A method, device, equipment and medium for constructing a topological relationship applied to a power distribution network

Abstract

The application discloses a kind of topological relationship construction method, device, equipment and medium applied to distribution network, wherein, the method comprises: obtaining at least one voltage node pair corresponding to at least one feeder associated with target area corresponding monthly voltage dataset;Wherein, voltage node pair is based on any two voltage nodes on each feeder to build;Monthly voltage dataset of each voltage node pair is input into the first node relationship identification model trained in advance in turn, and at least one node relationship identification result is obtained, wherein, node relationship identification result corresponds to each voltage node pair;Based on each node relationship identification result, the topological relationship network corresponding to target area is constructed.The technical scheme of the embodiment of the application realizes the identification of distribution network topological relationship and the construction of topological network, and improves the topological relationship identification accuracy.

Description

Technical Field

[0001] This invention relates to the field of big data analysis technology for power systems, and in particular to a method, apparatus, equipment, and medium for constructing topology relationships in power distribution networks. Background Technology

[0002] In recent years, with the promotion and development of smart grids, more and more power infrastructure applications rely on distribution network topology models, which reflect the topological relationships between numerous distribution network devices in the actual power grid.

[0003] Currently, when identifying the topology information of the power distribution network, the lack of comprehensive early planning in the construction of the power distribution network often leads to the failure to update the topology information in a timely manner due to line renovations, equipment maintenance, and other reasons, resulting in low accuracy of the topology information archive. Furthermore, when there are many underground cables, overhead lines, and other lines with severe obstruction, staff need to manually record the topology structure, which can affect users' electricity use and user experience. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and medium for constructing topology relationships in power distribution networks, thereby enabling the identification of topology relationships and the construction of topology networks, and improving the accuracy of topology relationship identification.

[0005] According to one aspect of the present invention, a method for constructing topology relationships in a power distribution network is provided, the method comprising:

[0006] Obtain the monthly voltage dataset for at least one voltage node pair corresponding to at least one feeder associated with the target region; wherein the voltage node pair is constructed based on any two voltage nodes on each feeder.

[0007] The monthly voltage datasets of each voltage node pair are sequentially input into the pre-trained first node relationship recognition model to obtain at least one node relationship recognition result, wherein the node relationship recognition result corresponds to each voltage node pair;

[0008] Based on the identification results of the relationships between each node, a topological relationship network corresponding to the target region is constructed.

[0009] According to another aspect of the present invention, a topology construction apparatus for a power distribution network is provided, the apparatus comprising:

[0010] The monthly voltage dataset acquisition module is used to acquire the monthly voltage dataset of at least one voltage node pair corresponding to at least one feeder associated with the target area; wherein, the voltage node pair is constructed based on any two voltage nodes on each feeder.

[0011] The identification result determination module is used to sequentially input the monthly voltage dataset of each voltage node pair into the pre-trained first node relationship identification model to obtain at least one node relationship identification result, wherein the node relationship identification result corresponds to each voltage node pair;

[0012] The topology network construction module is used to construct a topology network corresponding to the target region based on the identification results of the relationships between each node.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] At least one processor; and

[0015] A memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the topology construction method for power distribution networks as described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the topology construction method for a power distribution network as described in any embodiment of the present invention.

[0018] The technical solution of this invention obtains monthly voltage datasets of at least one voltage node pair corresponding to at least one feeder associated with a target area. Further, the monthly voltage datasets of each voltage node pair are sequentially input into a pre-trained first node relationship recognition model to obtain at least one node relationship recognition result. Finally, based on the node relationship recognition results, a topology network corresponding to the target area is constructed. This solves the problem in the prior art where the lack of comprehensive pre-construction planning in power distribution networks often leads to a failure to update topology information in a timely manner due to line modifications, equipment maintenance, etc., resulting in low accuracy of topology information archives. Furthermore, when there are many underground cables, overhead lines, or other lines with severe obstruction, manual recording of the topology structure is required, which can affect user electricity consumption. By introducing a model based on a self-attention mechanism to accelerate deep learning algorithms, the identification of power distribution network topology relationships and the construction of the topology network are realized, improving the accuracy of topology relationship recognition.

[0019] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of a topology construction method for a power distribution network according to Embodiment 1 of the present invention;

[0022] Figure 2 This is a schematic diagram of a node topology relationship type provided in Embodiment 1 of the present invention;

[0023] Figure 3 This is a flowchart of a topology construction method for a power distribution network according to Embodiment 2 of the present invention;

[0024] Figure 4 This is a schematic diagram of a topology construction device for a power distribution network according to Embodiment 3 of the present invention;

[0025] Figure 5 This is a schematic diagram of the structure of an electronic device that implements the topology construction method for power distribution networks according to embodiments of the present invention. Detailed Implementation

[0026] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises 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 processes, methods, products, or apparatus.

[0028] Example 1

[0029] Figure 1 This is a flowchart of a method for constructing topology relationships in a distribution network, provided in Embodiment 1 of the present invention. This embodiment is applicable to identifying the topology relationships of distribution network nodes in a target area and constructing a distribution network topology relationship network. This method can be executed by a topology relationship construction device for a distribution network, which can be implemented in hardware and / or software and can be configured in a terminal and / or server. Figure 1 As shown, the method includes:

[0030] S110. Obtain the monthly voltage dataset of at least one voltage node pair corresponding to at least one feeder associated with the target area.

[0031] The target area can be any region where distribution network node topology analysis is required. For example, the target area can be a city, a county, or other region using two different distribution networks. A feeder can be a distribution line branching off from the power bus and directly connected to the load. A feeder can also be called a cable. A feeder can include a bus and at least one branch line. Voltage node pairs are constructed based on any two voltage nodes associated with a feeder. It should be noted that a voltage node can be the convergence point of each branch line and the bus in the feeder, or the power terminal corresponding to the bus terminal and the end of each branch line, or the node corresponding to the main transformer installed on the bus, etc. Generally, combining all voltage nodes on each feeder in pairs yields at least one voltage node pair. The monthly voltage dataset can be a dataset constructed from the node voltage data of the two voltage nodes in each voltage node pair within one month.

[0032] It should be noted that in order to perform node topology analysis on the entire distribution network associated with the target area, it is necessary to determine each distribution network node and its corresponding node voltage data on each feeder associated with the target area. This allows for the construction of a monthly voltage dataset based on the node voltage data of each distribution network node, so that the distribution network topology of the target area can be analyzed based on the monthly voltage dataset of each voltage node pair.

[0033] Optionally, obtaining a monthly voltage dataset of at least one voltage node pair corresponding to at least one feeder associated with the target area includes: obtaining node voltage data of at least one voltage node corresponding to at least one feeder associated with the target area within a first preset time period before the current time; processing the node voltage data according to a preset dataset construction method to construct at least one voltage dataset; and constructing a monthly voltage dataset of at least one node voltage pair based on the voltage datasets of any two voltage nodes among the voltage nodes.

[0034] The first preset duration can be a pre-defined time range used to limit the collection time of historical voltage data. The first preset duration can be any length of time, optionally one month. Node voltage data is collected based on a preset sampling time interval. The preset sampling time interval can be a pre-defined sampling standard used to limit the voltage data sampling time. The preset sampling time interval can be any time interval, optionally 15 minutes. For example, each voltage node can collect voltage data once every 15 minutes, collecting 96 voltage data points per day. Taking a 30-day month as an example, the number of node voltage data points corresponding to this voltage node is 2880.

[0035] In this embodiment, the preset dataset construction method can be a pre-built data processing method used to organize and analyze node voltage data. It should be noted that, since the sampling duration of node voltage data within the first preset time period is relatively long, the connection relationships between nodes may change. To more clearly reflect this change and avoid misinterpreting it as data anomaly, the node voltage data needs to be transformed into a corresponding matrix to obtain the voltage dataset. The voltage dataset corresponds to the voltage nodes.

[0036] For example, the transformation process of node voltage data can be illustrated as follows: For a voltage node, voltage data is collected every 15 minutes, totaling 96 data points per day. Taking a 30-day month as an example, the number of node voltage data points corresponding to this voltage node is 2880. To demonstrate the difference between the monthly voltage curve corresponding to this voltage node and the daily voltage curve reflecting the voltage fluctuation pattern, the 2880 data points can be transformed using a matrix of dimension (32, 90). The matrix has 90 columns per row, structured as follows: One data point is selected every 8 hours, i.e., 3 data points are used per day, totaling 90 data points over 30 days. Taking the first three columns as an example, the first column contains data from 0:00-7:45, the second column from 8:00-15:45, the third column from 16:00-23:45, and so on. The data between the uplink and downlink are selected with an initial time at 15-minute intervals. For example, if the first row is selected with a sampling time of 0:00, the second row is selected with a sampling time of 0:15, the third row is selected with a sampling time of 0:30, and so on. Finally, the 32nd row is selected with a sampling time of 7:45.

[0037] Specifically, all voltage nodes corresponding to at least one feeder associated with the target area are identified, the node voltage data of these voltage nodes within a first preset time period at the current moment are obtained, and a corresponding voltage dataset is constructed based on the node voltage data. Furthermore, any two voltage nodes among these voltage nodes are combined to construct at least one voltage node pair, and a monthly voltage dataset corresponding to the voltage node pair is constructed based on the voltage dataset of each voltage node in the voltage node pair.

[0038] S120. Input the monthly voltage dataset of each voltage node pair into the pre-trained first node relationship recognition model in sequence to obtain at least one node relationship recognition result.

[0039] Among them, the node relationship identification results correspond to each voltage node pair.

[0040] In this embodiment, the first node relationship recognition model can be a pre-trained neural network model that can determine the relationship between voltage nodes based on a monthly voltage dataset of voltage node pairs. For example, the first node relationship recognition model can be an autoregressive sequence generation model. The autoregressive sequence generation model can be a transformer model with lightweight optimization of the sub-networks; this model is a self-attention mechanism-based model to accelerate deep learning algorithms. The autoregressive sequence generation model includes two normalization layers: Layer Norm and Batch Norm, a self-attention layer, and a multilayer perceptron layer. The normalization layer is an algorithm used in deep network models to accelerate neural network training, convergence speed, and stability. Specifically, during model training, the normalization layer normalizes the model training data to achieve a stable forward input distribution and faster model convergence.

[0041] The self-attention sublayer is a neural network layer used to implement the self-attention mechanism of the model. The self-attention mechanism is a mechanism for filtering out less important information from a large amount of information. Attention types are divided into spatial attention and temporal attention. In practical applications, it can also be divided into Soft Attention and Hard Attention. For Soft Attention, all data is paid attention to, and corresponding attention weights are calculated without setting filtering conditions. For Hard Attention, after generating all attention values, some attention values ​​that do not meet the conditions are filtered out, even if their attention weights are 0. This embodiment will not elaborate further. A multilayer perceptron is a feedforward artificial neural network model used to map multiple input datasets to a single output dataset. Those skilled in the art should understand that a typical multilayer perceptron includes three layers: an input layer, a hidden layer, and an output layer. Furthermore, the different layers of a multilayer perceptron neural network are fully connected, meaning that any neuron in one layer is connected to all neurons in the next layer. This embodiment will not elaborate further.

[0042] It should be noted that before applying the first node relationship recognition model of this embodiment, the first node relationship recognition model to be trained needs to be trained first. The specific model training process is as follows: obtain at least one training sample; wherein, the training sample includes, for each training sample, inputting the monthly voltage dataset of the sample voltage node pairs in the current training sample into the first node relationship recognition model to be trained to obtain the actual node relationship; based on the actual node relationship and the node relationship label in the current training sample, determine the loss value, and based on the loss value, correct the model parameters of the first node relationship recognition model to be trained, and take the convergence of the loss function in the first node relationship recognition model to be trained as the training objective, to obtain the trained first node relationship recognition model.

[0043] It should also be noted that the node relationship labels are obtained by annotating the topological relationship between the corresponding two nodes. For example, such as... Figure 2 As shown, the topological relationship between two nodes can be divided into four categories: adjacent nodes, the same line, adjacent lines, and non-adjacent lines. Figure 2 The relationships between node 1 and nodes 2 and 3 are as follows: adjacent nodes with the closest voltage relationship, marked as 0; the relationships between node 1 and nodes 4 and 5 are as follows: the same line with a relatively close voltage relationship, marked as 1; the relationships between node 1 and nodes 6 and 7 are as follows: adjacent lines with a certain voltage relationship, marked as 2; the relationships between node 1 and nodes 8, 9, and 10 are as follows: non-adjacent lines with a weak voltage relationship, marked as 3.

[0044] For example, consider two feeders powered by different transformer substations: Feeder 1 has 14 nodes, and Feeder 2 has 13 nodes. The voltage sampling period for these nodes is two months: March 2021 and March 2022, each month having 31 days, with a sampling interval of 15 minutes. First, construct the voltage dataset for each node in each month: Since March has 31 days, each node has 31 * 96 = 2976 data points per month, stored in a matrix of dimension (32, 93). Therefore, the 27 nodes have a total of 54 voltage datasets over the two months. Then, label the node relationships corresponding to these 54 datasets. First, randomly select two voltage datasets from the same month for labeling. Since the labeled sample set for any two nodes belonging to a certain node relationship in a given month can be... Two months can generate 702 labeled sample sets. Further, all 702 samples are randomly arranged and input into the first-node relationship recognition model for training, thus obtaining the trained first-node relationship recognition model. In this embodiment, to further highlight the classification performance of the first-node relationship recognition model compared to other models, an experiment was conducted comparing it with a CNN-based deep learning classification model. The experimental results are shown in Table 1.

[0045] Table 1 Comparison of model performance based on monthly voltage dataset

[0046]

[0047] As shown in Table 1, the first node relationship recognition model has a higher classification accuracy than the CNN model, and a significantly lower critical error rate (critical error rate refers to classification errors that cross level 1 or 2, such as classifying a topological relationship of class 0 as class 2 or 3).

[0048] It should be noted that when training the model, the sample dataset can be divided into training and testing datasets in a 4:1 ratio, and a five-fold cross-validation method can be used to obtain the optimal hyperparameters of the first node relationship recognition model.

[0049] In specific implementation, after obtaining the monthly voltage dataset of at least one voltage node pair, for each voltage node pair, the monthly voltage dataset of the current voltage node pair can be sequentially input into the first node relationship recognition model to obtain the node relationship recognition result corresponding to the current voltage node pair, thereby finally obtaining the node relationship recognition result corresponding to each voltage node pair.

[0050] S130. Based on the identification results of the relationships between each node, construct a topological relationship network corresponding to the target area.

[0051] Among them, the topological relationship network can be a network composed of connections and adjacency relationships between spatial data based on the principles of topological geometry, that is, the adjacency, association, inclusion and connectivity relationships between entities represented by nodes, arcs and polygons.

[0052] In practical applications, after obtaining the node relationship identification results corresponding to each voltage node pair, each node relationship identification result can be converted into its corresponding voltage node topology relationship based on the correspondence between the node relationship identification results and the voltage node topology relationship. For example, when the node relationship identification result is 0, its corresponding voltage node topology relationship is adjacent nodes; when the node relationship identification result is 1, its corresponding voltage node topology relationship is the same line; when the node relationship identification result is 2, its corresponding voltage node topology relationship is adjacent lines; and when the node relationship identification result is 3, its corresponding voltage node topology relationship is non-adjacent lines.

[0053] Furthermore, based on the voltage node topology relationships corresponding to the node relationship identification results, a topology network of the target region is constructed.

[0054] The technical solution of this invention obtains monthly voltage datasets of at least one voltage node pair corresponding to at least one feeder associated with a target area. Further, the monthly voltage datasets of each voltage node pair are sequentially input into a pre-trained first node relationship recognition model to obtain at least one node relationship recognition result. Finally, based on the node relationship recognition results, a topology network corresponding to the target area is constructed. This solves the problem in the prior art where the lack of comprehensive pre-construction planning in power distribution networks often leads to a failure to update topology information in a timely manner due to line modifications, equipment maintenance, etc., resulting in low accuracy of topology information archives. Furthermore, when there are many underground cables, overhead lines, or other lines with severe obstruction, manual recording of the topology structure is required, which can affect user electricity consumption. By introducing a model based on a self-attention mechanism to accelerate deep learning algorithms, the identification of power distribution network topology relationships and the construction of the topology network are realized, improving the accuracy of topology relationship recognition.

[0055] Example 2

[0056] Figure 3 This is a flowchart of a topology construction method for a power distribution network provided in Embodiment 2 of the present invention. Based on the aforementioned embodiments, step S130 is further refined. For specific implementation details, please refer to the technical solution of this embodiment. Technical terms that are the same as or corresponding to those in the above embodiments will not be repeated here. Figure 3 As shown, the method includes:

[0057] S210. Obtain the monthly voltage dataset of at least one voltage node pair corresponding to at least one feeder associated with the target area.

[0058] S220. Input the monthly voltage dataset of each voltage node pair into the pre-trained first node relationship recognition model in sequence to obtain at least one node relationship recognition result.

[0059] S230. Based on the identification results of the relationships between each node, determine the power grid topology relationship between the corresponding voltage node pairs and the corresponding confidence level.

[0060] In this embodiment, the power grid topology can be the relationships between various nodes in the distribution network, namely, the adjacency, association, inclusion, and connectivity relationships between nodes. It should be noted that the power grid topology is formed from digitized point, line, and surface data to enable users to perform graphical selection, overlay, or merging for query or application analysis requirements.

[0061] In practical applications, after the first node relationship recognition model processes the monthly voltage dataset, the output node relationship recognition results include not only the topological relationship classification labels between the corresponding node pairs, but also the corresponding confidence scores. The advantage of this setting is that the accuracy of the node relationship recognition results can be determined based on the confidence scores. Furthermore, when equipment maintenance or switching operations cause a jump in the topological relationship of a node, the corresponding confidence score will decrease, allowing the determination of whether the node topological relationship has changed based on the magnitude of the confidence score.

[0062] Specifically, after obtaining at least one node relationship identification result, the power grid topology relationship between the corresponding voltage node pairs and the corresponding confidence level can be determined based on the correspondence between the node relationship identification result and the voltage node topology relationship, so as to construct the power grid topology relationship network of the target area based on the topology relationship between each voltage node pair.

[0063] S240. Based on the topological relationships of each power grid and the corresponding confidence levels, a topological relationship network is constructed.

[0064] It should be noted that since the node relationship identification results include not only the topological relationship identifier between the corresponding voltage node pairs, but also the corresponding confidence level, the topological relationship between the corresponding voltage node pairs can be judged based on the confidence level, so that voltage node pairs with abnormal topological relationships can be dealt with in a timely manner.

[0065] Optionally, based on the topological relationships of each power grid and the corresponding confidence levels, a topological relationship network is constructed, including: dividing each voltage node pair according to a preset confidence threshold to obtain at least one normal voltage node pair with a confidence level greater than the preset confidence threshold, and at least one abnormal voltage node pair with a confidence level less than the preset confidence threshold; processing each abnormal voltage node pair according to a preset abnormal node pair processing method to obtain at least one voltage node pair to be processed; and constructing a topological relationship network based on each normal voltage node pair, each voltage node pair to be processed, and the corresponding power grid topological relationships.

[0066] The preset confidence threshold can be a pre-set confidence standard value used to determine whether the topological relationship between each voltage node pair is normal. The preset confidence threshold can be any value, optionally 0.95.

[0067] In practical applications, after obtaining the node relationship identification results, each voltage node pair can be divided according to a preset confidence threshold. Voltage node pairs with a confidence threshold greater than the preset threshold are considered normal voltage node pairs, while those with a confidence threshold less than the preset threshold are considered abnormal voltage node pairs.

[0068] It should be noted that, in order to construct a topology network based on the power grid topology relationships between voltage node pairs, after determining the normal voltage node pairs and abnormal voltage node pairs, each voltage node pair can be stored as a triple of "node-topology relationship identifier-node" as "normal voltage node pair" and "abnormal voltage node pair". For example, for nodes 1 and 2, the monthly voltage datasets corresponding to nodes 1 and 2 are input into the first node relationship identification model. The output result is identifier 3 and the corresponding confidence score. If the confidence score is greater than 0.95, this node pair can be stored as a "normal voltage node pair" in the form of "1-3-2"; if the confidence score is less than 0.95, this node pair can be stored as a "abnormal voltage node pair" in the form of "1-3-2".

[0069] Furthermore, for each voltage node pair classified as an anomalous voltage node pair, in order to determine why the confidence level corresponding to these voltage node pairs is less than a preset confidence threshold, and to determine the normal topology relationship identifier associated with these voltage node pairs, these anomalous voltage node pairs can be retrieved for subsequent processing.

[0070] Optionally, each abnormal voltage node pair is processed according to a preset abnormal node pair processing method to obtain at least one voltage node pair to be processed, including: detecting the voltage data corresponding to each abnormal voltage node in each abnormal voltage node pair; when the voltage difference between two adjacent voltage data is detected to be greater than a preset difference threshold, the corresponding abnormal node and the voltage abnormal time are determined; the daily voltage data of the abnormal node after the voltage abnormal time are obtained, and a daily voltage dataset is constructed based on the daily voltage data; and each voltage node pair to be processed is determined based on the daily voltage dataset and the second node relationship recognition model.

[0071] The preset difference threshold can be a pre-set standard value used to determine whether the voltage difference is normal. The voltage anomaly time can be the time corresponding to a voltage data jump, i.e., the time corresponding to the next voltage data point when the difference between two adjacent voltage data points exceeds the preset difference threshold. Daily voltage data can be the voltage data of a voltage node within 24 hours. For example, the sampling interval for daily voltage data is 15 minutes; therefore, the number of voltage data points corresponding to one voltage node is 96. Correspondingly, the daily voltage dataset can be a dataset composed of a matrix of dimension (1, 96). The second node relationship recognition model can be a pre-trained neural network model that can determine the relationship between voltage nodes based on the daily voltage dataset of voltage node pairs. For example, the second node relationship recognition model can be an autoregressive sequence generation model. It should be noted that the second node relationship recognition model and the first node relationship recognition model are two models with the same model structure. The difference between the two models is that the input of the first node relationship recognition model is the monthly voltage dataset of voltage node pairs, while the input of the second node relationship recognition model is the daily voltage dataset of voltage node pairs.

[0072] In practical applications, after identifying at least one abnormal voltage node pair, the historical voltage data corresponding to each abnormal voltage node in the current abnormal voltage node pair is detected. When the voltage difference between two adjacent voltage data is detected to be greater than a preset difference threshold, it can be determined that the abnormal node has experienced a voltage jump at the moment when the voltage data changes, and the corresponding abnormal node and the moment when the voltage data jump occurs are identified. Further, the daily voltage data after the abnormal voltage moment of this abnormal node are obtained, and a daily voltage dataset is constructed based on the daily voltage data. Finally, the daily voltage dataset is processed based on the second node relationship identification model to obtain each voltage node pair to be processed.

[0073] Optionally, based on the daily voltage dataset and the second node relationship identification model, each voltage node pair to be processed is determined, including: sequentially inputting the daily voltage dataset into the pre-trained second node relationship identification model to obtain the grid topology relationship to be applied and the corresponding confidence level corresponding to each abnormal node pair; updating the abnormal voltage node pairs based on the grid topology relationship to be applied to obtain each voltage node pair to be processed.

[0074] In practical applications, after obtaining the daily voltage datasets of each abnormal node pair, the daily voltage datasets are sequentially input into the second node relationship identification model to obtain the adjusted power grid topology relationship and corresponding confidence level corresponding to each abnormal node pair. This power grid topology relationship can be used as the power grid topology relationship to be applied. Furthermore, the power grid topology relationship between each abnormal node pair is updated according to the power grid topology relationship to be applied, and the updated voltage node pairs are stored again as voltage node pairs to be processed.

[0075] It should be noted that after determining the normal voltage node pairs, the voltage node pairs to be processed, and the corresponding power grid topology, in order to perform topology analysis on the entire distribution network in the target area and determine the distribution network topology connection relationship between each node, the connection sequence of each node and the distribution network can also be associated to obtain the topology network.

[0076] Optionally, a topology relationship network is constructed based on each normal voltage node pair, each voltage node pair to be processed, and the corresponding power grid topology relationship. This includes: processing each normal voltage node pair, each voltage node pair to be processed, and the corresponding power grid topology relationship according to a preset network construction algorithm to construct the topology relationship network.

[0077] The preset network construction algorithm can be a pre-defined method used to analyze node topology relationships and construct a topological network. Optionally, the preset network construction algorithm can be a graph search algorithm.

[0078] Specifically, based on the pre-set network construction algorithm, the normal node pairs, the node pairs to be processed, and the corresponding power grid topology are analyzed. One node is randomly selected from each node pair, and with that node as the starting node, each node is gradually inserted into the connection sequence of the distribution network, thereby finally constructing a topology network and realizing the fine identification of the distribution network topology.

[0079] The technical solution of this invention obtains monthly voltage datasets of at least one voltage node pair corresponding to at least one feeder associated with a target area. Further, the monthly voltage datasets of each voltage node pair are sequentially input into a pre-trained first node relationship recognition model to obtain at least one node relationship recognition result. Finally, based on the node relationship recognition results, the power grid topology relationship between the corresponding voltage node pairs and the corresponding confidence level are determined. Based on each power grid topology relationship and the corresponding confidence level, a topology relationship network is constructed. This solves the problem in the prior art where the lack of comprehensive pre-construction planning in the distribution network construction often leads to a failure to update topology information in a timely manner due to line modifications, equipment maintenance, etc., resulting in low accuracy of topology information archives. Furthermore, when there are many underground cables, overhead lines, or other lines with severe obstruction, manual recording of the topology structure is required, which can affect user electricity consumption. By introducing a model based on a self-attention mechanism to accelerate deep learning algorithms, the identification of distribution network topology relationships and the construction of the topology network are realized, improving the accuracy of topology relationship recognition. Moreover, it can identify node relationship jumps to a certain extent.

[0080] Example 3

[0081] Figure 4 This is a schematic diagram of a topology construction device for a power distribution network provided in Embodiment 3 of the present invention. Figure 4 As shown, the device includes: a monthly voltage dataset acquisition module 310, an identification result determination module 320, and a topology network construction module 330.

[0082] The monthly voltage dataset acquisition module 310 is used to acquire the monthly voltage dataset of at least one voltage node pair corresponding to at least one feeder associated with the target area; wherein, the voltage node pair is constructed based on any two voltage nodes on each feeder.

[0083] The identification result determination module 320 is used to sequentially input the monthly voltage dataset of each voltage node pair into the pre-trained first node relationship identification model to obtain at least one node relationship identification result, wherein the node relationship identification result corresponds to each voltage node pair;

[0084] The topology relationship network construction module 330 is used to construct a topology relationship network corresponding to the target area based on the relationship identification results of each node.

[0085] The technical solution of this invention obtains monthly voltage datasets of at least one voltage node pair corresponding to at least one feeder associated with a target area. Further, the monthly voltage datasets of each voltage node pair are sequentially input into a pre-trained first node relationship recognition model to obtain at least one node relationship recognition result. Finally, based on the node relationship recognition results, a topology network corresponding to the target area is constructed. This solves the problem in the prior art where the lack of comprehensive pre-construction planning in power distribution networks often leads to a failure to update topology information in a timely manner due to line modifications, equipment maintenance, etc., resulting in low accuracy of topology information archives. Furthermore, when there are many underground cables, overhead lines, or other lines with severe obstruction, manual recording of the topology structure is required, which can affect user electricity consumption. By introducing a model based on a self-attention mechanism to accelerate deep learning algorithms, the identification of power distribution network topology relationships and the construction of the topology network are realized, improving the accuracy of topology relationship recognition.

[0086] Optionally, the monthly voltage dataset acquisition module 310 includes a node voltage data acquisition unit, a voltage dataset construction unit, and a monthly voltage dataset construction unit.

[0087] The node voltage data acquisition unit is used to acquire node voltage data of at least one voltage node corresponding to at least one feeder associated with the target area within a first preset time period before the current time; wherein, the node voltage data is acquired based on a preset sampling time interval;

[0088] A voltage dataset construction unit is used to process the node voltage data according to a preset dataset construction method to construct at least one voltage dataset, wherein the voltage dataset corresponds to the voltage node;

[0089] The monthly voltage dataset construction unit is used to construct a monthly voltage dataset of at least one node voltage pair based on the voltage datasets of any two voltage nodes among the voltage nodes.

[0090] Optionally, the topology relationship network construction module 330 includes a power grid topology relationship determination submodule and a topology relationship network construction submodule.

[0091] The power grid topology determination submodule is used to determine the power grid topology relationship between corresponding voltage node pairs and the corresponding confidence level based on the node relationship identification results.

[0092] The topology network construction submodule is used to construct the topology network based on each of the power grid topology relationships and the corresponding confidence levels.

[0093] Optionally, the topology network construction submodule includes a voltage node pair partitioning unit, an abnormal voltage node pair processing unit, and a topology network construction unit.

[0094] A voltage node pair division unit is used to divide each voltage node pair according to a preset confidence threshold to obtain at least one normal voltage node pair with a confidence level greater than the preset confidence threshold, and at least one abnormal voltage node pair with a confidence level less than the preset confidence threshold.

[0095] An abnormal voltage node pair processing unit is used to process each of the abnormal voltage node pairs according to a preset abnormal voltage node pair processing method to obtain at least one voltage node pair to be processed.

[0096] The topology network construction unit is used to construct the topology network based on each normal voltage node pair, each voltage node pair to be processed, and the corresponding power grid topology relationship.

[0097] Optionally, the abnormal voltage node pair processing unit includes a voltage data detection subunit, an abnormal node determination subunit, a daily voltage dataset construction subunit, and a voltage node pair determination subunit to be processed.

[0098] The voltage data detection subunit is used to detect the voltage data corresponding to each abnormal voltage node in each of the abnormal voltage node pairs.

[0099] The abnormal node determination subunit is used to determine the corresponding abnormal node and the time of voltage abnormality when the voltage difference between two adjacent voltage data is detected to be greater than a preset difference threshold.

[0100] The daily voltage dataset construction subunit is used to obtain the daily voltage data of the abnormal node after the voltage abnormality time, and construct the daily voltage dataset based on the daily voltage data.

[0101] The voltage node pair determination subunit is used to determine each of the voltage node pairs to be processed based on the daily voltage dataset and the second node relationship identification model.

[0102] Optionally, the sub-unit for determining voltage node pairs to be processed is further configured to sequentially input the daily voltage dataset into a pre-trained second node relationship recognition model to obtain the grid topology relationship to be applied and the corresponding confidence level corresponding to each of the abnormal node pairs; and update the abnormal voltage node pairs based on the grid topology relationship to be applied to obtain each of the voltage node pairs to be processed.

[0103] Optionally, the topology network construction unit is further configured to process each normal voltage node pair, each voltage node pair to be processed, and the corresponding power grid topology relationship according to a preset network construction algorithm, so as to construct the topology network.

[0104] The topology construction device for power distribution networks provided in this embodiment of the invention can execute the topology construction method for power distribution networks provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0105] Example 4

[0106] Figure 5 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0107] like Figure 5 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0108] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0109] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as topology construction applied to a power distribution network.

[0110] In some embodiments, the topology construction applied to the distribution network can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the topology construction applied to the distribution network described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the topology construction applied to the distribution network by any other suitable means (e.g., by means of firmware).

[0111] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0112] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0113] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0114] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0115] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0116] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0117] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0118] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for constructing a topological relationship applied to a power distribution network, characterized in that, The method comprises: obtaining a monthly voltage data set of at least one voltage node pair corresponding to at least one feeder associated with a target area; wherein the voltage node pair is constructed based on any two voltage nodes on the feeder; inputting the monthly voltage data set of each voltage node pair into a first node relationship identification model trained in advance in sequence to obtain at least one node relationship identification result; wherein the node relationship identification result corresponds to each voltage node pair; for each node relationship identification result, determining the power grid topology relationship between the corresponding voltage node pair and the corresponding confidence degree according to the node relationship identification result; dividing each voltage node pair according to a pre-set confidence threshold to obtain at least one normal voltage node pair with a confidence degree greater than the pre-set confidence threshold and at least one abnormal voltage node pair with a confidence degree less than the pre-set confidence threshold; for each abnormal voltage node pair, detecting the voltage data corresponding to each abnormal voltage node in each abnormal voltage node pair; when the voltage difference between two adjacent voltage data is greater than a pre-set difference threshold, determining the corresponding abnormal node and voltage abnormal moment; obtaining daily voltage data of the abnormal node after the voltage abnormal moment, and constructing a daily voltage data set based on the daily voltage data; inputting the daily voltage data set into a second node relationship identification model trained in advance in sequence to obtain the to-be-applied power grid topology relationship corresponding to each abnormal node pair and the corresponding confidence degree; updating the abnormal voltage node pair based on the to-be-applied power grid topology relationship to obtain each to-be-processed voltage node pair; and constructing a topology relationship network based on each normal voltage node pair, each to-be-processed voltage node pair, and the corresponding power grid topology relationship.

2. The method of claim 1, wherein, The method comprises: obtaining node voltage data of at least one voltage node corresponding to at least one feeder associated with the target area within a first pre-set time length before the current time; wherein the node voltage data is obtained based on a pre-set sampling time interval; processing the node voltage data according to a pre-set data set construction method to construct at least one voltage data set; wherein the voltage data set corresponds to the voltage node; constructing a monthly voltage data set of at least one node voltage pair based on the voltage data set of any two voltage nodes in each voltage node.

3. The method of claim 1, wherein, The method comprises: processing each normal voltage node pair, each to-be-processed voltage node pair, and the corresponding power grid topology relationship according to a pre-set network construction algorithm to construct the topology relationship network.

4. A topology relationship construction device applied to a power distribution network, characterized in that, The method comprises: a monthly voltage data set obtaining module, configured to obtain a monthly voltage data set of at least one voltage node pair corresponding to at least one feeder associated with a target area; wherein the voltage node pair is constructed based on any two voltage nodes on the feeder; The identification result determination module is configured to input the monthly voltage data set of each voltage node pair into a first node relationship identification model that is pre-trained to obtain at least one node relationship identification result, wherein the node relationship identification result corresponds to each voltage node pair; The topology relationship network construction module is configured to construct a topology relationship network corresponding to the target region based on the node relationship identification result; The topology relationship network construction module includes a power grid topology relationship determination submodule and a topology relationship network construction submodule; The power grid topology relationship determination submodule is configured to determine a power grid topology relationship between corresponding voltage node pairs and a corresponding confidence level based on the node relationship identification result; The topology relationship network construction submodule is configured to construct the topology relationship network based on the power grid topology relationship and the corresponding confidence level; The topology relationship network construction submodule includes a voltage node pair division unit, an abnormal voltage node pair processing unit, and a topology relationship network construction unit; The voltage node pair division unit is configured to divide each voltage node pair based on a pre-set confidence level threshold to obtain at least one normal voltage node pair with a confidence level greater than the pre-set confidence level threshold and at least one abnormal voltage node pair with a confidence level less than the pre-set confidence level threshold; The abnormal voltage node pair processing unit is configured to process each abnormal voltage node pair based on a pre-set abnormal node pair processing method to obtain at least one to-be-processed voltage node pair; The topology relationship network construction unit is configured to construct a topology relationship network based on each normal voltage node pair, each to-be-processed voltage node pair, and the corresponding power grid topology relationship; The abnormal voltage node pair processing unit includes a voltage data detection submodule, an abnormal node determination submodule, a daily voltage data set construction submodule, and a to-be-processed voltage node pair determination submodule; The voltage data detection submodule is configured to detect voltage data corresponding to each abnormal voltage node in each abnormal voltage node pair; The abnormal node determination submodule is configured to determine a corresponding abnormal node and a voltage abnormal time when detecting that the voltage difference between two adjacent voltage data is greater than a pre-set difference threshold; The daily voltage data set construction submodule is configured to obtain daily voltage data of the abnormal node after the voltage abnormal time and construct a daily voltage data set based on the daily voltage data; The to-be-processed voltage node pair determination submodule is configured to determine each to-be-processed voltage node pair based on the daily voltage data set and a second node relationship identification model; The to-be-processed voltage node pair determination submodule is further configured to input the daily voltage data set into a pre-trained second node relationship identification model in sequence to obtain a to-be-applied power grid topology relationship corresponding to each abnormal node pair and a corresponding confidence level; and update the abnormal voltage node pair based on the to-be-applied power grid topology relationship to obtain each to-be-processed voltage node pair.

5. An electronic device, comprising: The electronic device includes: at least one processor; and A memory connected in communication with the at least one processor; wherein The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the method for constructing a topological relationship applied to a power distribution network according to any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for causing a processor to execute the method for constructing a topological relationship applied to a power distribution network according to any one of claims 1-3.

Images