Power grid intelligent checking method and device based on graph neural network and storage medium

Abstract

The application relates to a power grid intelligent checking method and device based on a graph neural network and a storage medium, wherein the power grid intelligent checking method based on the graph neural network comprises the following steps: determining connection lines between different power sites based on a power grid topology structure, and determining a feature vector of each power site, and constructing a power grid topology graph taking the power sites as nodes and the connection lines as edges; inputting the power grid topology graph into a graph neural network, iteratively updating each feature vector through a message passing mechanism until a preset iteration number is met to obtain node embedding of each node, and determining a power grid topology relationship graph according to the node embedding; determining target nodes of each checking task in the power grid topology relationship graph, simultaneously executing all checking tasks according to node embedding of all target nodes to obtain checking results; and obtaining scores of each checking result based on a preset scoring rule of each checking task, and generating a checking report according to the scores. Through the application, the efficiency of power grid checking is improved.

Description

Technical Field

[0001] This application relates to the field of power grid verification, and in particular to a method, apparatus and storage medium for intelligent power grid verification based on graph neural networks. Background Technology

[0002] Power grid verification refers to the technical assessment of the safety, stability, and reliability of the power grid during the operation, planning, or equipment connection of the power system.

[0003] In existing technologies, power grid verification relies on manual verification or rule-engine-based automated systems. Manual verification is time-consuming and labor-intensive, easily affected by subjective factors, and suffers from verification errors and low accuracy. While rule-engine-based automated systems can achieve partial automation, they struggle to adapt to complex and ever-changing power grid structures, resulting in poor accuracy. Furthermore, existing power grid verification methods are designed for single verification tasks, requiring multiple verifications, leading to low efficiency. Therefore, improving the efficiency of power grid verification while ensuring accuracy has become a pressing issue. Summary of the Invention

[0004] This application provides a method, apparatus, and storage medium for intelligent power grid verification based on graph neural networks, which at least addresses the urgent problem in related technologies of how to improve the efficiency of power grid verification while ensuring its accuracy.

[0005] In a first aspect, embodiments of this application provide a smart grid verification method based on graph neural networks, including: Based on the power grid topology, the connecting lines between different power stations are determined, and the feature vector of each power station is determined based on the power grid operation data. A power grid topology graph is constructed with the power stations as nodes and the connecting lines as edges. The power grid topology graph is input into a graph neural network, and each feature vector is iteratively updated through a message passing mechanism until a preset number of iterations is met to obtain the node embedding of each node. The power grid topology graph is determined based on the node and the node embedding. For any power grid verification task, the node used for power grid verification is taken as the target node in the power grid topology diagram. All the verification tasks are executed simultaneously according to the node embedding of all the target nodes to obtain the verification result. Based on the preset scoring rules for each of the verification tasks, a score is obtained for each verification result, and a verification report is generated based on the score.

[0006] In one embodiment, the power station includes power plants and substations. The verification task for any power grid, in the power grid topology graph, uses the node used for power grid verification as the target node, including: If the verification tasks are a remaining interval verification task and a continuity verification task, the power station is taken as the target node in the power grid topology diagram. If the verification task is a capacity ratio verification task, the substations of the district / county level power grid area are taken as the target nodes in the power grid topology diagram.

[0007] In one embodiment, the step of simultaneously executing all the verification tasks based on the node embeddings of all the target nodes to obtain the verification result includes: If the verification task is the remaining interval verification task, based on the node embedding of the target node, the difference between the preset total number of intervals of the power station and the number of intervals currently occupied is calculated to obtain the remaining interval number. If the verification task is a continuation verification task, the wiring type of the power station is obtained based on the node embedding of the target node; If the verification task is a capacity-to-load ratio verification task, based on the node embedding of the target node, the capacity-to-load ratio of the district / county-level power grid area is obtained according to the ratio of the total transformer capacity of the power grid in the area to the annual load of the area.

[0008] In one embodiment, the preset scoring rule includes a first preset scoring rule for the remaining interval verification task, and the first preset scoring rule is specifically configured as follows: The score for the remaining interval verification task decreases sequentially in the following order: the remaining interval satisfies the target remaining interval, the remaining interval is equal to 0, and the remaining interval is greater than the target remaining interval.

[0009] In one embodiment, the preset scoring rule includes a second preset scoring rule for the continuation verification task, and the second preset scoring rule is specifically configured as follows: The scoring of the connection verification task decreases in the following order: single-chain connection, non-co-tower double-circuit double-radial connection, co-tower double-circuit double-radial connection, one-sided T connection, double T connection, and single-wire connection.

[0010] In one embodiment, the preset scoring rule includes a third preset scoring rule for the capacity ratio verification task, and the third preset scoring rule is specifically configured as follows: The score for the load capacity ratio verification task decreases sequentially in the following order: the load capacity ratio is within a preset range, the load capacity ratio is within a first offset range of the preset range, the load capacity ratio is within a second offset range of the preset range, the load capacity ratio is within a third offset range of the preset range, and the load capacity ratio is within a fourth offset range of the preset range.

[0011] In one embodiment, generating a verification report based on the score includes: If any node has at least two of the above scores, all the scores are weighted and fused to obtain the target score of the node; If any of the nodes has only one rating, that rating will be used as the target rating. The overall evaluation of the power grid and the generation of a verification report are determined based on the scores of all targets.

[0012] In one embodiment, the feature vector is specifically configured as follows: The feature vector includes load data, equipment parameters, and power grid operation data; The equipment parameters include the parameters of the transformer and the line parameters located within the power station; The power grid operation data includes the switching status and electrical measurement values ​​of the power equipment within the power station; The load data includes historical power consumption data and predicted power consumption data of the power station.

[0013] Secondly, embodiments of this application provide a smart grid verification device based on graph neural networks, comprising: The power grid topology construction module is used to determine the connection lines between different power stations based on the power grid topology, determine the feature vector of each power station based on power grid operation data, and construct a power grid topology with the power stations as nodes and the connection lines as edges. A power grid topology graph construction module is used to input the power grid topology graph into a graph neural network, iteratively update each feature vector through a message passing mechanism until a preset number of iterations is met to obtain the node embedding of each node, and determine the power grid topology graph based on the node and the node embedding. The verification module is used for any power grid verification task. In the power grid topology diagram, the node used for power grid verification is taken as the target node. Based on the node embedding of all the target nodes, all the verification tasks are executed simultaneously to obtain the verification result. The scoring module is used to obtain a score for each verification result based on a preset scoring rule for each verification task, and to generate a verification report based on the score.

[0014] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the power grid intelligent verification method based on graph neural networks as described in the first aspect above.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the power grid intelligent verification method based on graph neural networks as described in the first aspect above.

[0016] The power grid intelligent verification method, device, and storage medium based on graph neural networks provided in this application have at least the following technical effects: The power grid topology graph is determined using a graph neural network. Each node in the graph has a node embedding containing both electrical and topological semantics, providing accurate data for subsequent verification and improving the accuracy of the verification task. The target node for each verification task is identified, and all verification tasks are executed simultaneously based on the node embedding of the target node to obtain the verification results. Executing multiple verification tasks in parallel improves the efficiency of power grid verification. According to the preset scoring rules for each verification task, a score is determined for each verification result, and a verification report is generated based on the score. This verification report supports subsequent power grid decision-making and engineering applications.

[0017] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating a power grid intelligent verification method based on a graph neural network according to an exemplary embodiment; Figure 2 This is a schematic diagram illustrating a power grid topology according to an exemplary embodiment; Figure 3 This is a schematic diagram of a power grid intelligent verification device based on a graph neural network, according to an exemplary embodiment. Figure 4 This is a block diagram of an electronic device according to an exemplary embodiment. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0020] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0021] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0022] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0023] Firstly, embodiments of this application provide a smart grid verification method based on graph neural networks. Figure 1 This is a flowchart illustrating a power grid intelligent verification method based on a graph neural network according to an exemplary embodiment, such as... Figure 1 As shown, the smart grid verification method based on graph neural networks includes: Step S101: Determine the connection lines between different power stations based on the power grid topology, determine the feature vector of each power station based on the power grid operation data, and construct a power grid topology graph with power stations as nodes and connection lines as edges.

[0024] A power grid topology map describes the physical connections between equipment at different power stations, identifying the connecting lines between them. Power grid operation data includes load data, equipment parameters, and overall grid operation data for each power station. Equipment parameters include transformer and line parameters within the power station. Grid operation data includes the switching status and electrical measurements of electrical equipment within the power station. Load data includes historical and predicted power consumption data for each power station. The load data, equipment parameters, and grid operation data for each power station are used as its own feature vector. The power station is used as a node, and the connecting lines between different power stations are used as edges to construct the power grid topology map. The power grid topology map is graph data used for graph neural network processing; therefore, each node carries its own feature vector. The power grid topology structure is derived from a Geographic Information System (GIS), and the power grid operation data is derived from a Supervisory Control and Data Acquisition (SCADA) system.

[0025] It should be noted that before constructing the power grid topology map, the power grid topology, load data, equipment parameters, and power grid operation data all need to be preprocessed. Since these data originate from different systems and equipment, they require data cleaning and normalization. The preprocessed data will then be used to construct the power grid topology map. Data cleaning includes aligning equipment IDs across different systems, handling outlier measurements, and completing missing data. Data normalization involves scaling numerical characteristics of different dimensions to a uniform numerical range.

[0026] In a power grid topology graph, each node possesses a feature vector, which represents the initial parameters of the current node and is isolated attribute information. Edges between nodes represent the connections between different nodes. By analyzing the nodes and edges in the power grid topology graph, the feature information of the nodes and the connections between them are preliminarily determined, providing a foundation for subsequent graph neural network processing.

[0027] Step S102: Input the power grid topology diagram into the graph neural network, iteratively update each feature vector through the message passing mechanism until the preset number of iterations is met to obtain the node embedding of each node, and determine the power grid topology diagram based on the nodes and node embeddings.

[0028] Using the power grid topology as input to a graph neural network (Graph Neural Network), the Graph Neural Network employs a message-passing mechanism and comprises multiple layers, starting with an initial layer and ending with a final layer. For each node, the neural network extends from the starting point to the ending point. In the initial layer, direct neighbors are determined based on edges, and information is exchanged and aggregated with these direct neighbors to update the feature vector and obtain the first intermediate feature vector. Based on this intermediate feature vector, the node determines its first indirect neighbor in the next layer based on edges from its direct neighbors. The node then exchanges and aggregates information with these direct neighbors to update the first intermediate feature vector and obtain the second intermediate feature vector. This process is repeated layer by layer to determine the node's more distant indirect neighbors in the power grid topology, continuously updating the intermediate feature vectors until the node embedding is obtained in the final layer. Here, the number of layers from the initial layer to the final layer represents the number of iterations.

[0029] Through graph neural networks, nodes are transformed from isolated feature vectors into feature vectors carrying complex relationships within the power grid. These are high-dimensional feature vectors rich in electrical relationships and topological semantics. Furthermore, they enable a node to perceive the impact of the power grid structure on the node.

[0030] Continuing with step S102, determine the power grid topology diagram based on each node and node embedding. Figure 2 This is a schematic diagram illustrating a power grid topology according to an exemplary embodiment, such as... Figure 2 As shown, power stations, power plants, user stations, and substations are considered nodes, and lines 1 to 6 represent the direct and indirect connections between different nodes, respectively. The power grid topology diagram allows a single node to perceive multiple indirect nodes, enabling nodes to not only possess their own attribute information but also capture the electrical distances between nodes and power sources, between nodes and loads, and the criticality of nodes in power flow paths within the diagram.

[0031] By processing the power grid topology graph using the graph neural network in step S102, nodes acquire node embeddings rich in electrical and topological semantics. This allows nodes to perceive information from multiple indirect nodes, improving the accuracy of their electrical data determination under various information conditions. This provides an accurate data foundation for subsequent power grid verification. Furthermore, the graph neural network processes the power grid topology graph, which reflects the grid's connectivity and updates rapidly with grid updates. This allows it to adapt to the high-frequency updates during power grid verification, providing reliable system and data support for subsequent dynamic power grid verification.

[0032] Step S103: For any power grid verification task, the node used for power grid verification is taken as the target node in the power grid topology diagram. All verification tasks are executed simultaneously according to the node embedding of all target nodes to obtain the verification result.

[0033] Based on different verification tasks, the target node for each verification task is determined in the power grid topology diagram; that is, the electrical substations verified by each verification task will be different. The electrical information of the node is determined based on the node embedding carried by the target node. The verification task is then executed based on this electrical information to obtain the verification results. In the power grid topology diagram, the node embeddings do not interfere with each other; for a verification task, only the electrical information of the node embedding needs to be obtained for verification. Therefore, all verification tasks can be executed synchronously.

[0034] The verification tasks include remaining interval verification, continuity verification, and capacity ratio verification. Determining the target node for each verification task specifically includes: If the verification task is a remaining interval verification task, the power station is used as the target node in the power grid topology diagram. If the verification task is a continuity verification task, the power station is used as the target node in the power grid topology diagram. If the verification task is a capacity ratio verification task, the substation of the district / county-level power grid area is used as the target node in the power grid topology diagram.

[0035] Based on the target node determined for each verification task, the node embedding of the target node is obtained, and the verification task is executed according to the node embedding. The content executed varies for different verification tasks, and may specifically include: If the verification task is a remaining interval count verification task, the remaining interval count is obtained by calculating the difference between the preset total number of intervals of the power station and the number of currently occupied intervals based on the node embedding of the target node. If the verification task is a connection verification task, the connection type of the power station is obtained based on the node embedding of the target node. If the verification task is a capacity-to-load ratio verification task, the capacity-to-load ratio of the county-level power grid area is obtained based on the ratio of the total substation capacity of the power grid in the area to the annual load of the area, based on the node embedding of the target node. The total substation capacity is the sum of the substation capacities of all grid-connected substations (including public and dedicated substations) in the county-level power grid area, and the annual load of the area is the annual measured maximum load data. The steps for obtaining the annual load of the area specifically include: obtaining the load value of each substation in the county-level power grid area at preset time intervals; constructing a load curve for each substation with time as the horizontal axis and load value as the vertical axis, with the horizontal axis of the load curve corresponding to a one-year time period; and ensuring that the unit length of the horizontal axis is the same for each substation's load curve. The load values ​​on the same horizontal axis of each load curve are summed to determine the regional curve for the county-level power grid area. The largest load value in the regional curve is taken as the annual load of the region.

[0036] Continuing with step S103, the target node for each verification task is determined. The node embedding based on the target node synchronously executes the verification task, which can effectively improve the speed of power grid verification. Furthermore, the data for executing the verification task comes from the power grid topology diagram and has accurate data. Therefore, synchronously executing the verification task can ensure the accuracy of power grid verification on the one hand, and improve the verification speed on the other hand, thus achieving the efficiency of power grid verification.

[0037] Step S104: Based on the preset scoring rules for each verification task, obtain the score for each verification result and generate a verification report based on the score.

[0038] Each verification task has a corresponding preset scoring rule, and the score for each scoring result is determined according to the respective preset scoring rule. Among them, the preset scoring rules include the first preset scoring rule for the remaining interval verification task, the second preset scoring rule for the subsequent verification task, and the third preset scoring rule for the capacity ratio verification task.

[0039] The first preset scoring rule for the remaining interval verification task specifically satisfies the following: the score for the remaining interval verification task decreases sequentially according to the remaining interval number meeting the target remaining interval number, the remaining interval number being equal to 0, and the remaining interval number being greater than the target remaining interval number. Here, the target remaining interval number is the number of remaining intervals that meet the reserved requirements of the power station, and the target remaining interval number is non-negative. Optionally, the target remaining interval number may include 1 or 2.

[0040] In one embodiment, if the number of remaining intervals meets the target number of remaining intervals, the score is 2 points. If the number of remaining intervals is 0, the score is 0 points. If the number of remaining intervals is greater than the target number of remaining intervals, the score is -1 points.

[0041] The second preset scoring rule for the continuity verification task specifically satisfies the following: the scoring of the continuity verification task decreases in the following order: single-chain connection, non-same-tower double-circuit double-radial connection, same-tower double-circuit double-radial connection, one-sided T connection, double T connection, and single-wire connection.

[0042] In one embodiment, if the wiring type is a single-chain connection, the score is 2 points; if the wiring type is a non-co-tower double-circuit double-radial connection, the score is 1 point; if the wiring type is a co-tower double-circuit double-radial connection, the score is 0 points; if the wiring type is a one-sided T connection, the score is -1 point; if the wiring type is a double T connection, the score is -2 points; and if the wiring type is a single-wire connection, the score is -3 points.

[0043] Among them, single-chain connections are characterized by simple wiring, convenient operation and maintenance, and high reliability. Non-co-tower double-circuit double-radial connections are characterized by good economy and wide power supply coverage. Co-tower double-circuit double-radial connections are characterized by compact structure and saving corridor resources, but require strengthened operation and maintenance safety management. One-sided T-connections are characterized by moderate reliability and require comprehensive protection measures. Double-T connections are characterized by low reliability and require control of the power supply radius. Single-line connections are characterized by simple structure and low investment.

[0044] The third preset scoring rule for the load capacity ratio verification task specifically satisfies the following: the score for the load capacity ratio verification task decreases sequentially in the following order: the load capacity ratio is within a preset range, the load capacity ratio is within the first offset range of the preset range, the load capacity ratio is within the second offset range of the preset range, the load capacity ratio is within the third offset range of the preset range, and the load capacity ratio is within the fourth offset range of the preset range.

[0045] In one embodiment, if the load capacity ratio is within a preset range, the score is 3 points. If the load capacity ratio is within a first offset range of the preset range, the score is 2 points. If the load capacity ratio is within a second offset range of the preset range, the score is 1 point. If the load capacity ratio is within a third offset range of the preset range, the score is 0 points. If the load capacity ratio is within a fourth offset range of the preset range, the score is -2 points.

[0046] The third preset scoring rule is also affected by the annual average load increase rate. When the annual average load increase rate changes, the preset range will also change accordingly. When the annual average load increase rate is configured to be 5.5%, the preset range is [1.7, 1.9]; when the annual average load increase rate is less than or equal to 2%, the preset range is [1.5, 1.7]; when the annual average load increase rate is less than or equal to 4% and greater than 2%, the preset range is [1.6, 1.8]; when the annual average load increase rate is less than or equal to 7% and greater than 4%, the preset range is [1.7, 1.9]; when the annual average load increase rate is greater than 7%, the preset range is [1.8, 2.0].

[0047] Under the same average annual load increase rate, the preset range remains the same for different voltage levels. If the voltage levels are 220kV, 110kV, and 35kV, and the load growth rate is 5.5%, their preset ranges are all [1.7, 1.9].

[0048] After determining the preset range, the first offset range, second offset range, and third offset range are different offset regions outside the preset range. Furthermore, the first offset range, second offset range, third offset range, and fourth offset range expand sequentially. For example, the first offset range is the interval between ±0.1 of the preset range, including the ±0.1 offset value. The second offset range is the interval between +0.1 and +0.2 of the preset range, and the interval between -0.1 and -0.2 of the preset range, including the ±0.2 offset value. The third offset range is the interval between +0.2 and +0.5 of the preset range, and the interval between -0.2 and -0.5 of the preset range, including the ±0.5 offset value. The fourth offset range is the interval between offsets greater than +0.5 of the preset range and the interval between offsets less than -0.5 of the preset range.

[0049] Based on preset scoring rules for different verification tasks, a score is obtained for each verification result. The score can provide a preliminary reflection of the operating status of different power stations and make a preliminary assessment of the operating status of power stations.

[0050] Continuing with step S104, a verification report is generated based on the scoring, and response strategies for each power station are formulated based on the verification report. The generation of the verification report specifically includes: If any node has at least two scores, all scores are weighted and merged to obtain the node's target score. By weighted merging of all scores for the same node, the overall status of that node can be assessed more accurately. If any node has only one score, that score is used as the target score. Based on all target scores, the overall evaluation of the power grid is determined and a verification report is generated.

[0051] The verification report includes a score based on different verification results to determine the problem type and severity statistics. Each problem is also presented as an item, including its location, details, violated rules, and recommended actions.

[0052] In addition, based on the issues identified in the verification report, problem areas are highlighted in the power grid topology diagram to provide visualization results. Specifically, the highlighted problem areas can be nodes or edges with problems marked in different colors.

[0053] In one embodiment, if the remaining interval count of the remaining interval count verification task is non-negative and meets the target remaining interval count, the verification report will show that the corresponding node has passed the remaining interval count verification task. If the remaining interval count is negative or does not meet the target remaining interval count, the verification report will show processing suggestions, namely, checking the completeness of the power plant interval count statistics or the rationality of the total power plant interval configuration.

[0054] In another embodiment, during the capacity-to-load ratio verification task, if the capacity-to-load ratio is lower than the lower limit of a preset range, the verification report displays processing suggestions, namely, assessing the power supply gap and determining a solution for adding substation capacity or a load transfer solution. If the capacity-to-load ratio is higher than the upper limit of the preset range, the verification report displays processing suggestions, namely, analyzing the idle substation capacity and optimizing the power grid layout or adjusting the power supply range. It should be noted that for special exemption areas, such as areas rich in new energy sources and areas with rapid load growth (areas with an annual load growth rate exceeding 10%), the preset range can be adjusted according to the actual situation.

[0055] In summary, the power grid intelligent verification based on graph neural networks provided in this application uses graph neural networks to process the power grid topology, enabling nodes to perceive information from multiple indirect nodes. This allows nodes to improve the accuracy of their electrical data determination under various information conditions, thus providing an accurate data foundation for subsequent power grid verification. The power grid topology reflects the grid's connectivity; if the grid is updated, the topology can be quickly updated to match the actual grid. Therefore, by adapting to high-frequency grid updates, the power grid verification can respond to grid updates, obtain the latest data, and perform dynamic verification, exhibiting scalability. Multiple verification tasks can be executed simultaneously, improving the verification rate and ensuring accurate verification based on node embedding, thereby increasing the efficiency of power grid verification. Visualized results and verification reports provide support for subsequent power grid decision-making and engineering applications.

[0056] Secondly, embodiments of this application provide a smart grid verification device based on graph neural networks. Figure 3 This is a schematic diagram of a power grid intelligent verification device based on a graph neural network, according to an exemplary embodiment. Figure 3 As shown, the intelligent power grid verification device based on graph neural networks includes: The power grid topology construction module is used to determine the connection lines between different power stations based on the power grid topology, determine the feature vector of each power station based on power grid operation data, and construct a power grid topology with power stations as nodes and connection lines as edges. The power grid topology graph construction module is used to input the power grid topology graph into the graph neural network, iteratively update each feature vector through a message passing mechanism until the preset number of iterations is met to obtain the node embedding of each node, and determine the power grid topology graph based on the nodes and node embeddings. The verification module is used to determine the target node for power grid verification in each verification task in the power grid topology diagram, and to execute all verification tasks simultaneously based on the node embedding of all target nodes to obtain the verification results. The scoring module is used to obtain a score for each verification result based on the preset scoring rules for each verification task, and generate a verification report based on the score.

[0057] In summary, this application provides a power grid intelligent verification device based on graph neural networks. The device determines the power grid topology graph using graph neural networks, and each node in the topology graph has a node embedding containing both electrical and topological semantics, providing accurate data for subsequent verification and improving the accuracy of verification tasks. The target node for each verification task is determined, and all verification tasks are executed simultaneously based on the node embedding of the target node to obtain verification results. The parallel execution of multiple verification tasks improves the efficiency of power grid verification. According to the preset scoring rules for each verification task, a score is determined for each verification result, and a verification report is generated based on the score. This verification report can support subsequent power grid decision-making and engineering applications.

[0058] It should be noted that the power grid intelligent verification device based on graph neural networks provided in this embodiment is used to implement the above-described embodiments, and details already described will not be repeated. As used above, the terms "module," "unit," and "subunit" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the above embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0059] Thirdly, embodiments of this application provide an electronic device, Figure 4 This is a block diagram illustrating an electronic device according to an exemplary embodiment. (e.g.) Figure 4 As shown, the electronic device may include a processor 81 and a memory 82 storing computer program instructions.

[0060] Specifically, the processor 81 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0061] The memory 82 may include a mass storage device for data or instructions. For example, and not limitingly, the memory 82 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 82 may include removable or non-removable (or fixed) media. Where appropriate, the memory 82 may be internal or external to a data processing device. In a particular embodiment, the memory 82 is non-volatile memory. In a particular embodiment, the memory 82 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random-Access Memory (FPMDRAM), Extended Data Out Dynamic Random-Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0062] The memory 82 can be used to store or cache various data files that need to be processed and / or communicated, as well as possible computer program instructions executed by the processor 81.

[0063] The processor 81 reads and executes computer program instructions stored in the memory 82 to implement any of the graph neural network-based smart grid verification methods in the above embodiments.

[0064] In one embodiment, the smart grid verification device based on graph neural networks may further include a communication interface 83 and a bus 80. Wherein, as... Figure 4 As shown, the processor 81, memory 82, and communication interface 83 are connected through bus 80 and complete communication with each other.

[0065] The communication interface 83 is used to enable communication between the various modules, devices, units, and / or equipment in the embodiments of this application. The communication interface 83 can also enable data communication with other components such as external devices, image / data acquisition devices, databases, external storage, and image / data processing workstations.

[0066] Bus 80 includes hardware, software, or both, that couples together components of a graph neural network-based smart grid verification device. Bus 80 includes, but is not limited to, at least one of the following: data bus, address bus, control bus, expansion bus, and local bus. For example, and not as a limitation, bus 80 may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 80 may include one or more buses. Although specific buses are described and illustrated in the embodiments of this application, this application considers any suitable bus or interconnection.

[0067] Fourthly, embodiments of this application provide a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the power grid intelligent verification method based on graph neural networks provided in the first aspect.

[0068] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0069] In a possible implementation, the present invention can also be implemented as a program product comprising program code, which, when the program product is run on a terminal device, is used to cause the terminal device to perform the steps of implementing the power grid intelligent verification method based on graph neural networks provided in the first aspect.

[0070] The program code for executing the present invention can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0071] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0072] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A smart power grid verification method based on graph neural networks, characterized in that, include: Based on the power grid topology, the connecting lines between different power stations are determined, and the feature vector of each power station is determined based on the power grid operation data. A power grid topology graph is constructed with the power stations as nodes and the connecting lines as edges. The power grid topology graph is input into a graph neural network, and each feature vector is iteratively updated through a message passing mechanism until a preset number of iterations is met to obtain the node embedding of each node. The power grid topology graph is determined based on the node and the node embedding. For any power grid verification task, the node used for power grid verification is taken as the target node in the power grid topology diagram. All the verification tasks are executed simultaneously according to the node embedding of all the target nodes to obtain the verification result. Based on the preset scoring rules for each of the verification tasks, a score is obtained for each verification result, and a verification report is generated based on the score.

2. The intelligent power grid verification method based on graph neural networks according to claim 1, characterized in that, The power stations include power plants and substations. The verification task for any power grid, in the power grid topology diagram, uses the nodes used for power grid verification as target nodes, including: If the verification tasks are a remaining interval verification task and a continuity verification task, the power station is taken as the target node in the power grid topology diagram. If the verification task is a capacity ratio verification task, the substation in the county-level power grid area is taken as the target node in the power grid topology diagram.

3. The intelligent power grid verification method based on graph neural networks according to claim 2, characterized in that, The step of simultaneously executing all the verification tasks based on the node embeddings of all the target nodes to obtain the verification result includes: If the verification task is the remaining interval verification task, based on the node embedding of the target node, the difference between the preset total number of intervals of the power station and the number of intervals currently occupied is calculated to obtain the remaining interval number. If the verification task is a continuation verification task, the wiring type of the power station is obtained based on the node embedding of the target node; If the verification task is a capacity-to-load ratio verification task, based on the node embedding of the target node, the capacity-to-load ratio of the district / county-level power grid area is obtained according to the ratio of the total transformer capacity of the power grid in the area to the annual load of the area.

4. The intelligent power grid verification method based on graph neural networks according to claim 3, characterized in that, The preset scoring rules include a first preset scoring rule for the remaining interval verification task, and the first preset scoring rule is specifically configured as follows: The score for the remaining interval verification task decreases sequentially in the following order: the remaining interval satisfies the target remaining interval, the remaining interval is equal to 0, and the remaining interval is greater than the target remaining interval.

5. The intelligent power grid verification method based on graph neural networks according to claim 3, characterized in that, The preset scoring rules include a second preset scoring rule for the follow-up verification task, and the second preset scoring rule is specifically configured as follows: The scoring of the connection verification task decreases in the following order: single-chain connection, non-co-tower double-circuit double-radial connection, co-tower double-circuit double-radial connection, one-sided T connection, double T connection, and single-wire connection.

6. The intelligent power grid verification method based on graph neural networks according to claim 3, characterized in that, The preset scoring rules include a third preset scoring rule for the capacity ratio verification task, and the third preset scoring rule is specifically configured as follows: The score for the load capacity ratio verification task decreases sequentially in the following order: the load capacity ratio is within a preset range, the load capacity ratio is within a first offset range of the preset range, the load capacity ratio is within a second offset range of the preset range, the load capacity ratio is within a third offset range of the preset range, and the load capacity ratio is within a fourth offset range of the preset range.

7. The intelligent power grid verification method based on graph neural networks according to claim 1, characterized in that, The step of generating a verification report based on the score includes: If any node has at least two of the above scores, all the scores are weighted and fused to obtain the target score of the node; If any of the nodes has only one rating, that rating will be used as the target rating. The overall evaluation of the power grid and the generation of a verification report are determined based on the scores of all targets.

8. The intelligent power grid verification method based on graph neural networks according to claim 1, characterized in that, The feature vector is specifically configured as follows: The feature vector includes load data, equipment parameters, and power grid operation data; The equipment parameters include the parameters of the transformer and the line parameters located within the power station; The power grid operation data includes the switching status and electrical measurement values ​​of the power equipment within the power station; The load data includes historical power consumption data and predicted power consumption data of the power station.

9. A smart power grid verification device based on graph neural networks, characterized in that, include: The power grid topology construction module is used to determine the connection lines between different power stations based on the power grid topology, determine the feature vector of each power station based on power grid operation data, and construct a power grid topology with the power stations as nodes and the connection lines as edges. A power grid topology graph construction module is used to input the power grid topology graph into a graph neural network, iteratively update each feature vector through a message passing mechanism until a preset number of iterations is met to obtain the node embedding of each node, and determine the power grid topology graph based on the node and the node embedding. The verification module is used for any power grid verification task. In the power grid topology diagram, the node used for power grid verification is taken as the target node. Based on the node embedding of all the target nodes, all the verification tasks are executed simultaneously to obtain the verification result. The scoring module is used to obtain a score for each verification result based on a preset scoring rule for each verification task, and to generate a verification report based on the score.

10. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the power grid intelligent verification method based on a graph neural network as described in any one of claims 1 to 8.

Images