of visits; For the evaluation cycle; is the number of usage scenarios; is the total number of supported scenarios; The number of data fields to be utilized; is the number of available data fields; w7, w8, and w9 are weight coefficients;

[0156] The potential mining capability sub-model,

[0157]

[0158] Among them, P ext 、P reuse 、P market They are scalability, reuse value and reuse value respectively; w 10 、w 11 、w 12 is the weight coefficient.

[0159] In this embodiment, the dynamic monitoring system includes a data stream acquisition and storage module, a model evaluation and calculation module, a knowledge graph collaboration module, and a feedback optimization module;

[0160] The data stream collection and storage module accesses data sources through Apache Kafka / Flink, including equipment operating status, market fluctuation information, energy process chain monitoring information, and meteorological data; and performs preliminary cleaning and verification;

[0161] The model evaluation calculation module uses a multi-dimensional value evaluation model to evaluate the value of data, and performs value calculation and analysis through periodic and event-triggered mechanisms; it dynamically calls the knowledge graph to assist in the inference calculation of value;

[0162] The knowledge graph collaboration module deduces the intrinsic value or problems of the data through the association relationship between entity nodes, obtains the nodes that need to be repaired or optimized, and corrects or optimizes the data.

[0163] In this embodiment, the intrinsic value or problem of the data is derived through the association relationship between entity nodes, as follows:

[0164] Through the association relationship and attribute weights between nodes, the intrinsic value or problem of the node is derived. For a target node v i , its intrinsic value V value (v i ) is derived through the influence of neighboring nodes:

[0165] ;

[0166] Among them, w(v i, v j ) is the node v i and v j The weight of the relationship between is the neighbor node v j Attribute value of Indicates v i Neighbor nodes of

[0167] If the target node v i There is an abnormal problem, and the target node v is deduced through the association relationship of neighbor nodes i The problem :

[0168] ;

[0169] in, is the attribute difference function, used to determine the node v i and v j unusual differences between attributes;

[0170] According to time attributes and transmission capacity C to dynamically adjust node association weights :

[0171] ;

[0172] in, is the initial weight; is the time-dependent weight function; is the transmission capacity related weight function.

[0173] In this embodiment, the node that needs to be repaired or optimized is obtained, and the data is corrected or optimized, as follows:

[0174] For missing data or abnormal problems of nodes, use neighboring nodes to infer and repair data:

[0175] ;

[0176] in, A fixed (v i ) is the attribute value after repair;

[0177] Transmission path optimization, based on energy process chain diagram G process , find the optimal path to minimize power loss :

[0178] ;

[0179] Among them, P loss (e) is the power loss value on edge e;

[0180] Node load balancing, redistributing power flow to nodes with transmission loads above a threshold:

[0181] ;

[0182] Among them, P surplus (v j ) is the neighbor node v j Can be reassigned to v i The remaining power P redistribute (v i ) is the node v i The power flow that can be recovered through neighboring nodes;

[0183] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0184] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0185] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0186] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0187] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other manner. Any person skilled in the art may utilize the above-disclosed technical content to modify or modify the present invention into equivalent embodiments. However, any simple modifications, equivalent variations, and modifications to the above embodiments that do not depart from the technical content of the present invention and are based on the technical essence of the present invention remain within the scope of protection of the present invention.

Claims

1. The energy and power data asset evaluation method based on artificial intelligence is characterized by: The following steps are involved: S1: Acquire multi-source heterogeneous data in the energy and power industry, including field equipment data, SCADA system data, historical records, market and external data; S2: Build a database based on a time series database and distributed storage to store multi-source heterogeneous data; S3: Build an energy knowledge graph based on multi-source heterogeneous data, modeling equipment associations, energy process chains, and market change information into a knowledge graph to provide context-enhanced information; S4: Build a multi-dimensional value assessment model to quantify the economic benefits, quality indicators, usage, and potential mining capabilities of data assets; S5: Establish a dynamic monitoring system to conduct real-time evaluation of the value of data assets throughout their life cycle based on a multi-dimensional value assessment model and energy knowledge graph; The energy knowledge graph is modeled through a directed attribute graph G: G=(V,E,A V ,A E ); Where V is the set of nodes in the graph; E is the set of edges in the graph; A V A collection of attributes for a node; A E The attribute set of the edge; Directed attribute graph G, including device association subgraph G device , Energy process chain diagram G process , market change subgraph G market , meteorological impact subgraph G env ; The energy process chain path is inferred from the path in the figure: ； Where P is the path; v j represents node j; v k represents node k; e ij Represents node v i With v j edge; e jk Represents node v j With v k edge; Path weight W p Expressed as the rated power loss of each link in the process: ； Among them, f loss Defined as a function used to calculate power loss between devices; The dynamic monitoring system includes a data flow acquisition and storage module, a model evaluation and calculation module, a knowledge graph collaboration module, and a feedback optimization module; The data stream collection and storage module accesses data sources through Apache Kafka or Flink, including equipment operating status, market fluctuation information, energy process chain monitoring information, and meteorological data; and performs preliminary cleaning and verification; The model evaluation calculation module uses a multi-dimensional value evaluation model to evaluate the value of data, and performs value calculation and analysis through periodic and event-triggered mechanisms; it dynamically calls the knowledge graph to assist in the inference calculation of value; The knowledge graph collaboration module derives the intrinsic value or problems of the data through the association relationship between entity nodes, obtains the nodes that need to be repaired or optimized, and corrects or optimizes the data; The multidimensional value assessment model G P , specifically as follows: G P =f(G eco ,G qual ,G util ,G pot ) ； Among them: G eco is the economic benefit sub-model; G qual is the quality index sub-model; G util is the usage sub-model; G pot is the potential mining capability sub-model; f(•) is the weighted model.

2. The energy and power data asset evaluation method based on artificial intelligence according to claim 1 is characterized in that: The field equipment data includes sensor data, smart meters, and relay protection device data; the SCADA system data includes grid dispatch, equipment operating status, and accident alarm data; market and external data includes electricity market prices, load forecast data, policy documents, climate and environmental data; historical record data includes electricity usage history and equipment operation and maintenance logs.

3. The energy and power data asset evaluation method based on artificial intelligence according to claim 1 is characterized in that: The database adopts a tiered storage design for hot and cold data. The time series database is used to store high-frequency, real-time time series data; the distributed storage is used to store large-scale, low-frequency access historical data. In a time series database, data is partitioned based on time and device ID; In distributed storage, data is partitioned and stored by time and type; Use the built-in compression algorithm of the time series database for time series data to save storage space; Regularly migrate hot data, i.e., data from the time series database, to distributed storage, and automate the migration process through Apache Nifi.

4. The energy and power data asset evaluation method based on artificial intelligence according to claim 1 is characterized in that: The energy knowledge graph is as follows: Let nodes V={v1,v2,…,v i ,...,v n }, where v i It is the i-th entity, the number of entities is n, and the node types include: device node v device 、Market node v market 、Process node v process , external environment node v env ; Edges represent the relationships between entities, including the connection relationship between devices, the sequence relationship of process chains, and the impact relationship between market dynamics and equipment operations. The edge set is defined as E: E={e1,e2,…,e m,... ,the M }; where e m =(v a ,v b ,r) represents the connection entity v a and v b The relationship r; Edge relationship classification: connection relationship between devices connect ; Process chain sequence relationship r process ; Market and equipment relationship market ; Relationship between environment and process env ; The attributes of relation r include: transmission capacity C, time attribute T and directionality attribute D; Device association subgraph G device , corresponding to the physical connection and logical relationship between devices: G device =(V device ,E connect ,A device ,A connect ); Among them, V device Indicates the device node, E connect Indicates the physical connection relationship of the device; A device Indicates the device node attributes, A connect Attributes representing the physical connection of the device; Energy process chain diagram G process , links that model energy flows: G process =(V process ,E process ,A process ,A flow ); Among them, E process Indicates the order of the process, A flow Indicates the power value of the participating nodes in the link; A process Indicates the attributes of the process node; V process Indicates a process node; Market change subgraph G market , which represents the relationship between market rules, price fluctuations and physical systems: G market =(V market ,E market ,A price ,A policy ); Among them, V market represents the market node; E market Indicates the correlation between market changes and equipment operation; A price is the characteristic factor of market price change; A policy policy attribute set; Weather impact subgraph G env , indicating the relationship between weather and equipment operation: G env =(V env ,E impact ,A weather ,A env ); Among them, V env is the external environment node; E impact Indicates the impact of meteorological conditions on equipment or processes; A weather A is a collection of Qi attributes; env A collection of environmental impact attributes.

5. The energy and power data asset evaluation method based on artificial intelligence according to claim 1 is characterized in that: The economic benefit sub-model is specifically: G eco =w1E rev +w2E cost +w3E fit ; ； Among them, E rev 、E cost 、E fit They are revenue improvement, cost savings and key scenario matching; R t represents the benefits brought by data support in scenario t; 、 Respectively The cost of individual data before and after optimization; 、 are the scene contribution data volume and the total data volume respectively; w1, w2, w3 are weight coefficients; T is the total number of scenes; is the total data volume; The quality indicator sub-model is specifically: G qual =w4Q acc +w5Q comp +w6Q cons ; ； Among them, Q acc , Q comp , Q cons They are data accuracy, data completeness and data consistency; For the Measured values; is the predicted value; is the amount of missing data; is the amount of data that is consistent in time or logic; w4, w5, and w6 are weight coefficients; The usage sub-model is specifically: G util =w7U freq +w8U scen +w9U depth ; ； Among them, U freq 、U scen 、U depth They are access frequency, scenario coverage and usage depth; is the total number of visits; For the evaluation cycle; is the number of usage scenarios; is the total number of supported scenarios; The number of data fields to be utilized; is the number of available data fields; w7, w8, and w9 are weight coefficients; The potential mining capability sub-model, G pot =w 10 P ext +w 11 P reuse +w 12 P market ; Among them, P ext 、P reuse 、P market They are scalability, reuse value and market value; w 10 、w 11 、w 12 is the weight coefficient.

6. The energy and power data asset evaluation method based on artificial intelligence according to claim 1 is characterized in that: The intrinsic value or problem of data is derived through the association relationship between entity nodes, as follows: Through the association relationship and attribute weights between nodes, the intrinsic value or problem of the node is derived. For a target node v i , its intrinsic value V value (v i ) is derived through the influence of neighboring nodes: ； Among them, w(v i, v j ) is the node v i and v j The weight of the relationship between is the neighbor node v j Attribute value of Indicates v i Neighbor nodes of If the target node v i There is an abnormal problem, and the target node v is deduced through the association relationship of neighbor nodes i The problem : ； in, is the attribute difference function, used to determine the node v i and v j unusual differences between attributes; Dynamically adjust node relationship weights based on time attribute T and transmission capacity C : ； Among them, w0(v i ,v j ) is the initial weight; is the time-dependent weight function; f C (C) is the transmission capacity related weight function.

7. The energy and power data asset evaluation method based on artificial intelligence according to claim 6 is characterized in that: The nodes that need to be repaired or optimized are obtained, and the data are corrected or optimized, as follows: For missing data or abnormal problems of nodes, use neighboring nodes to infer and repair data: ； in, A fixed (v i ) is the attribute value after repair; Transmission path optimization, based on energy process chain diagram G process , find the optimal path to minimize power loss : ； Among them, P loss (e) is the power loss value on edge e; Node load balancing, redistributing power flow to nodes with transmission loads above a threshold: ； Among them, P surplus (v j ) is the neighbor node v j Can be reassigned to v i The remaining power P redistribute (v i ) is the node v i The power flow that can be regained through neighboring nodes.

Images