Power grid dispatching knowledge spectrum map construction method and device, equipment, medium and product

Abstract

The application relates to a power grid dispatching knowledge spectrum diagram construction method, device, equipment, medium and product; the method comprises the following steps: obtaining power grid dispatching data of a target power grid in a power grid dispatching process; determining entity features corresponding to at least one entity object included in the power grid dispatching data and event features corresponding to at least one dispatching abnormal event included in the power grid dispatching data according to the power grid dispatching data; and determining a power grid dispatching knowledge spectrum diagram of the target power grid according to the entity features corresponding to the at least one entity object and the event features of each dispatching abnormal event participated in by the corresponding entity object. According to the method, the power grid dispatching knowledge spectrum diagram of the target power grid can be obtained in time, comprehensively and accurately after the power grid dispatching data of the target power grid in the power grid dispatching process is obtained, so that the power grid dispatching knowledge spectrum diagram is synchronized with the dispatching change of the target power grid.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, apparatus, equipment, medium and product for constructing a power grid dispatch knowledge graph. Background Technology

[0002] With the continuous expansion of my country's power grid, the increasing complexity of power system operation, and the significant increase in the proportion of renewable energy integration, the amount of data generated in power grid dispatching scenarios is gradually increasing. To provide support for power grid dispatching decision-making, power grid fault analysis, and system situational awareness, entity associations and operational patterns can be constructed based on the data generated in power grid dispatching scenarios to build a power grid dispatching knowledge graph.

[0003] However, the power grid dispatch knowledge graphs constructed in related technologies cannot better meet the update needs of dispatch scenarios, resulting in the power grid dispatch knowledge graphs not being synchronized with the changes in actual power grid dispatch. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, equipment, medium, and product for constructing a power grid dispatch knowledge graph to address the aforementioned technical problems.

[0005] Firstly, this application provides a method for constructing a power grid dispatch knowledge graph, including:

[0006] Acquire grid dispatch data of the target power grid during the grid dispatch process;

[0007] Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data;

[0008] Based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates, the power grid scheduling knowledge map of the target power grid is determined.

[0009] In one embodiment, the grid scheduling knowledge graph of the target power grid is determined based on the entity features corresponding to at least one entity object and the event features of each scheduling anomaly event in which the corresponding entity object participates. This includes: for each entity object, performing time decay attention enhancement on the event features of each scheduling anomaly event in which the entity object participates to obtain enhanced event features corresponding to the entity object; performing feature fusion on the entity features and enhanced event features to obtain fused features; and determining the grid scheduling knowledge graph of the target power grid based on the fused features.

[0010] In one embodiment, determining the event features corresponding to at least one scheduling anomaly event included in the power grid scheduling data based on the power grid scheduling data includes: performing event detection on the power grid scheduling data to obtain at least one scheduling anomaly event; and extracting the event features corresponding to the scheduling anomaly event included in the power grid scheduling data for each scheduling anomaly event.

[0011] In one embodiment, the power grid dispatch data includes power grid operation data; event detection is performed on the power grid dispatch data to obtain at least one dispatch anomaly event, including: acquiring the signal amplitude of the power grid operation data at each dispatch time and the signal amplitude change value at two adjacent dispatch times; if the signal amplitude at at least one dispatch time is greater than a first amplitude threshold or less than a second amplitude threshold, a dispatch anomaly event exceeding the limit is determined; the first amplitude threshold is greater than the second amplitude threshold; if the signal amplitude change value at two adjacent dispatch times is greater than the amplitude change threshold, a dispatch anomaly event of a sudden change is determined.

[0012] In one embodiment, the power grid dispatch data includes at least one of dispatch procedure data, power grid operation data, and wiring diagram data. Extracting event features corresponding to dispatch anomalies included in the power grid dispatch data includes: extracting dispatch features associated with dispatch anomalies from the dispatch procedure data, operation features associated with dispatch anomalies from the power grid operation data, and wiring features associated with dispatch anomalies from the wiring diagram data; and performing feature fusion on at least one of the dispatch features, operation features, and wiring features corresponding to the dispatch anomalies to obtain the event features corresponding to the dispatch anomalies.

[0013] In one embodiment, the power grid dispatch data includes power grid operation data and wiring diagram data: Based on the power grid dispatch data, the entity features corresponding to at least one entity object included in the power grid dispatch data are determined, including: for each entity object included in the power grid dispatch data, the span feature is extracted from the dispatch procedure data to obtain the entity span feature; the connection feature is extracted from the wiring diagram data to obtain the entity connection feature; and the entity span feature and the entity connection feature are fused to obtain the entity feature corresponding to the entity object.

[0014] Secondly, this application also provides a power grid dispatch knowledge graph construction device, comprising:

[0015] The acquisition module is used to acquire grid dispatch data of the target power grid during the grid dispatch process;

[0016] The first determining module is used to determine, based on the power grid dispatch data, the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data.

[0017] The second determining module is used to determine the grid dispatch knowledge map of the target power grid based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates.

[0018] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0019] Acquire grid dispatch data of the target power grid during the grid dispatch process;

[0020] Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data;

[0021] Based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates, the power grid scheduling knowledge map of the target power grid is determined.

[0022] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0023] Acquire grid dispatch data of the target power grid during the grid dispatch process;

[0024] Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data;

[0025] Based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates, the power grid scheduling knowledge map of the target power grid is determined.

[0026] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0027] Acquire grid dispatch data of the target power grid during the grid dispatch process;

[0028] Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data;

[0029] Based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates, the power grid scheduling knowledge map of the target power grid is determined.

[0030] The aforementioned method, apparatus, equipment, medium, and product for constructing a power grid dispatch knowledge graph acquire power grid dispatch data of the target power grid during the power grid dispatch process. Based on the power grid dispatch data, the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data, are determined. Based on the entity characteristics corresponding to at least one entity object and the event characteristics of each dispatch anomaly event in which the corresponding entity object participates, the power grid dispatch knowledge graph of the target power grid is determined. In the above process, after acquiring the power grid dispatch data of the target power grid during the power grid dispatch process, the power grid dispatch knowledge graph of the target power grid can be obtained in a timely, comprehensive, and accurate manner by combining the entity characteristics corresponding to at least one entity object included in the power grid dispatch data and the event characteristics corresponding to at least one dispatch anomaly event, thereby synchronizing the power grid dispatch knowledge graph with the dispatch changes of the target power grid. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is an application environment diagram of the power grid dispatch knowledge graph construction method in one embodiment;

[0033] Figure 2 This is a flowchart illustrating a method for constructing a power grid dispatch knowledge graph in one embodiment;

[0034] Figure 3 This is a flowchart illustrating the steps for determining the power grid dispatch knowledge graph in one embodiment;

[0035] Figure 4 This is a flowchart illustrating the event feature determination steps in one embodiment;

[0036] Figure 5 This is a flowchart illustrating the entity feature determination steps in one embodiment;

[0037] Figure 6 This is a flowchart illustrating the method for constructing a power grid dispatch knowledge graph in another embodiment;

[0038] Figure 7 This is a structural block diagram of a power grid dispatch knowledge graph construction device in one embodiment;

[0039] Figure 8This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with 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.

[0041] The power grid dispatch knowledge graph construction method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or located on the cloud or other network servers. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, projection devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. Server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0042] In one exemplary embodiment, such as Figure 2 As shown, a method for constructing a knowledge graph for power grid dispatching is provided, and this method is applied to... Figure 1 Taking the server in the example, the following steps are included:

[0043] S210, Obtain grid dispatch data of the target power grid during the grid dispatch process.

[0044] Among them, power grid dispatch data can be understood as data generated during the power grid dispatch process.

[0045] The power grid dispatch data can include at least one of the following: dispatch procedure data, power grid operation data, and wiring diagram data. Dispatch procedure data can include dispatch logs and procedure text data. Dispatch logs can be understood as text data recording the operating status, dispatch instructions, and dispatch situations of each device in the target power grid. Procedure text data can be understood as text data providing standardized explanations of power grid dispatch work. Power grid operation data can be understood as the equipment operation data of each device in the target power grid at each time point. Wiring diagram data can be understood as the node data (such as node identifier, type, location, etc.) and connection relationships contained in a wiring diagram constructed with each device in the target power grid as a node and the connection relationships between devices as edges.

[0046] In some embodiments, the aforementioned power grid dispatch data may be pre-stored in a database, and thus the power grid dispatch data can be retrieved from the database.

[0047] S220, Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data.

[0048] The entity object can include at least one of the following in the target power grid: equipment, lines, busbars, switches, protection devices, alarm types, and operating actions. Entity features can be understood as features used to characterize the entity span and entity connection relationships of an entity object. Entity features can include entity span features and entity connection features. Entity span features are used to characterize the entity span of an entity object. Entity connection features are used to characterize the entity connection relationships of an entity object.

[0049] In some embodiments, display relationship triples can be extracted from power grid operation data to obtain a first entity object included in the power grid operation data; a second entity object can be extracted from the wiring diagram data using a target detection algorithm. The first entity object and the second entity object are then merged to obtain at least one entity object included in the power grid dispatch data.

[0050] In this context, a scheduling anomaly event can be understood as an event in which the signal amplitude in the target power grid undergoes abnormal changes during power grid scheduling. Scheduling anomaly events can include at least one of two types: limit-crossing events and mutation events. A limit-crossing event can be understood as an event where the signal amplitude at any scheduling time is greater than a first amplitude threshold or less than a second amplitude threshold, where the first amplitude threshold is greater than the second amplitude threshold. A mutation event can be understood as an event where the change in signal amplitude between two adjacent scheduling times is greater than an amplitude change threshold. Event characteristics can be understood as at least one of the following features used to characterize a scheduling anomaly event: scheduling information, timing information, at least one involved entity object, and entity connection relationships between entity objects.

[0051] S230, Based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates, determine the grid scheduling knowledge map of the target power grid.

[0052] Among these features, entity features and event features can be fused together, and the grid dispatch knowledge graph of the target power grid can be determined based on the fused features.

[0053] In the aforementioned method for constructing a power grid dispatch knowledge graph, the following steps are taken: First, power grid dispatch data of the target power grid during the power grid dispatch process is acquired. Based on this data, the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data, are determined. Then, based on the entity characteristics corresponding to at least one entity object and the event characteristics of each dispatch anomaly event in which the corresponding entity object participates, the power grid dispatch knowledge graph of the target power grid is determined. In this process, after acquiring the power grid dispatch data of the target power grid during the power grid dispatch process, the power grid dispatch knowledge graph of the target power grid can be obtained promptly, comprehensively, and accurately by combining the entity characteristics corresponding to at least one entity object included in the power grid dispatch data and the event characteristics corresponding to at least one dispatch anomaly event, thereby synchronizing the power grid dispatch knowledge graph with the dispatch changes of the target power grid.

[0054] Based on the technical solutions of the above embodiments, this application also provides another optional embodiment, in which the power grid dispatch knowledge graph determination step of S230 is refined.

[0055] See Figure 3 The steps for determining the power grid dispatch knowledge graph shown include:

[0056] S310, for each entity object, time decay attention enhancement is performed on the event characteristics of each scheduling exception event in which the entity object participates, to obtain the enhanced event characteristics corresponding to the entity object.

[0057] In some embodiments, the following time-decayed attention weights can be used to enhance the event characteristics of each scheduling exception event in which the entity object participates:

[0058]

[0059] Where q represents a learnable vector; This represents the time difference between the occurrence of a scheduling anomaly and the current time. Indicates the attenuation coefficient; Represents entity objects The set of all scheduling exceptions involved; p represents the set. Any event index in the dataset.

[0060] S320 fuses entity features and enhanced event features to obtain fused features.

[0061] In some embodiments, entity features and enhanced event features can be concatenated to obtain fused features.

[0062] S330, based on the fusion characteristics, determines the grid dispatch knowledge map of the target power grid.

[0063] Based on the technical solutions of the above embodiments, this application also provides another optional embodiment, in which the event feature determination step of S220 is refined.

[0064] See Figure 4 The event feature determination steps shown include:

[0065] S410 performs event detection on the power grid dispatch data and obtains at least one dispatch anomaly event.

[0066] In some embodiments, grid dispatch data may include grid operation data.

[0067] In some embodiments, the signal amplitude of power grid operation data at each scheduling time and the change in signal amplitude between two adjacent scheduling times can be acquired. If the signal amplitude at at least one scheduling time is greater than a first amplitude threshold or less than a second amplitude threshold, a scheduling anomaly event exceeding the limit is determined; the first amplitude threshold is greater than the second amplitude threshold. If the change in signal amplitude between two adjacent scheduling times is greater than the amplitude change threshold, a scheduling anomaly event due to a sudden change is determined.

[0068] In some embodiments, the signal amplitude at scheduling time t can be... Scheduling time The signal amplitude below is .

[0069] exist satisfy In the case of an out-of-limit scheduling anomaly, determine the occurrence of the anomaly event; among which, This is the upper limit of the amplitude, i.e., the first amplitude threshold; This is the lower limit of the amplitude, i.e., the second amplitude threshold.

[0070] exist satisfy In the case of a sudden change in scheduling, identify the abnormal scheduling event; where, This is the threshold for amplitude change.

[0071] S420: For each scheduling anomaly event, extract the event features corresponding to the scheduling anomaly event included in the power grid scheduling data.

[0072] In some embodiments, power grid dispatch data may include at least one of dispatch procedure data, power grid operation data, and wiring diagram data.

[0073] In some embodiments, scheduling features associated with scheduling anomalies in scheduling procedure data, operation features associated with scheduling anomalies in power grid operation data, and wiring features associated with scheduling anomalies in wiring diagram data can be extracted.

[0074] In practice, for each scheduling exception event, the start time of that scheduling exception event can be taken as the event starting point. The endpoint of the event is defined as the time when the system returns to the normal range. According to the starting point of the event and the end of the event Generate event time window Based on the event time window, the corresponding scheduling procedure data segment, power grid operation data segment, and partial wiring diagram data can be located.

[0075] Then, the dispatching procedure data segments can be encoded to obtain the dispatching characteristics in the dispatching procedure data segments; the power grid operation data segments can be encoded to obtain the operation characteristics in the power grid operation data segments; and the wiring diagram local data can be encoded to obtain the wiring characteristics in the wiring diagram local data.

[0076] To eliminate the feature dimension differences between different modal features, in some embodiments, scheduling features, operation features, and wiring features can be mapped to a shared semantic space to make the feature dimensions of scheduling features, operation features, and wiring features consistent. Then, feature fusion is performed on at least one of the mapped scheduling features, operation features, and wiring features to obtain the event features corresponding to the scheduling abnormal event.

[0077] Based on the technical solutions of the above embodiments, this application also provides another optional embodiment, in which the power grid dispatch data is refined into dispatch procedure data and wiring diagram data: accordingly, the entity feature determination step of S220 is refined.

[0078] See Figure 5 The entity feature determination steps shown include:

[0079] S510: For each entity object contained in the power grid dispatch data, the span feature of the dispatch procedure data is extracted to obtain the entity span feature; the connection feature of the wiring diagram data is extracted to obtain the entity connection feature.

[0080] In some embodiments, the word segmentation sequence in the scheduling procedure data can be denoted as Then, context encoding is performed on the scheduling procedure data to obtain the semantic representation matrix of each token. The semantic representation matrix is ​​then input into the conditional random field layer to learn the transition constraints of neighboring labels and output the label sequence y. Maximum a posteriori decoding is then performed on y to obtain the optimal label sequence. .in, It can be used to label the categories of entities such as equipment, lines, busbars, switches, protection devices, alarm types, and operating actions.

[0081] For each entity object, the segmentation span of the scheduling procedure data fragment containing that entity object is obtained, thus yielding the entity span corresponding to that entity object. Based on the text span corresponding to that entity object, a weighted average is calculated for the optimal entity tag sequence corresponding to that entity object to obtain the entity span feature, as detailed below:

[0082]

[0083] in, Represents entity objects The corresponding segmentation span; t represents the t-th segment in the segmentation span; This represents the tag sequence of the t-th word segment.

[0084] For each entity object, the entity connection characteristics can be determined based on the connection edges between the entity object and other entity objects contained in the wiring diagram data.

[0085] S520 performs feature fusion on entity span features and entity connection features to obtain entity features corresponding to entity objects.

[0086] In some embodiments, entity span features and entity connection features can be concatenated to obtain entity features corresponding to entity objects.

[0087] Based on the technical solutions of the above embodiments, this application also provides another optional embodiment, in which the method for constructing a power grid dispatch knowledge graph is described in detail.

[0088] See Figure 6 The method for constructing the power grid dispatch knowledge graph shown includes:

[0089] S601, acquire the dispatching procedure data, power grid operation data and wiring diagram data of the target power grid during the power grid dispatching process.

[0090] S602, for each entity object contained in the power grid dispatch data, the span feature is extracted from the dispatch procedure data to obtain the entity span feature; the connection feature is extracted from the wiring diagram data to obtain the entity connection feature.

[0091] S603 performs feature fusion on entity span features and entity connection features to obtain entity features corresponding to entity objects.

[0092] S604 performs event detection on the power grid dispatch data and obtains at least one dispatch anomaly event.

[0093] S605 acquires the signal amplitude of the power grid operation data at each scheduling time and the change in signal amplitude between two adjacent scheduling times.

[0094] S606, if the signal amplitude at at least one scheduling moment is greater than a first amplitude threshold or less than a second amplitude threshold, a scheduling anomaly event exceeding the limit is determined to have occurred; the first amplitude threshold is greater than the second amplitude threshold.

[0095] S607: If the change in signal amplitude between two adjacent scheduling times is greater than the amplitude change threshold, a scheduling anomaly event is identified.

[0096] S608, extract the scheduling features associated with scheduling anomalies from the scheduling procedure data, the operation features associated with scheduling anomalies from the power grid operation data, and the wiring features associated with scheduling anomalies from the wiring diagram data.

[0097] S609, perform feature fusion on at least one of the scheduling features, operation features and wiring features corresponding to the scheduling abnormal event to obtain the event features corresponding to the scheduling abnormal event.

[0098] S610: For each entity object, time decay attention enhancement is performed on the event characteristics of each scheduling exception event in which the entity object participates, to obtain the enhanced event characteristics corresponding to the entity object.

[0099] S611 performs feature fusion on entity features and enhanced event features to obtain fused features.

[0100] S612, Based on the fusion characteristics, determine the grid dispatch knowledge map of the target power grid.

[0101] Based on the technical solutions of the above embodiments, this application also provides another optional embodiment, in which the method for constructing a power grid dispatch knowledge graph is described in detail. The power grid dispatch knowledge graph construction method provided by this application can be applied to a power grid dispatch knowledge graph construction model.

[0102] The power grid dispatch knowledge graph construction model can include an input layer, a feature extraction network, a feature fusion network, and a construction network.

[0103] The input layer is used to input grid dispatch data of the target power grid during the grid dispatch process.

[0104] The feature extraction network is used to determine the entity features corresponding to at least one entity object included in the power grid dispatch data, and the event features corresponding to at least one dispatch anomaly event included in the power grid dispatch data, based on the power grid dispatch data.

[0105] The feature fusion network can include an attention layer and a feature fusion layer. The attention layer, for each entity object, performs time-decay attention enhancement on the event features of each scheduling exception event in which the entity object participates, resulting in enhanced event features corresponding to the entity object. The feature fusion layer fuses the entity features and the enhanced event features to obtain fused features.

[0106] A network is constructed to determine the grid dispatch knowledge graph of the target power grid based on the fusion characteristics.

[0107] In some embodiments, the power grid dispatch knowledge graph construction model can be implemented based on a graph neural network. Based on the power grid dispatch data of the target power grid during the power grid dispatch process, entity objects and the connections between these entities are extracted from the dispatch data. An initial knowledge graph is constructed using entity objects as nodes and the connections between them as edges. In each layer of propagation, the features of neighboring nodes and edge features are aggregated for the current node to obtain an adjacency matrix. After each layer of propagation, the relation confidence of each edge is calculated based on the updated node features (entity features) and edge features (connection relationship features). Edges are dynamically added or deleted based on the relation confidence and multi-hop inference results, and the adjacency matrix is ​​updated. Based on the iterative propagation and dynamic evolution at each layer, a unified representation set of nodes and the final adjacency matrix are output. The power grid dispatch knowledge graph is then instantiated.

[0108] Based on the technical solutions of the above embodiments, this application also provides another optional embodiment, in which the training process of the power grid dispatch knowledge graph construction model is described in detail.

[0109] Acquire sample grid dispatch data and a sample grid dispatch knowledge graph of the reference grid during grid dispatching. Input the sample grid dispatch data into the grid dispatch knowledge graph construction model. Based on the sample grid dispatch data, determine the sample entity features corresponding to at least one entity object included in the sample grid dispatch data, and the sample event features corresponding to at least one dispatching anomaly event included in the sample grid dispatch data. Based on the sample entity features corresponding to at least one entity object and the sample event features of each dispatching anomaly event in which the corresponding entity object participates, determine the predicted grid dispatch knowledge graph of the reference grid. Based on the sample grid dispatch knowledge graph and the predicted grid dispatch knowledge graph, determine the training loss, and adjust the parameters in the grid dispatch knowledge graph construction model according to the training loss to train the grid dispatch knowledge graph construction model.

[0110] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0111] Based on the same inventive concept, this application also provides a power grid dispatching knowledge graph construction device for implementing the power grid dispatching knowledge graph construction method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more power grid dispatching knowledge graph construction device embodiments provided below can be found in the limitations of the power grid dispatching knowledge graph construction method described above, and will not be repeated here.

[0112] In one exemplary embodiment, such as Figure 7 As shown, a power grid dispatch knowledge graph construction device is provided, comprising: an acquisition module 710, a first determination module 720, and a second determination module 730, wherein:

[0113] The acquisition module 710 is used to acquire power grid dispatch data of the target power grid during the power grid dispatch process;

[0114] The first determining module 720 is used to determine, based on the power grid dispatch data, the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatching abnormal event included in the power grid dispatch data.

[0115] The second determining module 730 is used to determine the grid dispatch knowledge map of the target power grid based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates.

[0116] In some embodiments, the second determining module 730 is specifically used to perform time decay attention enhancement on the event features of each scheduling anomaly event in which the entity object participates for each entity object, to obtain the enhanced event features corresponding to the entity object; perform feature fusion on the entity features and the enhanced event features to obtain fused features; and determine the grid scheduling knowledge map of the target power grid based on the fused features.

[0117] In some embodiments, the first determining module 720 is specifically used to perform event detection on the power grid dispatch data to obtain at least one dispatch anomaly event; for each dispatch anomaly event, extract the event features corresponding to the dispatch anomaly event included in the power grid dispatch data.

[0118] In some embodiments, the power grid dispatch data includes power grid operation data; the first determining module 720 is specifically used to acquire the signal amplitude of the power grid operation data at each dispatch time and the signal amplitude change value at two adjacent dispatch times; if the signal amplitude at at least one dispatch time is greater than a first amplitude threshold or less than a second amplitude threshold, determine that a dispatch anomaly event has occurred; the first amplitude threshold is greater than the second amplitude threshold; if the signal amplitude change value at two adjacent dispatch times is greater than the amplitude change threshold, determine that a dispatch anomaly event has occurred.

[0119] In some embodiments, the power grid dispatch data includes at least one of dispatch procedure data, power grid operation data, and wiring diagram data: the first determining module 720 is specifically used to extract dispatch features associated with dispatch abnormal events from the dispatch procedure data, operation features associated with dispatch abnormal events from the power grid operation data, and wiring features associated with dispatch abnormal events from the wiring diagram data; and to perform feature fusion on at least one of the dispatch features, operation features, and wiring features corresponding to the dispatch abnormal event to obtain the event features corresponding to the dispatch abnormal event.

[0120] In some embodiments, the power grid dispatch data includes dispatch procedure data and wiring diagram data: the first determining module 720 is specifically used to extract the span feature from the dispatch procedure data for each entity object contained in the power grid dispatch data to obtain entity span features; extract the connection feature from the wiring diagram data to obtain entity connection features; and perform feature fusion on the entity span features and entity connection features to obtain entity features corresponding to the entity object.

[0121] Each module in the aforementioned power grid dispatch knowledge graph construction device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0122] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8 As shown, the computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores data such as power grid dispatching data, entity characteristics, event characteristics, and a power grid dispatching knowledge graph. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for constructing a power grid dispatching knowledge graph.

[0123] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0124] In one exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the power grid dispatch knowledge graph construction method provided in any of the above embodiments, or to implement the following steps:

[0125] Acquire grid dispatch data of the target power grid during the grid dispatch process;

[0126] Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data;

[0127] Based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates, the power grid scheduling knowledge map of the target power grid is determined.

[0128] In one embodiment, when the processor executes the computer program, it further performs the following steps: for each entity object, time decay attention enhancement is performed on the event features of each scheduling anomaly event in which the entity object participates, to obtain the enhanced event features corresponding to the entity object; feature fusion is performed on the entity features and the enhanced event features to obtain fused features; and the grid scheduling knowledge map of the target power grid is determined based on the fused features.

[0129] In one embodiment, when the processor executes the computer program, it further performs the following steps: performing event detection on the power grid dispatch data to obtain at least one dispatch anomaly event; and for each dispatch anomaly event, extracting the event features corresponding to the dispatch anomaly event included in the power grid dispatch data.

[0130] In one embodiment, the power grid dispatch data includes power grid operation data; when the processor executes the computer program, it further implements the following steps: acquiring the signal amplitude of the power grid operation data at each dispatch time and the signal amplitude change value at two adjacent dispatch times; determining that a dispatch anomaly event has occurred if the signal amplitude at at least one dispatch time is greater than a first amplitude threshold or less than a second amplitude threshold; the first amplitude threshold is greater than the second amplitude threshold; and determining that a dispatch anomaly event has occurred if the signal amplitude change value at two adjacent dispatch times is greater than the amplitude change threshold.

[0131] In one embodiment, the power grid dispatch data includes at least one of dispatch procedure data, power grid operation data, and wiring diagram data. When the processor executes the computer program, it further performs the following steps: extracting dispatch features associated with dispatch anomalies from the dispatch procedure data, operation features associated with dispatch anomalies from the power grid operation data, and wiring features associated with dispatch anomalies from the wiring diagram data; and performing feature fusion on at least one of the dispatch features, operation features, and wiring features corresponding to the dispatch anomaly event to obtain the event features corresponding to the dispatch anomaly event.

[0132] In one embodiment, the power grid dispatch data includes power grid operation data and wiring diagram data. When the processor executes the computer program, it also performs the following steps: for each entity object contained in the power grid dispatch data, the processor extracts the span feature from the dispatch procedure data to obtain the entity span feature; the processor extracts the connection feature from the wiring diagram data to obtain the entity connection feature; and the processor fuses the entity span feature and the entity connection feature to obtain the entity feature corresponding to the entity object.

[0133] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When executed by a processor, the computer program implements the steps of the power grid dispatch knowledge graph construction method provided in any of the above embodiments, or implements the following steps:

[0134] Acquire grid dispatch data of the target power grid during the grid dispatch process;

[0135] Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data;

[0136] Based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates, the power grid scheduling knowledge map of the target power grid is determined.

[0137] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: for each entity object, time decay attention enhancement is performed on the event features of each scheduling anomaly event in which the entity object participates, to obtain the enhanced event features corresponding to the entity object; feature fusion is performed on the entity features and the enhanced event features to obtain fused features; and the grid scheduling knowledge map of the target power grid is determined based on the fused features.

[0138] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: performing event detection on the power grid dispatch data to obtain at least one dispatch anomaly event; and for each dispatch anomaly event, extracting the event features corresponding to the dispatch anomaly event included in the power grid dispatch data.

[0139] In one embodiment, the power grid dispatch data includes power grid operation data; when the computer program is executed by the processor, it further performs the following steps: acquiring the signal amplitude of the power grid operation data at each dispatch time and the signal amplitude change value at two adjacent dispatch times; determining that a dispatch anomaly event has occurred if the signal amplitude at at least one dispatch time is greater than a first amplitude threshold or less than a second amplitude threshold; the first amplitude threshold is greater than the second amplitude threshold; and determining that a dispatch anomaly event has occurred if the signal amplitude change value at two adjacent dispatch times is greater than the amplitude change threshold.

[0140] In one embodiment, the power grid dispatch data includes at least one of dispatch procedure data, power grid operation data, and wiring diagram data. When the computer program is executed by the processor, it further performs the following steps: extracting dispatch features associated with dispatch anomalies from the dispatch procedure data, operation features associated with dispatch anomalies from the power grid operation data, and wiring features associated with dispatch anomalies from the wiring diagram data; and performing feature fusion on at least one of the dispatch features, operation features, and wiring features corresponding to the dispatch anomaly event to obtain the event features corresponding to the dispatch anomaly event.

[0141] In one embodiment, the power grid dispatch data includes power grid operation data and wiring diagram data. When the computer program is executed by the processor, it also performs the following steps: for each entity object contained in the power grid dispatch data, the span feature is extracted from the dispatch procedure data to obtain the entity span feature; the connection feature is extracted from the wiring diagram data to obtain the entity connection feature; and the entity span feature and entity connection feature are fused to obtain the entity feature corresponding to the entity object.

[0142] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the power grid dispatch knowledge graph construction method provided in any of the above embodiments, or implements the following steps:

[0143] Acquire grid dispatch data of the target power grid during the grid dispatch process;

[0144] Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data;

[0145] Based on the entity characteristics corresponding to at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates, the power grid scheduling knowledge map of the target power grid is determined.

[0146] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: for each entity object, time decay attention enhancement is performed on the event features of each scheduling anomaly event in which the entity object participates, to obtain the enhanced event features corresponding to the entity object; feature fusion is performed on the entity features and the enhanced event features to obtain fused features; and the grid scheduling knowledge map of the target power grid is determined based on the fused features.

[0147] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: performing event detection on the power grid dispatch data to obtain at least one dispatch anomaly event; and for each dispatch anomaly event, extracting the event features corresponding to the dispatch anomaly event included in the power grid dispatch data.

[0148] In one embodiment, the power grid dispatch data includes power grid operation data; when the computer program is executed by the processor, it further performs the following steps: acquiring the signal amplitude of the power grid operation data at each dispatch time and the signal amplitude change value at two adjacent dispatch times; determining that a dispatch anomaly event has occurred if the signal amplitude at at least one dispatch time is greater than a first amplitude threshold or less than a second amplitude threshold; the first amplitude threshold is greater than the second amplitude threshold; and determining that a dispatch anomaly event has occurred if the signal amplitude change value at two adjacent dispatch times is greater than the amplitude change threshold.

[0149] In one embodiment, the power grid dispatch data includes at least one of dispatch procedure data, power grid operation data, and wiring diagram data. When the computer program is executed by the processor, it further performs the following steps: extracting dispatch features associated with dispatch anomalies from the dispatch procedure data, operation features associated with dispatch anomalies from the power grid operation data, and wiring features associated with dispatch anomalies from the wiring diagram data; and performing feature fusion on at least one of the dispatch features, operation features, and wiring features corresponding to the dispatch anomaly event to obtain the event features corresponding to the dispatch anomaly event.

[0150] In one embodiment, the power grid dispatch data includes power grid operation data and wiring diagram data. When the computer program is executed by the processor, it also performs the following steps: for each entity object contained in the power grid dispatch data, the span feature is extracted from the dispatch procedure data to obtain the entity span feature; the connection feature is extracted from the wiring diagram data to obtain the entity connection feature; and the entity span feature and entity connection feature are fused to obtain the entity feature corresponding to the entity object.

[0151] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0152] 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 application.

[0153] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. 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 application should be determined by the appended claims.

Claims

1. A method for constructing a knowledge graph for power grid dispatching, characterized in that, The method includes: Acquire grid dispatch data of the target power grid during the grid dispatch process; Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data; Based on the entity characteristics corresponding to the at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates, the power grid scheduling knowledge map of the target power grid is determined.

2. The method according to claim 1, characterized in that, The step of determining the grid dispatch knowledge map of the target power grid based on the entity characteristics corresponding to the at least one entity object and the event characteristics of each dispatch anomaly event in which the corresponding entity object participates includes: For each entity object, time decay attention enhancement is performed on the event characteristics of each scheduling anomaly event in which the entity object participates, to obtain the enhanced event characteristics corresponding to the entity object. The entity features and the enhanced event features are fused to obtain fused features; Based on the fusion characteristics, the power grid dispatch knowledge map of the target power grid is determined.

3. The method according to claim 1, characterized in that, Based on the power grid dispatch data, determine the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data, including: Event detection is performed on the power grid dispatch data to obtain at least one dispatch anomaly event; For each scheduling anomaly event, the event features corresponding to the scheduling anomaly event included in the power grid scheduling data are extracted.

4. The method according to claim 3, characterized in that, The power grid dispatch data includes power grid operation data; The event detection of the power grid dispatch data yields at least one dispatch anomaly event, including: The signal amplitude of the power grid operation data at each scheduling time and the change in signal amplitude between two adjacent scheduling times are obtained. If the signal amplitude at at least one scheduling moment is greater than a first amplitude threshold or less than a second amplitude threshold, a scheduling anomaly event exceeding the limit is determined to have occurred; the first amplitude threshold is greater than the second amplitude threshold. If the change in signal amplitude between two adjacent scheduling times exceeds the amplitude change threshold, a scheduling anomaly event is identified.

5. The method according to claim 3, characterized in that, The power grid dispatch data includes at least one of the following: dispatch procedure data, power grid operation data, and wiring diagram data: Extracting event features corresponding to the scheduling anomaly events included in the power grid scheduling data, including: Extract the scheduling features associated with the scheduling anomaly from the scheduling procedure data, the operation features associated with the scheduling anomaly from the power grid operation data, and the wiring features associated with the scheduling anomaly from the wiring diagram data; The scheduling anomaly event is subjected to feature fusion of at least one of the scheduling features, the operation features, and the wiring features to obtain the event features corresponding to the scheduling anomaly event.

6. The method according to claim 1, characterized in that, The power grid dispatch data includes dispatch procedure data and wiring diagram data: Based on the power grid dispatch data, determine the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, including: For each entity object contained in the power grid dispatch data, the span feature is extracted from the dispatch procedure data to obtain entity span features; the connection feature is extracted from the wiring diagram data to obtain entity connection features. The entity span feature and the entity connection feature are fused to obtain the entity feature corresponding to the entity object.

7. A device for constructing a knowledge graph for power grid dispatching, characterized in that, The device includes: The acquisition module is used to acquire grid dispatch data of the target power grid during the grid dispatch process; The first determining module is used to determine, based on the power grid dispatch data, the entity characteristics corresponding to at least one entity object included in the power grid dispatch data, and the event characteristics corresponding to at least one dispatch anomaly event included in the power grid dispatch data. The second determining module is used to determine the grid scheduling knowledge map of the target power grid based on the entity characteristics corresponding to the at least one entity object and the event characteristics of each scheduling anomaly event in which the corresponding entity object participates.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images