Low-voltage area feeder branch topology identification method and device, storage medium and equipment

Abstract

This application relates to a method for identifying the topology of feeder branches in low-voltage distribution substations. The method includes: collecting time-series voltage data from N measurement points in the low-voltage feeder topology and calculating the Euclidean distance between each measurement point based on the time-series voltage data; determining an abstract topology and the corresponding correlation matrix of the abstract topology based on the low-voltage feeder topology; performing cluster analysis on the Euclidean distances between each measurement point to obtain a cluster Euclidean distance matrix; and determining the specific location of each measurement point in the abstract topology based on the cluster Euclidean distance matrix and the correlation matrix. Compared with the prior art, this invention can quickly identify the topology of feeder branches in low-voltage distribution substations without increasing the number of line branch terminals, meeting the needs of practical applications.

Description

Technical Field

[0001] This application relates to the field of power distribution system technology, and in particular to a method, apparatus, storage medium and power equipment for identifying the feeder branch topology of a low-voltage distribution area. Background Technology

[0002] In a power system, a distribution area refers to the power supply range or region of a single transformer. Low-voltage distribution areas, as the "last 100 meters" directly facing users, connect countless households, and the level of intelligence in their operation, maintenance, and management directly impacts customer satisfaction. Correctly identifying the topology of low-voltage distribution areas is crucial for power supply departments to calculate power flow, update switch status changes, analyze and assess faults, implement remote fee control, analyze line losses, and propose optimization strategies. It is a prerequisite for the safe operation and efficient management of the power grid.

[0003] However, urban low-voltage power supply areas are numerous and their connections are often chaotic, with some even exhibiting non-compliant wiring. With the increasing application of distributed power sources, controllable loads, electric vehicle integration, and demand response technologies, the safety and economic efficiency of the power grid have improved. Simultaneously, the problem of frequent topology changes in urban low-voltage power supply areas has become increasingly prominent. Consequently, the topology records kept by power supply departments often suffer from data loss and errors, requiring manual investigation of topology relationships between users. This process is not only costly in terms of manpower and resources but also inefficient. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, device, storage medium, and power equipment for identifying the feeder branch topology of a low-voltage distribution substation, which can quickly identify the feeder branch topology of the low-voltage distribution substation without increasing the number of line branch terminals, in order to address the above-mentioned technical problems.

[0005] This invention provides a method for identifying the feeder branch topology in a low-voltage distribution area, the method comprising the following steps:

[0006] Collect time-series voltage data from N measuring points in the low-voltage feeder topology, and calculate the Euclidean distance between each measuring point based on the time-series voltage data;

[0007] An abstract topology and the corresponding correlation matrix are determined by the low-voltage feeder topology.

[0008] Cluster analysis was performed on the Euclidean distances between the measurement points to obtain the Euclidean distance matrix of the cluster points.

[0009] The specific location of the measurement point in the abstract topology is determined based on the Euclidean distance matrix and the incidence matrix of the cluster point.

[0010] Furthermore, the method for determining an abstract topology and the corresponding correlation matrix of the abstract topology through the low-voltage feeder topology includes:

[0011] The number of feeder branch nodes and the number of branches in each branch node are determined by the shape of the low-voltage feeder topology.

[0012] Based on the number of feeder branch nodes and the number of branches of each branch node, an abstract topology diagram and the corresponding correlation matrix are determined.

[0013] Furthermore, the method for collecting time-series voltage data from N measurement points in a low-voltage feeder topology and calculating the Euclidean distance between each measurement point based on the time-series voltage data includes:

[0014] Generate a time-series voltage matrix based on the time-series voltage data of N measurement points within the target time period;

[0015] The time-series voltage matrix is ​​normalized using a normalization algorithm to obtain the voltage data set of all measurement points at all times.

[0016] Obtain the correlation matrix of each element in the voltage data set, as well as the eigenvector matrix, contribution rate matrix and eigenvalue matrix of the correlation matrix, and filter out the principal component data;

[0017] Calculate the Euclidean distance matrix between each measurement point based on the principal component data.

[0018] Furthermore, methods for obtaining the Euclidean distance matrix of cluster points by performing cluster analysis on the Euclidean distances between each measurement point include:

[0019] The clustering tree matrix of each measurement point is calculated using the sum of squared deviations algorithm and the Euclidean distance between measurement points;

[0020] Cluster the N measurement points into n classes, and calculate the average Euclidean distance of the measurement point voltages in each cluster to obtain the average time-series voltage set of all clusters;

[0021] The Euclidean distance matrix of the clusters of n types is obtained by using the average time-series voltage set of all clusters.

[0022] Furthermore, the method for determining the specific location of the measurement point in the abstract topology based on the Euclidean distance matrix and the incidence matrix of the cluster points includes:

[0023] Search the column number of the non-zero element in the correlation matrix according to the row number of the measurement point to obtain the first set with the first number of elements;

[0024] Search all elements in the Euclidean distance matrix of the cluster points according to the row number of the measurement points to obtain a second set with the first number of elements;

[0025] Based on the first set and the second set, the cluster points in the Euclidean distance matrix are associated with the measurement points in the correlation matrix to obtain the specific location of the measurement points in the abstract topology.

[0026] Furthermore, the method also includes:

[0027] Define a large number, and when the cluster point corresponds to the measurement point, replace the elements in the Euclidean distance moments of the cluster point with the large number;

[0028] When the number of elements in the first set is 0, output the cluster number in the Euclidean distance matrix, the node number in the correlation matrix corresponding to the cluster, and the cluster number connected to the cluster, to determine the specific position of the cluster in the abstract matrix.

[0029] Furthermore, the method also includes:

[0030] When the number of elements in the first set is not 0, return to the step of searching all elements in the Euclidean distance matrix of the cluster points according to the row number of the measurement point to obtain the second set with the first number of elements.

[0031] Another embodiment of the present invention provides a low-voltage distribution area feeder branch topology identification device, the device comprising:

[0032] The data acquisition module is used to acquire time-series voltage data from N measuring points in the low-voltage feeder topology, and to calculate the Euclidean distance between each measuring point based on the time-series voltage data.

[0033] The clustering analysis module is used to perform clustering analysis on the Euclidean distance between each measurement point to obtain the Euclidean distance matrix of the cluster points.

[0034] The topology determination module is used to determine an abstract topology and the corresponding correlation matrix of the abstract topology based on the low-voltage feeder topology.

[0035] The location determination module is used to determine the specific location of the measurement point in the abstract topology based on the Euclidean distance matrix and the correlation matrix of the cluster points.

[0036] Another embodiment of the present invention provides a computer-readable storage medium comprising a stored computer program; wherein, when the computer program is executed, it controls the device in which the computer-readable storage medium is located to perform the low-voltage substation feeder branch topology identification method as described above.

[0037] Another embodiment of the present invention provides a power device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the low-voltage distribution area feeder branch topology identification method as described above.

[0038] The aforementioned method for identifying the topology of low-voltage distribution transformer feeder branches involves collecting time-series voltage data from N measurement points along the low-voltage feeder topology and calculating the Euclidean distance between these points based on the time-series voltage data. It then determines an abstract topology and the corresponding correlation matrix based on the low-voltage feeder topology. Next, it performs cluster analysis on the Euclidean distances between the measurement points to obtain a cluster Euclidean distance matrix. Finally, it determines the specific location of each measurement point within the abstract topology based on the cluster Euclidean distance matrix and the correlation matrix. Compared to existing technologies, this invention can quickly identify the topology of low-voltage distribution transformer feeder branches without increasing the number of line branch terminals, thus meeting practical application requirements. Attached Figure Description

[0039] Figure 1 A schematic flowchart of a low-voltage distribution area feeder branch topology identification method provided in an embodiment of the present invention;

[0040] Figure 2 A schematic diagram illustrating the specific process of the low-voltage distribution area feeder branch topology identification method provided in an embodiment of the present invention;

[0041] Figure 3 Original feeder topology (left) and abstract topology (right);

[0042] Figure 4 This is an example diagram of the feeder topology for a transformer substation.

[0043] Figure 5 Abstract topology diagram of feeder lines in the transformer area

[0044] Figure 6 Generate a tree for the feeder topology clustering system of the transformer area;

[0045] Figure 7 This is a structural block diagram of the low-voltage substation feeder branch topology identification device provided in an embodiment of the present invention;

[0046] Figure 8 This is a structural diagram of the device terminal provided in an embodiment of the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0048] It should be noted that the step numbers in this document are only for the convenience of explaining the specific embodiments and are not intended to limit the order in which the steps are executed. The method provided in this embodiment can be executed by a relevant server, and the following description will use a server as the execution subject.

[0049] like Figures 1 to 2 As shown, the low-voltage distribution area feeder branch topology identification method provided in this embodiment of the invention includes steps S11 to S14:

[0050] Step S11: Collect time-series voltage data from N measurement points in the low-voltage feeder topology, and calculate the Euclidean distance between each measurement point based on the time-series voltage data.

[0051] Specifically, a time-series voltage matrix is ​​generated based on the time-series voltage data of N measuring points within the target time period; the time-series voltage matrix is ​​normalized using a normalization algorithm to obtain a set of voltage data for all measuring points at all times; the correlation matrix of each element in the voltage data set, as well as the eigenvector matrix, contribution rate matrix, and eigenvalue matrix of the correlation matrix, are obtained, and principal component data are selected; the Euclidean distance matrix between each measuring point is calculated based on the principal component data.

[0052] Furthermore, suppose there are N voltage measurement points on a feeder line in a certain low-voltage distribution area, including the voltage measurement point at the beginning of the feeder line; collect the effective voltage values ​​of each measurement point in a certain time sequence over a period of time. For example, based on the time-series voltage data collected from the measurement points from time 1 to time m, generate the measurement point time-series voltage matrix U, which can be expressed as:

[0053] (6)

[0054] in, This represents the effective voltage value collected at measurement point N at time m.

[0055] The Z-Score normalization method preserves the original distribution characteristics of the time-series voltage matrix U at the measurement points. However, the voltage fluctuation characteristics differ between different measurement points, necessitating the elimination of the influence of statistical variance. The Z-Score normalization method, which eliminates the mean and standardized variance, can address this issue. Its expression is:

[0056] (7)

[0057] (8)

[0058] In the formula, This represents the standardized voltage value at measurement point N at time u. This represents the mean of the effective voltage values ​​at all measuring points at time u; It represents the standard deviation of the effective voltage value of measuring point N at time u; It represents the original effective voltage value of measuring point N at time u, It represents the set of voltage data of all measuring points at all times. It retains the original distribution characteristics of the original time-sequence voltage matrix U of the original measuring points and amplifies the relative differences between the measuring points, which is conducive to the judgment of subsequent classification and identification.

[0059] By calculating the set The correlation relationship between each measuring point is obtained by calculating the correlation coefficient of each element in it, and it is used as the classification criterion for measuring points; [[ID=1...]] The correlation matrix of

[0060] (9)

[0061] Among them, R represents the correlation coefficient between two measuring points i and j, and N is the dimension of the voltage matrix 1 < t, ii < u represents the acquisition time, i represents the number of the measuring point, It represents the normalized voltage corresponding to measuring point i at time t.

[0062] According to the characteristic matrix R, the eigenvector matrix V is calculated,

[0063] (10)

[0064] (11)

[0065] Among them, It represents the eigenvector of measuring point 1 of the correlation matrix R; is the corresponding eigenvalue. [[ID=...]]

[0066] The elements in the eigenvalue matrix can be sorted and analyzed. When the eigenvalue is less than 1, this principal component needs to be removed because the explanatory power of this principal component is very low and its contribution to the overall data is very small.

[0067] According to the eigenvalue matrix L, the principal component contribution rate matrix T of each measuring point of the distribution transformer feeder is calculated, and the calculation formula is:

[0068] (12)

[0069] (13)

[0070] Among them, 1 < b < u, t

[0071] ,

[0070] , ,

[0072] , , b , , represents the principal component contribution rate of the measuring point, and 1 < m < u is the acquisition time.

[0071] [[ID=6...]]Calculate the cumulative contribution rate of each component

[0072] (14)

[0073] Among them, 1 < tt < u represents the cumulative contribution rate of the principal components.

[0074] Calculate the scores of each component, and the expression is

[0075] (15)

[0076] Among them, df N represents the score of the principal component, U' NM is the standardized voltage matrix, l N is the eigenvector, m is the time, and N is the measurement point.

[0077] Select the number of principal components whose cumulative contribution rate is greater than 80% and denote it as . Obtain the standardized measurement point voltage data matrix, reduce the dimension of the measurement point matrix, and use the data with few samples to reflect the overall characteristics. If the original data is used, the dimension of the voltage matrix is , N is the total number of measurement points, and u is the dimension of the voltage matrix of the measurement time of the measurement points, which is . Compared with the original voltage matrix, the dimension reduction is obvious because the number of principal components is much smaller than the measurement time of the measurement points.

[0078] Calculate the Euclidean distance of the effective voltage data between each measurement point according to the principal component data:

[0079] (1)

[0080] Among them represents the Euclidean distance obtained from the time-series effective voltage data of measurement point <00002​​​​​​​​​​​​​​​​​​​​​​​Specifically, the number of feeder branch nodes and the number of branches of each branch node are determined by the shape of the low-voltage feeder topology; based on the number of feeder branch nodes and the number of branches of each branch node, the abstract topology diagram and the topology correlation matrix corresponding to the abstract topology are determined.

[0085] Furthermore, based on the approximate topology of the feeder in this area, the number of branch nodes of the feeder is determined, and denoted as [missing information]. The feeder is numbered sequentially from the beginning (number 1) to the end (numbered from nearest to farthest): 1, 2, 3, ... +1、 ; where node 1 and the last node are numbered These are the beginning and end nodes of the feeder; nodes 2, 3, ... +1 represents each branch node. Node 1 to node... The main line that constitutes this feeder.

[0086] Based on the approximate topology of the feeder, each branch node... p , ( +1) Determine the number of branches and continue numbering the terminal nodes of the branches sequentially: e.g. Figure 2 As shown on the left. This forms the abstract topology of the feeder, and the total number of nodes in the abstract topology of the feeder is denoted as n (including the first, last, and branch nodes of the trunk line and all branch end nodes).

[0087] In the abstract topology diagram of the feeder, nodes 1, 2, ..., n-3 represent trunk nodes, where node 1 represents the beginning of the trunk, node n-3 represents the end node, and nodes n-1, ..., n represent branch node numbers.

[0088] Based on the above rules, the abstract topology is obtained as follows: Figure 3 As shown on the right, the node number of the main line is the first node 1 in its corresponding association matrix C, which must be placed in the network master table (with the sequence number as follows). The measuring point can then be determined. At node The position is denoted as: Because the incidence matrix C reflects the connections between abstract topologies, it is a symmetric matrix, and is expressed as:

[0089] (3)

[0090] Among them, the main diagonal elements All are 0. Nodes representing abstract topology p and q Directly connected, This indicates that the nodes are not directly connected.

[0091] (4)

[0092] All elements with a value of 1 in the incidence matrix C of the abstract topology are placed into set Q.

[0093] (5)

[0094] Understandably, clustering is used to classify the measurement points of the feeder lines in the transformer area, in order to prepare for the next step of determining the location of the cluster points in the abstract topology.

[0095] Step S13: Perform cluster analysis on the Euclidean distances between each measurement point to obtain the Euclidean distance matrix of the cluster points.

[0096] Specifically, the clustering tree matrix of each measuring point is calculated using the sum of squared deviations algorithm and the Euclidean distance between measuring points; the N measuring points are clustered into n classes, and the average Euclidean distance of the measuring point voltage in each cluster is calculated to obtain the average time-series voltage set of all clusters; the cluster Euclidean distance matrix of the n clusters is obtained based on the average time-series voltage set of all clusters.

[0097] Furthermore, the Euclidean distance between each measuring point is calculated using the sum of squared deviations method. Based on the distance, the points are classified and a clustering tree matrix M is generated. The clustering tree matrix M contains the binary leaf node numbers of the generated clustering tree and the minimum distance between the binary leaf nodes.

[0098] The expression for calculating the sum of squared deviations is:

[0099] (16)

[0100] in,

[0101] According to the formula:

[0102] (17)

[0103] Where G1 and G2 are the scores of the principal components of the measurement points, D1 and D2 represent the distances between the measurement points in G1 and G2 and their respective cluster centers, and D12 represents the distance between the two classes to facilitate classification. n1 and n2 are the number of samples in G1 and G2, respectively.

[0104] The clustering result obtained from the magnitude of D is denoted as n.

[0105] The N measuring points of the transformer feeder are divided into n categories, and the average Euclidean distance of the measuring point voltage in each category is calculated. The average time-series voltage set of all measuring points is calculated according to equation (15) based on the measuring point number in each category, and denoted as […]. ,

[0106] (18)

[0107] in l Indicates the first x The measurement point numbers included in the cluster points, Indicates the measurement point number l The time-series voltage set, N p This involves clustering the measurement points into n groups and calculating the average voltage at each cluster. It is a focal point x The average time-series voltage set.

[0108] The Euclidean distance matrix D of the clusters of n types is obtained from the set of average time-series voltages of all clusters.

[0109] Step S14: Determine the specific location of the measurement point in the abstract topology based on the Euclidean distance matrix and the correlation matrix of the cluster point.

[0110] As mentioned above, based on the Euclidean distance between the cluster points The topological correlation matrix C is used to determine the specific location of each measurement point in the abstract topology.

[0111] Specifically, the column number of the element in the correlation matrix that is not equal to zero is searched according to the row number of the measurement point to obtain a first set with a first number of elements; all elements in the Euclidean distance matrix of the cluster points are searched according to the row number of the measurement point to obtain a second set with a first number of elements; based on the first set and the second set, the cluster points in the Euclidean distance matrix of the cluster points are associated with the measurement points in the correlation matrix to obtain the specific location of the measurement points in the abstract topology.

[0112] First, based on the abstract topological association matrix C, abstract topological nodes are generated. p A set of connection relationships.

[0113] From the association matrix C of the abstract topology, according to the nodes p Search for matrix elements based on their row number. We need to find the elements in the correlation matrix C. At that time, element The column numbers are used to form a set P, arranged in ascending order of column number.

[0114] Set P can be represented as:

[0115] (19)

[0116] The number of elements in set P is denoted as . .

[0117] Then, the specific location of the measurement point in the abstract topology is obtained through the Euclidean distance matrix D of the cluster points.

[0118] Establish the relationship between the elements in the Euclidean distance matrix D and the abstract topological incidence matrix C, assuming... That is, the nodes in the correlation matrix C are The nodes and the cluster points in the Euclidean distance matrix D x In connection with this, we need to extract the first element from the Euclidean distance matrix D. All elements of the row ,turn up Euclidean distances in ascending order The elements are arranged in order of their column numbers according to the magnitude of their Euclidean distance, forming a set J.

[0119] Set J is represented as:

[0120] (20)

[0121] in This represents the column index of the z-th smallest element in the x-th row of the Euclidean distance matrix D.

[0122] make That is, finding the cluster point in the Euclidean distance matrix D. x Therefore, the gathering point x Measurement points in the correlation matrix C Then they can be matched up. ,node x The location of the abstract topology is also the measurement point. The location.

[0123] The method further includes: defining a large number; when a cluster point corresponds to a measurement point, replacing the elements in the Euclidean distance matrix of the cluster point with the large number; when the number of elements in the first set is 0, outputting the cluster point number in the Euclidean distance matrix, the node number in the correlation matrix corresponding to the cluster point, and the cluster point number connected to the cluster point, to determine the specific position of the cluster point in the abstract matrix. When the number of elements in the first set is not 0, returning to the step of searching all elements in the Euclidean distance matrix of the cluster point according to the row number of the measurement point to obtain a second set with the number of elements of the first number.

[0124] Specifically, we define a large number K, which is a number larger than any element in the Euclidean distance matrix D, i.e.:

[0125] (twenty one)

[0126] When found When doing this, it is necessary to select the first element in the corresponding Euclidean distance matrix D. x The first line Replace the elements in column n with a large number K, representing the nth element of that node. The branch has already been searched, so we will not traverse the branch again.

[0127] Repeat the above steps, each time ending the above steps. x Increment by 1; until a certain node exists. x Number of elements found The loop operation terminates at a certain time, and the number of loops is h; the cluster points generated in each loop operation x， Gathering point x The connection node number and the node of the abstract topology p The set of connection relationships I J and P are counted as , the loop result Store the data in sets O, M, and N. Represent the sets O, M, and N obtained from each iteration as follows:

[0128] (twenty two)

[0129] (twenty three)

[0130] (twenty four)

[0131] The output set O is the cluster point in the Euclidean distance matrix D. x The numbering, the elements in set M are the cluster points. x The corresponding node numbers in the correlation matrix C, and the set N represents the cluster points. x The cluster number connected to the node.

[0132] Where z and x represent cluster points, which correspond to the Euclidean distance matrix D; p and q represent abstract topological nodes, and i and j represent measurement points;

[0133] For example, according to Figure 3 The feeder topology diagram of a certain transformer area shown is numbered according to the rules in step S13. The main trunk node numbers are 1, 2, ..., 7, and nodes 8 to 42 are branch line nodes. Nodes 1 and 7 are the first and last nodes of the main trunk.

[0134] according to Figure 4 The feeder topology shown can be abstracted into a simplified feeder topology diagram, as follows: Figure 5 As shown, where Figure 4 The test points inside the dashed box can be clustered as follows: Figure 5 In the cluster, node n' is the main trunk node number; nodes 1' to 7' are the first and last nodes of the main trunk, and nodes 8', 9', and 10' are the branch node numbers.

[0135] Depend on Figure 5As can be seen from the abstract topology, the topological correlation matrix C is obtained through step S13.

[0136] The correlation matrix C is:

[0137] (twenty four)

[0138] In this case, all elements on the main diagonal are 0. Represents a node i and j Directly connected, This indicates that the nodes are not directly connected. Since the incidence matrix is ​​a symmetric matrix, we only need to analyze the upper triangular portion of the incidence matrix C. Therefore, we place all elements with a value of 1 in the incidence matrix C of the abstract topology into set Q.

[0139] (25)

[0140] After the operation in step S13, the clustering results of 42 measurement points of the feeder in the transformer area and the average Euclidean distance matrix D of the cluster voltage can be obtained.

[0141] The 42 measuring points of a feeder topology in a certain transformer area can be divided into 10 categories, and its clustering system spanning tree is as follows: Figure 6 The distribution and classification are shown in Table 1:

[0142] Table 1. Clustering of Feeder Topology in Transformer Areas

[0143]

[0144] Meanwhile, the Euclidean distance matrix D corresponding to the cluster nodes is:

[0145]

[0146] Determine the location of the cluster point in the abstract topology.

[0147] According to step S14, from the association matrix C of the abstract topology, according to the nodes... p Search for matrix elements based on their row number. We need to find the elements in the correlation matrix C. At that time, element The column numbers are used to form a set P, arranged in ascending order of column number.

[0148] (1) In the first loop, the non-zero element found at the row number corresponding to node 1 is Therefore, set P is, The number of elements in set P is denoted as ;

[0149] (2) In the second loop, the non-zero element found at the row number corresponding to node 2 is Therefore, set P is, The number of elements in set P is denoted as . .

[0150] (3) In the third loop, the non-zero element found at the row number corresponding to node 3. Therefore, set P is, The number of elements in set P is denoted as . .

[0151] (4) In the fourth iteration, the non-zero element found at the row number corresponding to node 4. Therefore, set P is, The number of elements in set P is denoted as . .

[0152] (5) In the fifth iteration, the non-zero element found is the row number corresponding to node 5. Therefore, set P is, The number of elements in set P is denoted as . .

[0153] (6) In the sixth iteration, the non-zero element found at the row number corresponding to node 6. Therefore, set P is, The number of elements in set P is denoted as . .

[0154] Establish the relationship between the elements in the Euclidean distance matrix D and the abstract topological incidence matrix C, assuming... That is, the nodes in the correlation matrix C are The nodes and the cluster points in the Euclidean distance matrix D x In connection with this, we need to extract the first element from the Euclidean distance matrix D. All elements of the row ,turn up Euclidean distances in ascending order The elements are arranged into a set J according to the column numbers of the Euclidean distance, and the corresponding elements in the clustering Euclidean distance matrix are modified to K, where K is 1000.

[0155] (1) The result obtained in the first loop x =1, set J is J={3}, that is, k(2)=3; at the same time, replace the elements in the 1st to 10th columns of the 3rd row of the Euclidean distance matrix D with 1000.

[0156] (2) The result of the second cycle x=3, set J is J={5,4,2}, that is, k(3)=5, k(8)=4, k(9)=2; replace the elements in the 5th row, 4th row and 2nd row of the Euclidean distance matrix D with 1000 respectively.

[0157] (3) The result of the third cycle x =5, set J is J={7,6}, that is, k(4)=7, k(10)=6; replace the elements in the 7th row and the 3rd to 10th columns of the 6th row of the Euclidean distance matrix D with 1000 respectively.

[0158] (4) The result obtained in the fourth cycle x =7, set J is J={8}, that is, k(5)=8; replace the elements in the 4th to 10th columns of the 8th row of the Euclidean distance matrix D with 1000.

[0159] (5) The result obtained in the fifth cycle x =8, set J is J={9}, that is, k(6)=9; replace the elements in the 5th to 10th columns of the 9th row of the Euclidean distance matrix D with 1000.

[0160] (6) The result obtained in the sixth cycle x =9, set J is J={10}, that is, k(7)=10; replace the elements in the 6th to 7th columns of the 10th row of the Euclidean distance matrix D with 1000 respectively.

[0161] Get the loop operand h;

[0162] (1) The first loop yields h=1;

[0163] (2) The second cycle yields h=3;

[0164] (3) The third cycle yields h=2

[0165] (4) The h=1 obtained in the fourth cycle

[0166] (5) The fifth iteration yields h=1

[0167] (6) The h=1 obtained in the sixth cycle

[0168] Generate sets M, N, and O;

[0169] (1) The set obtained in the first loop is M={1}, N={3}, O={1};

[0170] (2) The set of the second cycle is M={2}, N={5,4,2}, O={3}.

[0171] (3) The set of the third cycle is M={3}, N={7,6}, O={5}.

[0172] (4) The set of the fourth cycle is M={4}, N={8}, O={7}.

[0173] (5) The set of the fifth cycle is M={5}, N={9}, O={8}.

[0174] (6) The set of the sixth cycle is M={6}, N={10}, O={9}.

[0175] Based on sets M, N, and O, determine the specific location of the cluster points in the abstract matrix. Set O contains the set of cluster points, set M contains cluster point elements corresponding to node numbers in the abstract topology, and set N contains elements representing node numbers connected to the cluster points.

[0176] The aforementioned method for identifying the topology of low-voltage distribution transformer feeder branches involves collecting time-series voltage data from N measurement points in the low-voltage feeder topology and calculating the Euclidean distance between each measurement point based on the time-series voltage data. Cluster analysis is then performed on the Euclidean distances between the measurement points to obtain a cluster Euclidean distance matrix. An abstract topology and its corresponding correlation matrix are then determined based on the low-voltage feeder topology. Finally, the specific location of each measurement point within the abstract topology is determined based on the cluster Euclidean distance matrix and the correlation matrix. Compared to existing technologies, this invention can quickly identify the topology of low-voltage distribution transformer feeder branches without increasing the number of line branch terminals, thus meeting practical application requirements.

[0177] It should be understood that although the steps in the flowchart above are shown sequentially as indicated by 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 flowchart above may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0178] Please see Figure 7 The present invention also provides a low-voltage distribution area feeder branch topology identification device, the device comprising:

[0179] The data acquisition module 21 is used to acquire time-series voltage data of N measurement points in the low-voltage feeder topology, and calculate the Euclidean distance between each measurement point based on the time-series voltage data.

[0180] Specifically, a time-series voltage matrix is ​​generated based on the time-series voltage data of N measuring points within the target time period; the time-series voltage matrix is ​​normalized using a normalization algorithm to obtain a set of voltage data for all measuring points at all times; the correlation matrix of each element in the voltage data set, as well as the eigenvector matrix, contribution rate matrix, and eigenvalue matrix of the correlation matrix, are obtained, and principal component data are selected; the Euclidean distance matrix between each measuring point is calculated based on the principal component data.

[0181] The topology determination module 22 is used to determine an abstract topology and the correlation matrix corresponding to the abstract topology through the low-voltage feeder topology.

[0182] Specifically, the number of feeder branch nodes and the number of branches of each branch node are determined by the shape of the low-voltage feeder topology; based on the number of feeder branch nodes and the number of branches of each branch node, the abstract topology diagram and the topology correlation matrix corresponding to the abstract topology are determined.

[0183] Cluster analysis module 23 is used to perform cluster analysis on the Euclidean distance between each measurement point to obtain the Euclidean distance matrix of the cluster points.

[0184] Specifically, the clustering tree matrix of each measuring point is calculated using the sum of squared deviations algorithm and the Euclidean distance between measuring points; the N measuring points are clustered into n classes, and the average Euclidean distance of the measuring point voltage in each cluster is calculated to obtain the average time-series voltage set of all clusters; the cluster Euclidean distance matrix of the n clusters is obtained based on the average time-series voltage set of all clusters.

[0185] The location determination module 24 is used to determine the specific location of the measurement point in the abstract topology based on the Euclidean distance matrix and the correlation matrix of the cluster point.

[0186] Specifically, the column number of the element in the correlation matrix that is not equal to zero is searched according to the row number of the measurement point to obtain a first set with a first number of elements; all elements in the Euclidean distance matrix of the cluster points are searched according to the row number of the measurement point to obtain a second set with a first number of elements; based on the first set and the second set, the cluster points in the Euclidean distance matrix of the cluster points are associated with the measurement points in the correlation matrix to obtain the specific location of the measurement points in the abstract topology.

[0187] Furthermore, a large number is defined. When a cluster point corresponds to a measurement point, the elements in the Euclidean distance matrix of the cluster point are replaced with the large number. When the number of elements in the first set is 0, the cluster point number in the Euclidean distance matrix, the node number in the correlation matrix corresponding to the cluster point, and the cluster point number connected to the cluster point are output to determine the specific position of the cluster point in the abstract matrix. When the number of elements in the first set is not 0, the process returns to the step of searching all elements in the Euclidean distance matrix of the cluster point according to the row number of the measurement point to obtain the second set with the first number of elements.

[0188] The low-voltage distribution substation feeder branch topology identification device provided in this invention collects time-series voltage data from N measurement points of the low-voltage feeder topology and calculates the Euclidean distance between each measurement point based on the time-series voltage data; performs cluster analysis on the Euclidean distances between each measurement point to obtain a cluster Euclidean distance matrix; determines an abstract topology and the corresponding correlation matrix of the abstract topology through the low-voltage feeder topology; and determines the specific location of the measurement point in the abstract topology based on the cluster Euclidean distance matrix and the correlation matrix. Compared with the prior art, this invention can quickly identify the topology of low-voltage distribution substation feeder branches without increasing the number of line branch terminals, meeting practical application requirements.

[0189] This invention also provides a computer-readable storage medium, which includes a stored computer program; wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the low-voltage distribution area feeder branch topology identification method as described above.

[0190] This invention also provides an electrical device, see [link to relevant documentation]. Figure 8 The diagram shown is a structural block diagram of a preferred embodiment of a power device provided by the present invention. The power device includes a processor 10, a memory 20, and a computer program stored in the memory 20 and configured to be executed by the processor 10. When the processor 10 executes the computer program, it implements the low-voltage distribution area feeder branch topology identification method as described above.

[0191] Preferably, the computer program can be divided into one or more modules / units (such as computer program 1, computer program 2, ...), and the one or more modules / units are stored in the memory 20 and executed by the processor 10 to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program in the power equipment.

[0192] The processor 10 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor, or the processor 10 may be any conventional processor. The processor 10 is the control center of the power equipment, and connects various parts of the power equipment using various interfaces and lines.

[0193] The memory 20 mainly includes a program storage area and a data storage area. The program storage area can store the operating system, applications required for at least one function, etc., while the data storage area can store related data, etc. Furthermore, the memory 20 can be a high-speed random access memory, or a non-volatile memory, such as a plug-in hard drive, a smart media card (SMC), a secure digital card (SD), and a flash card, or other volatile solid-state storage devices.

[0194] It should be noted that the aforementioned power equipment may include, but is not limited to, processors and memory, as will be understood by those skilled in the art. Figure 8 The structural block diagram is merely an example of electrical equipment and does not constitute a limitation on the electrical equipment. It may include more or fewer components than shown, or combine certain components, or use different components.

[0195] In summary, the low-voltage distribution substation feeder branch topology identification method, device, storage medium, and power equipment provided by this invention first collect time-series voltage data from N measurement points of the low-voltage feeder topology and calculate the Euclidean distance between each measurement point based on the time-series voltage data; perform cluster analysis on the Euclidean distances between each measurement point to obtain a cluster Euclidean distance matrix; determine an abstract topology and the corresponding correlation matrix of the abstract topology through the low-voltage feeder topology; and determine the specific location of the measurement point in the abstract topology based on the cluster Euclidean distance matrix and the correlation matrix. Compared with the prior art, this invention can quickly identify the topology of low-voltage distribution substation feeder branches without increasing the number of line branch terminals, meeting practical application requirements.

[0196] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for identifying the feeder branch topology in a low-voltage distribution area, characterized in that, The method includes the following steps: Collect time-series voltage data from N measuring points in the low-voltage feeder topology, and calculate the Euclidean distance between each measuring point based on the time-series voltage data; An abstract topology and the corresponding correlation matrix are determined by the low-voltage feeder topology. The process of determining an abstract topology through low-voltage feeder topology, and the corresponding correlation matrix of the abstract topology, includes: The number of feeder branch nodes and the number of branches in each branch node are determined by the shape of the low-voltage feeder topology. Based on the number of feeder branch nodes and the number of branches of each branch node, an abstract topology diagram and the corresponding correlation matrix of the abstract topology are determined. Cluster analysis is performed on the Euclidean distances between each measuring point to obtain the Euclidean distance matrix of the cluster points. This includes: calculating the cluster tree matrix of each measuring point using the sum of squared deviations algorithm and the Euclidean distances between the measuring points; clustering the N measuring points into n classes and calculating the average Euclidean distance of the measuring point voltages in each cluster to obtain the average time-series voltage set of all cluster points; and obtaining the Euclidean distance matrix of the cluster points of the n classes based on the average time-series voltage set of all cluster points. Determining the specific location of a measurement point in the abstract topology based on the Euclidean distance matrix and the incidence matrix includes: searching the column number of an element in the incidence matrix that is not equal to zero according to the row number of the measurement point, to obtain a first set with a first number of elements; searching all elements in the Euclidean distance matrix according to the row number of the measurement point, to obtain a second set with a first number of elements; and associating the cluster points in the Euclidean distance matrix with the measurement points in the incidence matrix based on the first set and the second set, to obtain the specific location of the measurement point in the abstract topology.

2. The method for identifying the feeder branch topology of a low-voltage distribution area according to claim 1, characterized in that, The method for collecting time-series voltage data from N measuring points in a low-voltage feeder topology and calculating the Euclidean distance between the measuring points based on the time-series voltage data includes: Generate a time-series voltage matrix based on the time-series voltage data of N measurement points within the target time period; The time-series voltage matrix is ​​normalized using a normalization algorithm to obtain the voltage data set of all measurement points at all times. Obtain the correlation matrix of each element in the voltage data set, as well as the eigenvector matrix, contribution rate matrix and eigenvalue matrix of the correlation matrix, and filter out the principal component data; Calculate the Euclidean distance matrix between each measurement point based on the principal component data.

3. The method for identifying the feeder branch topology of a low-voltage distribution area according to claim 1, characterized in that, The method further includes: Define a large number, and when the cluster point corresponds to the measurement point, replace the elements in the Euclidean distance moments of the cluster point with the large number; When the number of elements in the first set is 0, output the cluster number in the Euclidean distance matrix, the node number in the correlation matrix corresponding to the cluster, and the cluster number connected to the cluster, to determine the specific position of the cluster in the abstract matrix.

4. The method for identifying the feeder branch topology of a low-voltage distribution area according to claim 3, characterized in that, The method further includes: When the number of elements in the first set is not 0, return to the step of searching all elements in the Euclidean distance matrix of the cluster points according to the row number of the measurement point to obtain the second set with the first number of elements.

5. A low-voltage distribution area feeder branch topology identification device, characterized in that, The device includes: The data acquisition module is used to acquire time-series voltage data from N measuring points in a low-voltage feeder topology, and to calculate the Euclidean distance between each measuring point based on the time-series voltage data. The clustering analysis module is used to perform clustering analysis on the Euclidean distances between various measurement points to obtain the Euclidean distance matrix of the cluster points. This includes: calculating the clustering tree matrix for each measurement point using the sum of squared deviations algorithm and the Euclidean distances between measurement points; clustering N measurement points into n classes and calculating the average Euclidean distance of the measurement point voltages in each cluster to obtain the average time-series voltage set of all cluster points; and obtaining the Euclidean distance matrix of the cluster points for the n classes based on the average time-series voltage set of all cluster points. The topology determination module is used to determine an abstract topology and the corresponding correlation matrix of the abstract topology based on the low-voltage feeder topology. The process of determining an abstract topology through low-voltage feeder topology, and the corresponding correlation matrix of the abstract topology, includes: The number of feeder branch nodes and the number of branches in each branch node are determined by the shape of the low-voltage feeder topology. Based on the number of feeder branch nodes and the number of branches of each branch node, an abstract topology diagram and the corresponding correlation matrix of the abstract topology are determined. The location determination module is used to determine the specific location of the measurement point in the abstract topology based on the Euclidean distance matrix and the incidence matrix of the cluster points; including: searching the column number of the element in the incidence matrix when the element is not equal to zero according to the row number of the measurement point, to obtain a first set with a first number of elements; searching all elements in the Euclidean distance matrix of the cluster points according to the row number of the measurement point, to obtain a second set with a first number of elements; and associating the cluster points in the Euclidean distance matrix with the measurement points in the incidence matrix according to the first set and the second set, to obtain the specific location of the measurement point in the abstract topology.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program; wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform the low-voltage distribution area feeder branch topology identification method as described in any one of claims 1 to 4.

7. An electrical device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the low-voltage substation feeder branch topology identification method as described in any one of claims 1 to 4.

Images