A method, device, terminal and medium for identifying key nodes in a distribution network

Abstract

The present application discloses a method, device, terminal and medium for identifying key nodes in a distribution network, and relates to the technical field of distribution networks. The solution provided by the present application is based on the node topology data and historical operation data of the distribution network. First, different node relationship graph data are constructed according to the node topology data. Then, the spatial features obtained based on the node relationship graph data and the historical operation data representing the time series features are used to obtain a time-space feature vector through a time series aggregation model. Then, an arbitrary target node is selected, and the time-space feature vectors of the target node and related neighboring nodes are shielded. A simulation operation is performed through a multi-task prediction model to obtain the loss error before and after the target node is shielded. The difference between the two groups of loss errors is used to measure the impact of the failure of the target node on the overall distribution network, thereby identifying the key vulnerable nodes in the distribution network, and solving the technical problem of low accuracy in the existing key node identification.

Description

Technical Field

[0001] The present application relates to the technical field of distribution network, and in particular to a method, device, terminal and medium for identifying key nodes in a distribution network. Background Art

[0002] With the expansion of power systems and the frequent occurrence of extreme weather events, assessing the resilience of power systems and identifying key nodes have become important issues for ensuring power supply security. Traditional methods rely primarily on physical model simulation and complex network theory, but these methods struggle to cope with complex spatiotemporal coupled dynamic scenarios. Mainstream methods in the field of key node identification include complex network theory, machine learning models, and interpretability methods. In recent years, technologies based on graph neural networks (GNNs) and spatiotemporal modeling have been gradually applied to power system analysis. However, existing methods suffer from a single dimension, poor multi-task coordination, and insufficient interpretability in key node identification, resulting in low accuracy in key node identification in practical applications. Summary of the Invention

[0003] The present application provides a method, device, terminal and medium for identifying key nodes in a distribution network, which are used to solve the technical problem of low accuracy in existing key node identification.

[0004] To solve the above technical problems, the first aspect of the present application provides a method for identifying key nodes in a distribution network, comprising:

[0005] Obtain node topology data and historical operation data of the distribution network;

[0006] According to the node topology data, different node relationship graph data are constructed through a plurality of preset graph data construction conditions;

[0007] The node relationship graph data are integrated to obtain the spatial characteristics of the distribution network, and then the spatiotemporal characteristic vector of the distribution network is obtained based on the historical operation data and the spatial characteristics through a preset time series aggregation model;

[0008] Based on the spatiotemporal feature vectors corresponding to each node in the distribution network, a simulation operation is performed in combination with a multi-task prediction model to obtain a multi-task loss error benchmark value;

[0009] Determine any target node from the distribution network, combine the neighboring nodes corresponding to the target node, shield the spatiotemporal feature vectors of the target node and the neighboring nodes, and then perform simulation calculations through a multi-task prediction model to obtain a multi-task loss error check value;

[0010] A key node identification result of the target node is determined according to an error difference between the multi-task loss error reference value and the multi-task loss error check value.

[0011] Preferably, the node relationship graph data includes: topological adjacency graph data, geographical distance graph data and node type similarity graph data.

[0012] Preferably, the fusing of the node relationship graph data to obtain the spatial characteristics of the distribution network specifically includes:

[0013] The spatial embedding features in each node relationship graph data are extracted through a graph neural network, and then the spatial embedding features of each node relationship graph data are fused to obtain the spatial features of the distribution network.

[0014] Preferably, the multi-task loss error benchmark value is obtained by performing simulation calculations based on the spatiotemporal feature vectors corresponding to each node in the distribution network in combination with a multi-task prediction model, specifically comprising:

[0015] Based on the spatiotemporal feature vectors corresponding to each node in the distribution network, a simulation operation is performed in combination with a multi-task prediction model to obtain multiple single-task prediction data;

[0016] A multi-task loss error benchmark value is obtained based on the cumulative sum of the loss errors of each single-task prediction data and the corresponding single-task real data.

[0017] Preferably, the single-task prediction data includes at least two of: load gap prediction data, failure probability prediction data, recovery time prediction data, transmission flow prediction data and charging demand prediction data.

[0018] Preferably, determining any target node from the distribution network, combining the neighboring nodes corresponding to the target node, shielding the spatiotemporal feature vectors of the target node and the neighboring nodes, and then performing simulation calculations through a multi-task prediction model to obtain a multi-task loss error check value specifically includes:

[0019] Determine any target node from the distribution network, combine the neighboring nodes corresponding to the target node, and shield the spatiotemporal feature vectors of the target node and the neighboring nodes according to Extra shielding logic and / or Intra shielding logic.

[0020] Based on the spatiotemporal characteristic vector of the distribution network after node shielding processing, a simulation operation is performed through a multi-task prediction model to obtain a first multi-task loss error check value and / or a second multi-task loss error check value, wherein the first multi-task loss error check value is a multi-task loss error check value obtained based on the spatiotemporal characteristic vector of the distribution network processed according to the Extra shielding logic, and the second multi-task loss error check value is a multi-task loss error check value obtained based on the spatiotemporal characteristic vector of the distribution network processed according to the Intra shielding logic.

[0021] Preferably, determining the key node identification result of the target node according to the error difference between the multi-task loss error baseline value and the multi-task loss error check value specifically includes:

[0022] According to the error difference between the multi-task loss error reference value and the multi-task loss error check value, by comparing the error differences corresponding to each target node, a key node identification result of the target node is determined.

[0023] At the same time, the second aspect of the present application provides a distribution network key node identification device, including:

[0024] A distribution network data acquisition unit, used to acquire node topology data and historical operation data of the distribution network;

[0025] A graph data construction unit, configured to construct different node relationship graph data according to the node topology data and through a plurality of preset graph data construction conditions;

[0026] a spatiotemporal feature extraction unit, configured to fuse the node relationship graph data to obtain the spatial features of the distribution network, and then obtain the spatiotemporal feature vector of the distribution network based on the historical operation data and the spatial features through a preset time series aggregation model;

[0027] The first error calculation unit is used to perform simulation calculations based on the spatiotemporal feature vectors corresponding to each node in the distribution network in combination with the multi-task prediction model to obtain a multi-task loss error benchmark value;

[0028] A second error calculation unit is configured to determine an arbitrary target node from the distribution network, combine the neighboring nodes corresponding to the target node, shield the spatiotemporal feature vectors of the target node and the neighboring nodes, and then perform simulation calculations through a multi-task prediction model to obtain a multi-task loss error check value;

[0029] The key node identification unit is used to determine the key node identification result of the target node according to the error difference between the multi-task loss error reference value and the multi-task loss error check value.

[0030] A third aspect of the present application provides a distribution network key node identification terminal, comprising: a memory and a processor;

[0031] The memory is used to store program code, and the program code is used to implement a method for identifying key nodes in a distribution network as provided in the first aspect of the present application;

[0032] The processor is configured to read and execute the program code.

[0033] The fourth aspect of the present application provides a computer-readable storage medium, in which program code is stored. The program code is used to be read and executed by a processor to implement a distribution network key node identification method as provided in the first aspect of the present application.

[0034] It can be seen from the above technical solutions that this application has the following advantages:

[0035] The solution provided in this application is based on the node topology data and historical operation data of the distribution network. First, different node relationship graph data are constructed according to the node topology data, and then the spatial characteristics of the distribution network are obtained by fusing the node relationship graph data. The spatial characteristics and the historical operation data representing the time series characteristics are used to perform calculations through a preset time series aggregation model to obtain a spatiotemporal feature vector. Then, an arbitrary target node is determined from the distribution network to shield the spatiotemporal feature vector of the target node and related neighboring nodes. The multi-task prediction model is used to perform simulation operations to obtain the loss errors before and after the target node is shielded. The difference between the two groups of loss errors is used to measure the impact of the failure of the target node on the overall distribution network, and then the key nodes in the distribution network are identified. This solution can comprehensively capture the spatiotemporal dynamic characteristics of the power system through multi-dimensional modeling, task collaborative training, efficient computing and transparent decision support, improve the accuracy of node resilience prediction, and solve the technical problem of low accuracy in the existing key node identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0037] Figure 1 A flowchart of an embodiment of a method for identifying key nodes in a distribution network provided in this application.

[0038] Figure 2 This is a schematic diagram of the multi-graph structure and its fusion in the spatiotemporal model in a distribution network key node identification method provided in this application.

[0039] Figure 3 Schematic diagram of Extra and Intra shielding in interpretable analysis.

[0040] Figure 4 This is a structural diagram of an embodiment of a device for identifying key nodes in a distribution network provided in this application.

[0041] Figure 5This is a structural diagram of an embodiment of a distribution network key node identification terminal embodiment provided by this application. DETAILED DESCRIPTION

[0042] The embodiments of the present application provide a method, device, terminal and medium for identifying key nodes in a distribution network, which are used to solve the technical problem of low accuracy in existing key node identification.

[0043] In order to make the purpose, features, and advantages of the invention of this application more obvious and easy to understand, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the embodiments described below are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0044] First, a detailed description of an embodiment of a method for identifying key nodes in a distribution network provided by this application is as follows:

[0045] See also Figure 1 and Figure 2 , an embodiment of the present application provides a method for identifying key nodes in a distribution network, comprising:

[0046] Step 101: Obtain node topology data and historical operation data of the distribution network;

[0047] It should be noted that, first, according to the distribution network structure to be analyzed, the node topology data and historical operation data of the distribution network system are obtained.

[0048] Step 102: construct different node relationship graph data based on the node topology data and through a plurality of preset graph data construction conditions;

[0049] It should be noted that in step 102, based on the node topology data obtained in the previous step, different node relationship graph data are constructed. For example, assuming that the power system contains M nodes (such as substations, switch stations, distribution substations, important load centers, etc.). This embodiment presets multiple graph data construction conditions based on different perspectives to construct K pieces of node relationship graph data. , each picture Node Set and the corresponding adjacency matrix The adjacency matrix of each graph data represents the different association relationships between the nodes in the distribution network.

[0050] More specifically, the node relationship graph data mentioned in this embodiment mainly includes three types: topological adjacency graph data, geographical distance graph data, and node type similarity graph data;

[0051] Among them, the adjacency matrix of the topological adjacency graph is It is specifically used to reflect the topological connection relationship between each node. For example, if node i and node j are directly connected in a network topology, then ; otherwise 0;

[0052] Adjacency Matrix of Geographic Distance Graph Specifically used to reflect the geographical distance between nodes, for example, based on the geographical distance Normalize and get , can be used or etc.

[0053] Adjacency Matrix of Node Type Similarity Graph It is used to reflect the similarity of functional attributes or node types between nodes. If nodes i and j belong to the same type of nodes or nodes with similar functions, then , otherwise 0.

[0054] In addition to the three main graph data items mentioned above, more graphs can be expanded according to the needs of actual scenarios, such as "historical fault correlation graph" and "weather zone graph", so as to reflect node associations in more dimensions. I will not go into details here.

[0055] Step 103: The data of the node relationship graph are integrated to obtain the spatial characteristics of the distribution network. Then, based on the historical operation data and the spatial characteristics, a preset time series aggregation model is used to obtain the spatiotemporal characteristic vector of the distribution network.

[0056] It should be noted that when performing graph fusion, the GraphSAGE module is mainly used to extract the spatial feature representation of each node under each graph through the graph neural network. For each graph, GraphSAGE calculates the embedded representation of each node by aggregating the neighborhood of each node. Assume that at a certain layer, the representation of node i is , then the update process of this layer is:

[0057]

[0058] Where, is the learnable parameter matrix, is the embedding representation of node i at layer l, is a nonlinear activation function (such as ReLU), is the neighbor set of node i in a graph, is an aggregation function, such as mean, max-pool, or LSTM aggregation.

[0059] The input information of the aggregation function is the feature matrix and the adjacency matrix In the AGG function, the features of the neighboring nodes of node i are combined together by aggregation. For node i, it aggregates its neighboring nodes In the first layer of the GraphSAGE module, the feature representation of the node is extracted from the node feature matrix X_(i,t). In other words, the initial feature representation is obtained directly from the input matrix.

[0060] If there are K graphs, we can perform several layers of GraphSAGE operations on each graph to get the representation of node i in each graph. , and then concatenate them into:

[0061]

[0062] Among them, square brackets and “||” represent feature splicing operations.

[0063] Step 104: Based on the spatiotemporal feature vectors corresponding to each node in the distribution network, a simulation operation is performed in combination with the multi-task prediction model to obtain a multi-task loss error benchmark value;

[0064] It should be noted that the next step is multi-task output and loss processing. Assuming there are N tasks, the prediction target of each task n is recorded as At the last layer of the network, a small network is designed for each task:

[0065]

[0066] in, represents the predicted value of the nth task, represents the exclusive output structure of task 𝑛, is the learnable parameter of the network, let the true label be , the multi-task loss can be defined (using mean square error MSE), and the expression is as follows:

[0067]

[0068] in, Represents all parameters of the shared layer (spatial + temporal module); Indicates the parameters of the nth task header; represents the set of time indexes covered by the training set. During backpropagation, the gradients of multiple tasks jointly update the shared layer and their respective dedicated layers, thereby enabling the simultaneous prediction of multiple resilience-related indicators.

[0069] The model then outputs a predicted value for each task and an error value The specific error value type depends on the input information of the model. If the input is unshielded information, the initial error will be obtained. , which is the multi-task loss error benchmark mentioned in this application, which represents the total error of multi-task prediction when the node is not blocked. Assume Taking three tasks as an example (A: load gap, B: recovery time, C: fault range), the expression is as follows:

[0070]

[0071] Among them, load gap: represents the gap between the electricity demand and supply of node i at time t in an extreme scenario;

[0072] Recovery time: refers to the time required from power outage to restoration of normal power supply after a node or line failure;

[0073] Fault impact range: Indicates the potential impact of a node failure on surrounding nodes or the entire system, or the depth of failure propagation.

[0074] Step 105: Determine any target node from the distribution network, combine the neighboring nodes corresponding to the target node, shield the spatiotemporal feature vectors of the target node and the neighboring nodes, and then perform simulation calculations through the multi-task prediction model to obtain a multi-task loss error check value;

[0075] Step 106: Determine the key node identification result of the target node according to the error difference between the multi-task loss error baseline value and the multi-task loss error check value.

[0076] It should be noted that in steps 105 and 106, the interpretable analysis framework "Mask-Compute-Analyze" is used to identify key nodes and their importance to the overall or local prediction. After training is complete, further exploration of the importance of a node (or neighborhood) in "resilience prediction" is carried out using the following three steps:

[0077] Mask: Select the target node and its k-hop neighborhood, setting the input data of these nodes to zero to simulate a serious failure or data loss at the node;

[0078] Compute: Re-infer the model based on the masked input to obtain new multi-task prediction results and then calculate the error;

[0079] Specifically, for the node There are two shielding strategies, including Extra shielding (shielding nodes Self + Neighborhood) and Intra Shield (shield only neighbors, excluding ), the schematic diagram is as follows Figure 3 As shown, Figure 3 Part (a) is the node Schematic diagram of the shielding effect of the area and its neighborhood under the Extra shielding strategy. Figure 3 Part (b) is the node Schematic diagram of the shielding effect of the area and its neighborhood under the Intra shielding strategy.

[0080] Among them, the Extra shielding logic mainly analyzes the impact on the entire network. Define nodes The k-hop neighborhood of (including itself) is From the small-world network theory, we know that the influence range of a node is mainly concentrated within its 3-hop neighborhood. The contribution of nodes outside this range to the target node decays exponentially. Assuming k=3, the error output by the multi-task spatiotemporal learning model is:

[0081]

[0082] in, is the target node and the total error after 3 hops of neighborhood, i.e., the first multi-task loss error check value mentioned in this application, is a node The set of 3-hop neighboring nodes.

[0083] Intra shielding mainly analyzes the impact on the node itself. This analysis only shields the node The neighborhood of , but does not shield itself. At this time, the error output by the multi-task spatiotemporal learning model is:

[0084]

[0085] in, is the target node and the total error after 3 hops of neighborhood, i.e., the second multi-task loss error check value mentioned in this application, Indicates only shielding 3-hop neighborhood (retain itself).

[0086] It is understandable that in actual node analysis applications, any node shielding strategy can be selected according to actual needs, or two shielding strategies can be used in combination. For ease of explanation, this embodiment only uses the case of using two strategies in combination as an example.

[0087] Analyze: Compare the error changes before and after zeroing to measure the marginal contribution or influence of this node / neighborhood on the prediction result.

[0088] It should be noted that when performing the analyze analysis, the comparison error changes are all obtained by subtracting the initial error from the error after shielding. Specifically, they are:

[0089]

[0090]

[0091] in, is a node The marginal contribution to the prediction error of the entire network. The larger the positive value, the more significant the impact of the node failure on the overall resilience of the system. is a node The dependence on its own prediction error, the larger the positive value is, the more dependent the node resilience is on the surrounding area.

[0092] Based on Extra shield analysis, select Several nodes with the largest values (such as the top 10, which can be selected based on the actual node scale) should be prioritized for emergency power supply configuration. Failure of these nodes will significantly increase the total loss of multiple tasks.

[0093] Based on Intra shielding analysis, if ( is a preset threshold value, which can be selected based on actual conditions), indicating that the node is susceptible to neighboring failures and that local topological connections need to be strengthened;

[0094] Finally, the multi-task analysis result information is integrated and the results calculated by each node are compared. and , if the node Calculated in multiple tasks and The larger the value is, the greater the probability that the node is a key vulnerability point at the system level.

[0095] The solution provided in this application can more accurately find the most critical nodes in the distribution network, that is, identify those nodes that will have a major impact on the entire power grid if problems occur, and evaluate the resilience of the power grid, that is, evaluate the recovery efficiency of the power grid when encountering faults, and the places where problems are most likely to occur. It can help power operation and maintenance departments quickly identify key nodes and give priority to emergency repairs when disasters occur, reducing the duration and scope of power outages; guide reinforcement and expansion in power grid planning to improve overall resilience; monitor the status of key nodes in real time during daily operation and maintenance, and warn of potential risks in advance to avoid large-scale power outages.

[0096] The above is a detailed description of an embodiment of a method for identifying key nodes in a distribution network provided by this application. The following is a detailed description of an apparatus for identifying key nodes in a distribution network provided by this application, which is as follows:

[0097] See also Figure 4 , an embodiment of the present application provides a device for identifying key nodes in a distribution network, comprising:

[0098] The distribution network data acquisition unit 201 is used to acquire node topology data and historical operation data of the distribution network;

[0099] A graph data construction unit 202 is configured to construct different node relationship graph data according to the node topology data and through a plurality of preset graph data construction conditions;

[0100] The spatiotemporal feature extraction unit 203 is used to fuse the data of each node relationship graph to obtain the spatial characteristics of the distribution network, and then obtain the spatiotemporal feature vector of the distribution network based on the historical operation data and the spatial characteristics through a preset time series aggregation model;

[0101] The first error calculation unit 204 is used to perform simulation calculations based on the spatiotemporal feature vectors corresponding to each node in the distribution network in combination with the multi-task prediction model to obtain a multi-task loss error benchmark value;

[0102] The second error calculation unit 205 is used to determine any target node from the distribution network, combine the neighboring nodes corresponding to the target node, shield the spatiotemporal feature vectors of the target node and the neighboring nodes, and then perform simulation calculations through the multi-task prediction model to obtain a multi-task loss error check value;

[0103] The key node identification unit 206 is configured to determine a key node identification result of the target node according to an error difference between the multi-task loss error reference value and the multi-task loss error check value.

[0104] In addition, this application also provides a detailed description of embodiments of a distribution network key node identification terminal and a computer-readable storage medium, as follows:

[0105] like Figure 5 As shown, the present application provides an embodiment of a distribution network key node identification terminal. The terminal implementation types include but are not limited to: personal computers, industrial computers, servers and embedded intelligent devices. The main components of the terminal include: a memory 33 and a processor 31. The memory 33 and the processor 31 can be connected via a communication bus 34;

[0106] The memory 33 is used to store program codes, and the program codes are used to implement a method for identifying key nodes in a distribution network as provided in the above embodiment;

[0107] The processor 31 is used to read and execute program codes.

[0108] The present application provides a computer-readable storage medium, in which program code is stored. The program code is used to be read and executed by a processor to implement a distribution network key node identification method as provided in the above embodiment.

[0109] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the terminals, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0110] In the several embodiments provided in this application, it should be understood that the disclosed terminals, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0111] The terms "first," "second," "third," "fourth," and the like (if any) in the specification of the present application and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of the present application described herein, for example, can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such process, method, product, or apparatus.

[0112] It should be understood that in this application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0113] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0114] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0115] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0116] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for identifying key nodes in a distribution network, characterized in that: include: Obtain node topology data and historical operation data of the distribution network; According to the node topology data, different node relationship graph data are constructed through a plurality of preset graph data construction conditions; The node relationship graph data are integrated to obtain the spatial characteristics of the distribution network, and then the spatiotemporal characteristic vector of the distribution network is obtained based on the historical operation data and the spatial characteristics through a preset time series aggregation model; Based on the spatiotemporal feature vectors corresponding to each node in the distribution network, a simulation operation is performed in combination with a multi-task prediction model to obtain a multi-task loss error benchmark value; Determine any target node from the distribution network, combine the neighboring nodes corresponding to the target node, shield the spatiotemporal feature vectors of the target node and the neighboring nodes, and then perform simulation calculations through a multi-task prediction model to obtain a multi-task loss error check value; A key node identification result of the target node is determined according to an error difference between the multi-task loss error reference value and the multi-task loss error check value.

2. A method for identifying key nodes in a distribution network according to claim 1, characterized in that: The node relationship graph data includes: topological adjacency graph data, geographical distance graph data and node type similarity graph data.

3. A method for identifying key nodes in a distribution network according to claim 1, characterized in that: The fusing of the node relationship graph data to obtain the spatial characteristics of the distribution network specifically includes: The spatial embedding features in each node relationship graph data are extracted through a graph neural network, and then the spatial embedding features of each node relationship graph data are fused to obtain the spatial features of the distribution network.

4. A method for identifying key nodes in a distribution network according to claim 1, characterized in that: The multi-task loss error benchmark value obtained by performing simulation calculations based on the spatiotemporal feature vectors corresponding to each node in the distribution network and combining the multi-task prediction model specifically includes: Based on the spatiotemporal feature vectors corresponding to each node in the distribution network, a simulation operation is performed in combination with a multi-task prediction model to obtain multiple single-task prediction data; A multi-task loss error benchmark value is obtained based on the cumulative sum of the loss errors of each single-task prediction data and the corresponding single-task real data.

5. A method for identifying key nodes in a distribution network according to claim 4, characterized in that: The single-task prediction data includes at least two of load gap prediction data, failure probability prediction data, recovery time prediction data, transmission flow prediction data and charging demand prediction data.

6. A method for identifying key nodes in a distribution network according to claim 4, characterized in that: The step of determining any target node from the distribution network, combining the neighboring nodes corresponding to the target node, shielding the spatiotemporal feature vectors of the target node and the neighboring nodes, and performing simulation calculations through a multi-task prediction model to obtain a multi-task loss error check value specifically includes: Determine any target node from the distribution network, combine the neighboring nodes corresponding to the target node, and shield the spatiotemporal feature vectors of the target node and the neighboring nodes according to Extra shielding logic and / or Intra shielding logic. Based on the spatiotemporal characteristic vector of the distribution network after node shielding processing, a simulation operation is performed through a multi-task prediction model to obtain a first multi-task loss error check value and / or a second multi-task loss error check value, wherein the first multi-task loss error check value is a multi-task loss error check value obtained based on the spatiotemporal characteristic vector of the distribution network processed according to the Extra shielding logic, and the second multi-task loss error check value is a multi-task loss error check value obtained based on the spatiotemporal characteristic vector of the distribution network processed according to the Intra shielding logic.

7. A method for identifying key nodes in a distribution network according to claim 6, characterized in that: Determining the key node identification result of the target node according to the error difference between the multi-task loss error baseline value and the multi-task loss error check value specifically includes: According to the error difference between the multi-task loss error reference value and the multi-task loss error check value, by comparing the error differences corresponding to each target node, a key node identification result of the target node is determined.

8. A device for identifying key nodes in a distribution network, characterized in that: include: A distribution network data acquisition unit, used to acquire node topology data and historical operation data of the distribution network; A graph data construction unit, configured to construct different node relationship graph data according to the node topology data and through a plurality of preset graph data construction conditions; a spatiotemporal feature extraction unit, configured to fuse the node relationship graph data to obtain the spatial features of the distribution network, and then obtain the spatiotemporal feature vector of the distribution network based on the historical operation data and the spatial features through a preset time series aggregation model; The first error calculation unit is used to perform simulation calculations based on the spatiotemporal feature vectors corresponding to each node in the distribution network in combination with the multi-task prediction model to obtain a multi-task loss error benchmark value; A second error calculation unit is configured to determine an arbitrary target node from the distribution network, combine the neighboring nodes corresponding to the target node, shield the spatiotemporal feature vectors of the target node and the neighboring nodes, and then perform simulation calculations through a multi-task prediction model to obtain a multi-task loss error check value; The key node identification unit is used to determine the key node identification result of the target node according to the error difference between the multi-task loss error reference value and the multi-task loss error check value.

9. A distribution network key node identification terminal, characterized in that: include: memory and processor; The memory is used to store program code, and the program code is used to implement a method for identifying key nodes in a distribution network according to any one of claims 1 to 7; The processor is configured to read and execute the program code.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores program code, and the program code is used to be read and executed by a processor to implement a method for identifying key nodes in a distribution network according to any one of claims 1 to 7.

Images