A low-voltage area topology identification method based on current and electric quantity data

Abstract

The application discloses a low-voltage transformer area topology identification method based on current and electric quantity data, relates to the technical field of low-voltage transformer area topology identification, and comprises data preprocessing, a current algorithm module, an electric quantity algorithm module, a topology fusion module, a rehearsal and verification module and an output result module. The application provides the low-voltage transformer area topology identification method based on the current and electric quantity data, and aims at the deficiency of the prior art, and automatically infers the low-voltage transformer area topology connection relationship through the identification method.

Description

Technical Field

[0001] This invention relates to the field of low-voltage distribution area topology identification technology, specifically to a low-voltage distribution area topology identification method based on current and power data. Background Technology

[0002] Electrical topology is the foundation for the normal operation and maintenance of power grids.

[0003] Currently, the identification of low-voltage transformer area topology relationships is mainly divided into three categories:

[0004] (a) Manual identification: 1. Manual on-site inspection; 2. Power outage and disconnection at the low-voltage outgoing line of the distribution transformer, and check whether the user side is powered to confirm whether the user is supplied by the transformer outgoing line, and so on.

[0005] (II) Install hardware identification module: 1. Install signal transmission module and detection module. The transmission module injects special signals (such as voltage pulse signals and current characteristic signals) on the power carrier line on the distribution transformer side. The detection module reads and demodulates the special signals to determine the electrical circuit where the user is located.

[0006] (III) Data calculation and identification: 1. Voltage identification; 2. Current identification; the effective value of the current is approximately equal or the harmonic content of the current is approximately equal; 3. Active power.

[0007] Among these methods, manual on-site inspection is time-consuming, labor-intensive, and inefficient; once the line changes, the inspection must be repeated. Power outages and circuit breakers are prone to causing user dissatisfaction and still require reoperation if the line changes. Special signal identification requires the identification terminal in the transformer area to obtain the special signal for identification, which relies on carrier communication and can only be transmitted within the same transformer area. Voltage identification is not a valid criterion because the voltage drop at nearby points is within the equipment's error range. Current harmonic content and active power are not collected and output by every type of equipment, lacking universality.

[0008] Currently, the identification of low-voltage distribution network topology in transformer substations mainly relies on on-site inspections or the installation of testing equipment, both of which require manual intervention. This is not only tedious and time-consuming, but also cannot achieve automatic identification and updating. There is an urgent need for a method for identifying low-voltage distribution network topology in transformer substations based on existing equipment to solve the above problems. Summary of the Invention

[0009] This invention provides a method for identifying the topology of low-voltage distribution areas based on current and power data. The purpose is to address the shortcomings of existing technologies by automatically inferring the topological connection relationships of low-voltage distribution areas through this identification method.

[0010] To achieve the above objectives, the technical solution of the present invention is as follows:

[0011] A method for identifying the topology of low-voltage distribution areas based on current and power data includes the following steps:

[0012] S11: Obtain minute-by-minute frozen data D11 of current and energy consumption of power distribution equipment LTU and circuit breaker within the time period T11 (calculated backwards from this time).

[0013] S12: Preprocess the frozen minute data D11 of current and energy, and fill in the missing time points. At this time, the data is F11.

[0014] S13:F11 serves as the data source, and in the input current algorithm module, the output topology connection relationship T11 is used.

[0015] S14:F11 serves as the data source, inputting into the power algorithm module and outputting the topology connection relationship T12;

[0016] S15: T11 and T12 are used as parameters and input into the topology fusion module. The topology connection relationship T13 is output and used as the first output topology result.

[0017] S21: Obtain frozen minute data D12 of current and energy consumption of power distribution equipment LTU and circuit breaker within the time period T12 (calculated backwards from this time);

[0018] S22: Preprocess the current and energy minute-by-minute frozen D12 data and fill in the missing time points. At this time, the data is F12.

[0019] S23: The topology connection relationship T13 and data F12 are used as input parameters to the exercise and verification module. The output parameters are the current algorithm module result, the power algorithm module result parameter, the weight adjustment coefficient w21, and the topology result for this exercise.

[0020] S31:w21 is used as an adjustment coefficient and is input into the topology fusion module as an input parameter for the next algorithm and the exercise verification module.

[0021] S41: Repeat steps S21-S31 and output the results of the topology simulation.

[0022] Preferably, in step S11, the frozen data D11 includes the device type code, device address, terminal address, freezing time, current values ​​of the three phases A / B / C, total positive active power, and total reverse active power.

[0023] Preferably, in step S12, abnormal data is deleted during data preprocessing.

[0024] Preferably, in step S13, the current algorithm module compares the pairwise current data of each node. During the comparison, useless data is deleted, and the correlation between each pairwise node is determined based on whether there is an intersection relationship in the effective data. The similarity between each pairwise node is determined based on the similarity of the fluctuation trends of the effective data. By combining the similarity values ​​of the three-phase currents A, B, and C of each node and their current values, all possible parent nodes of the child node are determined and sorted according to the similarity values ​​from largest to smallest. At the same time, according to Kirchhoff's first law, abnormal possible parent nodes can be eliminated. Based on the consistency of current fluctuations, the "one-to-one" parent-child nodes can be determined. The relationship between nodes is determined; then, based on the similarity values ​​between child nodes and possible parent nodes, the probability that each possible parent node is the true parent node is determined; the probability of all parent-child node relationships is converted into distance relationships between parent and child nodes, with the distance value being |D[i]|, where D[i] is a vector pointing from the parent node to the child node. If there is no distance between nodes, then |D[i]| is ∞; the child node IDs correspond to numbers v0-vk, and S is the set of endpoints of the shortest path starting from v0, i∈0-k; the most likely parent-child node relationship is calculated by finding the minimum sum of distances between each node; the calculation method is: distance D[j] = Min The path starting at node v0 is the shortest path (v0, vj). If the destination is vk, this path is either (v0, vk) or (v0, vj, vk). Its distance is the distance from v0 to vk, or the sum of the distances from v0 to vj and from vj to vk. The length of the next shortest path must be D[m] = Min{D[i]|vi∈Sv0-vk}, where D[j] is the distance from v0 to vm, or the distance from v0 to vm, or the distance from vj to vk. If we use ar cs represents the distance relationship between parent and child nodes, and arcs[i][j] represents the value at distance <vi,vj>. If <vi,vj> does not exist, then arcs[i][j] is set to ∞. Then, the initial value of the shortest path length that can be reached from v0 to each other node vi on the graph is D[i] = arcs[LocateVertex(G,v0)][i], where G represents the number of nodes, v0 is the starting node, and vi∈Sv0-vk. Choose vj such that D[j] = Min{D[i]|vi∈Sv0-vk}. vj is the endpoint of the shortest path from v0 that has been found.

[0025] Let \(S_{v0 - vk}=S_{v0 - vk}\cup\{vj\}\), and modify the shortest path length from \(v0\) to any vertex \(vk\) reachable in the set \(S_{v0 - vk}\); if \(D[j]+arcs[j][k]<D[k]\), then modify \(D[k]\) to \(D[k]=D[j]+arcs[j][k]\) ②; repeat operations ① and ② for \(n - 1\) times, thus obtaining the shortest path sequence from \(v0\) to the remaining nodes in increasing order of path length; sort according to the minimum distance between the child node and the parent node to determine the one-to-one correspondence between the parent and child nodes.

[0026] Preferably, in step S13, when determining the similarity between each pair of nodes, the similarity can be determined according to the value range of the k value or, taking the current data of the nodes at different times as vectors, judge the similarity between two nodes by comparing the proximity of the vector angles between the two nodes. The k value refers to the quotient of the covariance and standard deviation of the current data of two nodes. When k is greater than 0 and less than 1, it indicates that the two nodes are positively correlated; when k is greater than -1 and less than 0, it indicates that the two nodes are negatively correlated; when k = 1, it indicates that the two nodes are completely positively correlated; when k = -1, it indicates that the two nodes are completely negatively correlated; when k = 0, the two nodes are not correlated.

[0027] Preferably, in step S14, the power quantity algorithm module compares the power quantity data of each node, deletes the useless data, and calculates the difference to obtain the power quantity increment. According to whether there is a cross relationship in the increment data, determine whether there is a correlation between each pair of nodes; calculate the similarity according to the power quantity increment, convert the similarity between nodes into a list of possible child nodes corresponding to the parent node, and traverse by adding the power quantity increment to delete the unqualified child nodes to determine the parent-child node relationship.

[0028] Preferably, in step S14, the similarity calculation according to the power quantity increment adopts the similarity calculation method in step S13.

[0029] The beneficial effects of a low-voltage distribution network topology identification method based on current and power quantity data in the present invention:

[0030] 1. Compared with the traditional topology identification method, using the low-voltage distribution network topology identification method based on current and power quantity data can greatly reduce the manual and subsequent maintenance work;

[0031] 2. Compared with the topology identification scheme that increases the hardware cost, the low-voltage distribution network topology identification method based on current and power quantity data greatly reduces the equipment cost investment;

[0032] 3. Compared with the current harmonic and active power identification methods, the low-voltage distribution network topology identification method based on current and power quantity data can effectively solve the promotion and adaptation problems. At the same time, the identification rate based on the ratio of current and power quantity is more accurate and stable than that of single attributes. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 , one Design framework diagram of a low-voltage distribution area topology identification method based on current and power data; Detailed Implementation

[0034] The following description provides a detailed explanation of the embodiments of the present invention in a step-by-step manner. This description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0035] In the description of this invention, it should be noted that the terms "upper," "lower," "left," "right," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or a specific orientational structure and operation. Therefore, they should not be construed as limiting this invention.

[0036] To better understand this invention, the following description is provided in conjunction with the accompanying drawings. Figure 1 The present invention, “A method for identifying the topology of low-voltage distribution areas based on current and power data,” will be described in further detail.

[0037] (1) Freeze current and energy data by minute; (2) Data preprocessing;

[0038] (3) Algorithm - Current Algorithm Module; (4) Algorithm - Energy Algorithm Module;

[0039] (5) Topology fusion; (6) Drills and verification;

[0040] (7) Output results.

[0041] (i) Obtain minute-by-minute frozen data D11(1) of the current and energy of the distribution equipment LTU and circuit breaker within the time period T11. The data includes equipment type code, equipment address, terminal address, freezing time, current value (including three phases A, B and C), total positive active energy, and total reverse active energy.

[0042] (ii) Data D11 undergoes preprocessing (2) to remove abnormal data and fill in defective data. At this time, the data is counted as F11.

[0043] (III) Data F11 is used as the data source and input into the current algorithm module (3). The current algorithm module (3) compares the pairwise current data of each node. During the comparison, useless data is deleted. Based on whether there is a cross relationship between the effective data, it is determined whether there is a correlation between each pairwise node. Based on the similarity of the fluctuation trend of the effective data, the similarity between each pairwise node is determined. The similarity can be determined according to the range of values ​​of k (the quotient of the covariance and standard deviation of the current data of two nodes). (When k is greater than 0 and less than 1, it indicates that the two nodes are positively correlated. When k is greater than -1 and less than 0, it indicates that the two nodes are positively correlated.) The relationship between two nodes is indicated by k=1 (completely positive correlation), k=-1 (completely negative correlation), and k=0 (no correlation). Alternatively, the similarity between two nodes can be determined by comparing the similarity of their vector angles (using current data at different times as vectors). By combining the similarity values ​​of the three-phase currents (A, B, and C) of each node with their magnitudes, all possible parent nodes of the child node can be determined and sorted from largest to smallest similarity value. Furthermore, Kirchhoff's first law can be used to eliminate abnormal parent nodes. Based on the consistency of current fluctuations... The parent-child node relationship can be determined; then, based on the similarity value between the child node and the possible parent node, the probability that each possible parent node is the real parent node is determined; the probability of all parent-child node relationships is converted into the distance relationship between the parent and child nodes, with the distance value being |D[i]| (D[i] is a vector), and the direction is from the parent node to the child node. If there is no distance value between nodes, then |D[i]| is ∞; the child node IDs correspond to the numbers v0-vk respectively, and S is the set of the endpoints of the shortest path starting from v0, i∈0-k; by calculating the minimum distance between each node, Calculate the most likely parent-child node relationship; the calculation method is: distance is D[j] = Min{D[i]|vi∈Sv0-vk}, the path starting from node v0 is the shortest path starting from v0, this path is (v0,vj). If the destination is vk, this path is either (v0,vk) or (v0,vj,vk), its distance is the distance from v0 to vk, or the sum of the distances from v0 to vj and from vj to vk. Then the length of the next shortest path must be D[m] = Min{D[i]|vi∈Sv0-vk}.

[0044] Sv0 - vk}, where D[j] is the distance value from v0 to vm, or the distance value of (v0, vm, vk). If arcs is used to represent the distance relationship between parent and child nodes, arcs[i][j] represents the value on the distance <vi, vj>; if <vi, vj> does not exist, arcs[i][j] is set to ∞. Then, the initial value of the possible shortest path length from v0 to the remaining nodes vi on the graph is D[i] = arcs[LocateVertex(G, v0)][i], G represents the number of nodes, v0 is the starting node, vi ∈ Sv0 - vk. Select vj such that D[j] = Min{D[i]|vi ∈ Sv0 - vk}, vj is the end point of a shortest path starting from v0 obtained currently ①;

[0045] Let Sv0 - vk = Sv0 - vk ∪ {vj}, and modify the shortest path length from v0 to any vertex vk reachable in the set Sv0 - vk. If D[j] + arcs[j][k] < D[k], then modify D[k] to D[k] = D[j] + arcs[j][k] ②; Repeat operations ① and ② a total of n - 1 times, and thus obtain the shortest path sequence from v0 to the remaining nodes in increasing order of path length. For example, the specific input parameters are as follows:

[0046] Input the number of nodes and the number of direct connections between nodes (input format: number of nodes, number of connections): 6, 8

[0047] Input the corresponding numbers of the nodes (input format: corresponding numbers of the nodes): 0 1 2 3 4 5

[0049] Input the corresponding numbers of two adjacent nodes and the distance value, with the direction from the parent node to the child node (input format: i j w): <0OO00105>0 5 100 4 5 60 0 4 30 0 2 10 1 2 5 2 3 50 3 5 10 4 3 20

[0058] Output the result of the shortest distance matrix: 0000 9999 0010 0050 0030 0060 9999 0000 0005 0055 9999 0065 9999 9999 0000 0050 9999 0060 9999 9999 9999 0000 9999 0010 9999 9999 9999 0020 0000 0030 9999 9999 9999 9999 0000

[0065] Sort by the minimum distance between child nodes and parent nodes to determine the one-to-one correspondence between parent and child nodes.

[0066] (iv) Data F11 is used as the data source and input into the power algorithm module (4). The power algorithm module (4) compares the power data of each node, deletes useless data, and calculates the difference to obtain the power increment. Based on whether there is a cross relationship between the incremental data, it determines whether there is a correlation between each pair of nodes; based on the similarity of the power increment (the similarity calculation method is similar to that of current), the similarity between nodes is transformed into a list of possible child nodes corresponding to the parent node, and by summing and traversing the power increment, child nodes that do not meet the requirements are deleted to determine the parent-child node relationship.

[0067] (v) The current algorithm module and the power algorithm module merge the topology results of each algorithm module into a single topology result based on the weighting coefficients and modified parameters.

[0068] (vi) If there are prediction results, the prediction results can be used as input, combined with current and power data for simulation and verification, and output the current algorithm module result, power algorithm module result parameters, weight adjustment coefficients, and the topology result for this time; otherwise, no output is output, and only the default parameters and coefficients are used.

[0069] (vii) Output the final topology result.

[0070] Working principle of the invention:

[0071] To familiarize you with the working principle of this invention, the working principle of this invention will be described in detail below.

[0072] The intelligent terminal in the distribution area initiates a self-organizing network through centralized scheduling. After the network is established, it reads and obtains the equipment type, current, power consumption, and other minute-by-minute frozen data of the collection terminals at each monitoring point through low-voltage power broadband carrier or low-power wireless carrier communication, and stores them in the data center. It then obtains the equipment type, current, power consumption, and other minute-by-minute frozen data from the data center, and realizes the automatic identification of physical topology relationships based on the low-voltage distribution area topology identification method based on the current and power consumption data.

Claims

1. A low-voltage area topology identification method based on current and power data, characterized by comprising the following steps: S11: Obtain current and power minute frozen data D11 of distribution equipment LTU and circuit breakers in time period T11; S12: Preprocess the current and power minute frozen data D11 and fill in the missing time points, at which time the data is F11; S13: Use F11 as the data source and input it into the current algorithm module to output the topology connection relationship T11; S14: Use F11 as the data source and input it into the power algorithm module to output the topology connection relationship T12; S15: Use T11 and T12 as parameters and input them into the topology fusion module to output the topology connection relationship T13 and use it as the first output topology result; S21: Obtain current and power minute frozen data D12 of distribution equipment LTU and circuit breakers in time period T12; S22: Preprocess the current and power minute frozen data D12 and fill in the missing time points, at which time the data is F12; S23: Use the topology connection relationship T13 and the data F12 as input parameters and input them into the rehearsal and verification module to output the current algorithm module result, the power algorithm module result parameter, the weight adjustment coefficient w21, and the topology result of this time; S31: Use w21 as the adjustment coefficient and input it into the topology fusion module as the input parameter of the next algorithm and the rehearsal and verification module; S41: Repeat the steps S21-S31 to output the topology rehearsal result after that; In step S13, the current algorithm module compares the current data of each node with each other, deletes useless data in the comparison process, determines whether there is a correlation between each node with each other according to whether there is a cross relationship of the effective data, determines the similarity between each node with each other according to the similarity degree of the fluctuation trend of the effective data, determines all possible parent nodes of the sub node by comprehensively considering the similarity value of the three-phase current of each node A, B and C and the current value, and sorts the possible parent nodes according to the similarity value from large to small; meanwhile, according to the first law of Kirchhoff, the possible parent nodes can be excluded; according to the consistency of current fluctuation, the parent-child node relationship of "one belt one" can be determined; then, according to the similarity value of the sub node and the possible parent node, the possibility probability of each possible parent node being the real parent node is judged; the possibility probability of all parent-child node relationships is converted into the distance relationship of the parent-child node, the distance value is |D[i]|, D[i] is a vector, the direction is from the parent node to the sub node, if there is no distance value between the nodes, |D[i]| is ∞; the sub node ID corresponds to the number v0-vk respectively, S is the set of the terminal point of the shortest path from v0, i∈0-k; the most possible parent-child node relationship is calculated by calculating the minimum distance between each node; the calculation method is: the distance is D[j]=Min{D[i]|vi∈Sv0-vk}, the path of the starting node v0 is the shortest path from v0, this path is (v0, vj), if the terminal point is vk, this path is (v0, vk) or (v0, vj, vk), its distance is the distance value from v0 to vk or the sum of the distance value from v0 to vj and the distance value from vj to vk, the length of the next shortest path must be D[m]=Min{D[i]|vi∈Sv0-vk}, wherein, D[j] is the distance value of v0 to vm or the distance value of (v0, vm, vk); if arcs represents the parent-child node distance relationship, arcs[i][j] represents the value on the distance ; if does not exist, arcs[i][j] is set to ∞, then the initial value of the shortest path length from v0 to the remaining nodes vi on the graph is D[i] = arcs[LocateVertex(G, v0)][i], G represents the number of nodes, v0 is the starting node, and vi ∈ Sv0-vk is selected so that D[j] = Min{D[i] | vi ∈ Sv0-vk} vj is the terminal point of the currently obtained shortest path from v0 ①; Sv0-vk = Sv0-vk ∪ {vj} is set, and the shortest path length from v0 to any vertex vk in the set Sv0-vk is modified; if D[j] + arcs[j][k] < D[k], then D[k] is modified to D[k] = D[j] + arcs[j][k] ②; The operations ① and ② are repeated for n-1 times, and thus the shortest path sequence from v0 to the remaining nodes in the order of increasing path length is obtained; the one-to-one parent-child node correspondence relationship is determined according to the minimum distance sorting of the child nodes and the parent nodes.

2. The low-voltage area topology identification method based on current and electric quantity data according to claim 1, characterized in that: In the step S11, the frozen data D11 includes the equipment type code, the equipment address, the terminal address, the freezing time, the current values of A / B / C three-phase, the forward active total power, and the reverse active total power.

3. The low-voltage transformer area topology identification method based on current and electric quantity data according to claim 2, characterized in that: The step S12, the data preprocessing is to delete the abnormal data.

4. The low-voltage area topology identification method based on current and electric quantity data according to claim 3, characterized in that: In the step S13, when determining the similarity between each node, the similarity can be determined according to the value range of k value or by taking the current data of the nodes at different times as vectors, and the similarity between the two nodes is determined by comparing the closeness of the vector angle between the two nodes. The k value refers to the quotient of the covariance and the standard deviation of the current data of the two nodes. When k is greater than 0 and less than 1, it indicates that the two nodes are positively correlated. When k is greater than -1 and less than 0, it indicates that the two nodes are negatively correlated. When k = 1, it indicates that the two nodes are completely positively correlated. When k = -1, it indicates that the two nodes are completely negatively correlated. When k = 0, the two nodes are not correlated.

5. The low-voltage area topology identification method based on current and electric quantity data according to claim 4, characterized in that: In the step S14, the power algorithm module compares the power data of each node, deletes the useless data, and obtains the power increment by difference. According to whether the increment data exists the cross relationship, it is determined whether the correlation exists between each node. According to the similarity calculation of the power increment, the possible child node list corresponding to the parent node is converted from the similarity between the nodes, and the incoherent child nodes are deleted by adding and traversing the power increment to determine the parent-child node relationship.

6. The low-voltage area topology identification method based on current and electric quantity data according to claim 5, characterized in that: In the step S14, the similarity calculation method in the step S13 is used for the similarity calculation of the power increment.

Images