Enhanced low-voltage topology identification method based on multidimensional data combination

Abstract

The invention discloses an enhanced low-voltage topology identification method based on multi-dimensional data combination, and the method comprises the following steps: collecting the physical wiring information of a terminal, and obtaining a wiring database; acquiring terminal position information, and obtaining a community database by using a community division algorithm; collecting electricity utilization characteristic data, and generating an algorithm result database according to the electricity utilization characteristic data by using a topology recognition algorithm; fitting the wiring database, the community database and the algorithm result database to obtain topological data of the current transformer area; and drawing a topological graph of the current transformer area by using the topological data of the current transformer area obtained in the step 2. According to the invention, no unidentified terminal appears in the obtained topological graph, and all terminals are placed at reasonable positions of the topological graph. The virtual nodes obtained through calculation and fitting disappear and are replaced by the actual power cabinet. And each terminal is clearly affiliated to the corresponding power cabinet, so that convenience is provided for finding problems. And the wiring data is associated, so that the drawn topological graph is more vivid and real.

Description

Technical Field

[0001] The invention relates to an enhanced low-voltage topology identification method based on the combination of multi-dimensional data, and belongs to the technical field of low-voltage topology. Background Art

[0002] At present, the topology recognition algorithm of the substation is generally to generate the topology by collecting the electrical signals on the wires using a specific algorithm. It is essentially a dynamic detection method. Figure 1 Due to factors such as the strength of the electrical signal, the accuracy of the acquisition equipment, and the feature calculation method, the topology map calculated by this method often deviates significantly from the actual topology.

[0003] Therefore, a new enhanced low-voltage topology identification method based on the combination of multi-dimensional data is needed to solve the above problems. Summary of the invention

[0004] Purpose of the invention: In view of the problems existing in the prior art, the present invention provides an enhanced low-voltage topology identification method based on the combination of multi-dimensional data.

[0005] An enhanced low-voltage topology identification method based on multi-dimensional data combination includes the following steps:

[0006] Step 1: Collect the physical wiring information of the terminal to obtain a wiring database; collect the terminal location information and use the community division algorithm to obtain a community database; collect power consumption feature data and use the topology recognition algorithm to generate an algorithm result database based on the power consumption feature data;

[0007] Step 2: Fit the connection database, the community database and the algorithm result database to obtain the topological data of the current substation area;

[0008] Step 3: Use the topological data of the current area obtained in step 2 to draw a topological map of the current area.

[0009] Furthermore, the data in the wiring database includes terminal number, wiring type and cabinet number, the data in the community database includes terminal number and community number, and the data in the algorithm result database includes terminal number, terminal type and superior number.

[0010] Furthermore, the wiring types include outlet cabinets, branch boxes and meter boxes, and the terminal types include transformers, general nodes and virtual nodes.

[0011] Furthermore, in the terminal number, 0 represents virtual node 0, and 65535 represents a transformer.

[0012] Furthermore, fitting the connection database, the community database and the algorithm result database in step 2 includes the following steps:

[0013] Step 21: Create a topology result table.

[0014] Step 22, processing unidentified terminals: finding the wiring neighbors of the unidentified terminals through the wiring database, finding the community neighbors of the unidentified terminals through the community database, taking the intersection of the wiring neighbors and the community neighbors to obtain the real first neighbor set, counting the upper terminals of the first neighbor set according to the algorithm result database, taking the upper terminal with the largest number of appearances as the upper terminal of the unidentified terminal, and adding the result to the topology result table;

[0015] Step 23, mapping the non-zero virtual node to the real terminal: searching the algorithm result database to obtain the non-virtual subordinate terminal and the non-virtual topological neighbor, taking the intersection of the wiring database and the non-virtual subordinate terminal to obtain the first wiring type of the non-virtual subordinate terminal, subtracting 1 from the first wiring type to obtain the second wiring type of the virtual node, then searching the wiring database for the non-virtual wiring neighbor of the second wiring type, and finally taking the intersection of the non-virtual wiring neighbor and the non-virtual topological neighbor to obtain the real second neighbor set, taking the terminal with the smallest number in the second neighbor set as the real terminal of the non-zero virtual node, and adding the result to the topological result table;

[0016] Step 24, processing the first terminal that is not in the algorithm result database: obtain the first wiring neighbor of the first terminal through the wiring database, and then search the upper terminal of the first wiring neighbor in the algorithm result database, and take the upper terminal with the largest number of appearances as the upper terminal of the first terminal; if the community number of the first terminal cannot be found from the community database, then search the community number of the first wiring neighbor in the community database, take the community number with the largest number of appearances as the community number of the first terminal, and add the result to the topology result table to obtain the topology data of the current substation.

[0017] Furthermore, the topology result table includes all data of the wiring database, the community database and the algorithm result database as well as data status attributes, and the data status attributes are used to indicate whether the data is fitted.

[0018] Furthermore, in step 22, when the wiring neighbor of the unidentified terminal is not found through the wiring database, the wiring type and cabinet number of the community neighbor are found by using the wiring database; when the community neighbor of the unidentified terminal is not found through the community database, the community number of the wiring neighbor is found by using the community database.

[0019] Beneficial effects: The enhanced low-voltage topology identification method based on multi-dimensional data of the present invention has no unidentified terminals in the obtained topology map, and all terminals are placed in reasonable positions in the topology map. The virtual nodes calculated and fitted disappear and are replaced by actual power cabinets. Each terminal is clearly attributed to the corresponding power cabinet, which facilitates the search for problems. The wiring data is associated, and the drawn topology map is more vivid and realistic. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flow chart of a method for topological identification of a substation in the prior art;

[0021] Figure 2 It is a flow chart of the enhanced low voltage topology identification method based on the combination of multi-dimensional data of the present invention;

[0022] Figure 3 This is the result diagram of the community database in Example 1;

[0023] Figure 4 This is the result diagram of the wiring database of Example 1;

[0024] Figure 5 This is a result diagram of the algorithm result database of Example 1;

[0025] Figure 6 A flowchart for processing where both wiring neighbors and group neighbors exist;

[0026] Figure 7 This is the processing flow chart for the presence of only community neighbors;

[0027] Figure 8 It is a processing flow chart for the existence of only wiring neighbors;

[0028] Fig. 9 It is a flowchart for mapping non-zero virtual nodes;

[0029] Fig.10 Flowchart for handling terminals not included in the algorithm result;

[0030] Fig.11 This is the topological diagram obtained in Example 1. DETAILED DESCRIPTION

[0031] The preferred embodiments of the present invention will be described below in conjunction with the accompanying drawings to more clearly and completely illustrate the technical solutions of the present invention.

[0032] See also Figure 2 As shown, the enhanced low-voltage topology identification method based on the combination of multi-dimensional data of the present invention first establishes a community database, a wiring database and an algorithm result database, and then fits the data of multiple dimensions to finally obtain a topology map that is more in line with the current actual substation.

[0033] The core of the present invention is to fit multi-dimensional data using a specific algorithm, and finally obtain a topological map that is more in line with the current actual station area.

[0034] Embodiment 1:

[0035] The enhanced low-voltage topology identification method based on the combination of multi-dimensional data of the present invention comprises the following steps:

[0036] Assume that there are 10 collection terminals in the current area.

[0037] Step 1: Create a community database:

[0038] Objective: Divide the entire collection terminal set in the area into different collection terminal communities.

[0039] Method: Not limited to Beidou, GLONASS, GPS, Galileo and other positioning systems, nor limited to geographic information systems such as enterprise GIS, nor limited to wireless base stations, Bluetooth, WIFI, Zigbee and other wireless communication positioning methods. There are many implementations of aggregation algorithms for various positioning methods, such as; Euclidean distance, K nearest neighbor search and intra-voxel nearest neighbor search. It is not within the scope of this patent, so it will not be introduced in detail. The following lists only two implementation ideas:

[0040] (a) Using Beidou Positioning System

[0041] Each acquisition terminal has a built-in Beidou positioning chip (or when installing and deploying the acquisition terminal, the current Beidou positioning information is obtained through other devices and then stored in the acquisition terminal), and the location information is sent periodically (such as daily freezing) to the topology calculation server (not limited to concentrators, intelligent fusion terminals, SCU, master stations and other devices or servers). The topology calculation server uses a specific algorithm based on the coordinate information (longitude, latitude, altitude) of each acquisition terminal to divide the entire collection of acquisition terminals into different collection terminal communities. The specific steps are as follows:

[0042] Step 1: The server collects the coordinate information (longitude, latitude, and altitude) of all terminals in the area and obtains the coordinate set of the terminals in the area.

[0043] Step 2: Randomly select a terminal from the area terminal coordinate set and put it into the calculation set, then calculate the Euclidean distance with all remaining terminals and record them one by one.

[0044] Step 4: Repeat step 2 until the area terminal coordinate set is empty.

[0045] Step 5: Randomly select a terminal from the calculation set and put it into the cluster 1 set, then check the Euclidean distance between the terminal and other terminals, and put the terminals whose Euclidean distance is less than a threshold (such as 1 meter) into the cluster 1 set.

[0046] Step 6: Repeat step 5 until all the terminals in the calculation set are divided into the corresponding community sets.

[0047] (b) Using Bluetooth positioning

[0048] Each collection terminal has a built-in Bluetooth module. Each collection terminal has its own Bluetooth name. Each collection terminal can scan and discover other collection terminals and establish communication with them. The collection terminal calculates the relative position with other collection terminals through a specific algorithm, and then sends the information to the topology calculation server. The topology calculation server uses a specific algorithm based on the relative position information of each collection terminal with other collection terminals to divide the entire collection terminal set into different collection terminal communities. The specific steps are as follows:

[0049] Step 1: The collection terminal uses the Bluetooth ranging algorithm to calculate the distance to all surrounding collection terminals and send the information to the server.

[0050] Step 2: The server collects the distance information of all terminals and adjacent terminals in the station area, and obtains the station area terminal distance calculation set.

[0051] Step 3: Randomly select a terminal from the calculation set and put it into the cluster 1 set, then check the distance between the terminal and other terminals, and put the terminals whose distance is less than a threshold (for example, 1 meter) into the cluster 1 set.

[0052] Step 4: Repeat step 3 until all the terminals in the calculation set are divided into the corresponding community sets.

[0053] Result: The 9 collection terminals in the area were divided into four groups, see Figure 3 .

[0054] Step 2: Create a wiring database:

[0055] Objective: Obtain the physical wiring information of the acquisition terminal.

[0056] Method: It is not limited to the on-site engineering implementation personnel to register the acquisition terminal wiring information. Anyone who can provide real and effective acquisition terminal wiring information can register. Although it is not necessary to conduct a second verification and confirmation of the registration information, it is a highly recommended practice. The way to input the registration information into the topology calculation server is not limited to the master station sending it to the topology calculation server, the topology calculation server background import, using other devices to store it in the acquisition terminal and then reporting it to the topology calculation server by the acquisition terminal.

[0057] Result: Assuming that Figure 4 Wiring data shown.

[0058] Step 3: Establish the algorithm result database:

[0059] Objective: Use a topology recognition algorithm to generate a database of algorithm results.

[0060] Method: It is not limited to any specific algorithm implementation for generating topology based on power consumption feature data, as long as the topology can be generated based on power consumption feature information. Of course, the faster the topology generated by the topology recognition library converges to the real topology and the more consistent it is with the real topology, the better it is. The specific implementation of the topology recognition library includes the Pearson correlation coefficient method and the iterative principal component analysis method, which are not within the scope of this patent and will not be introduced in detail.

[0061] Results: According to the inherent characteristics of the power system, the topological structure under the substation must be a directed tree structure. Assume that the output of the topological algorithm is as follows, where the virtual node represents the bifurcation point that the algorithm believes should exist here. Corresponding to the actual scene, it may be a distribution cabinet, branch box, and hinge point. The unidentified terminal under the virtual node 0 is the collection terminal whose location cannot be determined by the algorithm.

[0062] Step 4: Fitting multidimensional data:

[0063] Objective: Use the above three sets of data and a specific algorithm to fit and obtain the topological data of the current substation.

[0064] Table 1. Community database table

[0065] Terminal ID 1 2 3 4 5 6 7 8 9 tribeId 1 3 2 3 4 3 2 3 4

[0066] Table 2. Wiring database table

[0067] Terminal ID 1 2 3 4 5 6 7 9 10 Connection type linkType 1 3 2 3 3 3 2 3 3 Box ID boxId 1 1 1 1 2 1 1 2 2

[0068] Table 3. Algorithm result database

[0069] Terminal ID 1 2 3 4 5 6 7 8 9 65535 0 11 Terminal type 1 1 1 1 1 1 1 1 1 0 2 2 UpId 65535 0 11 3 7 3 11 0 7 null 65535 65535

[0070] a. Terminal number: a unique identifier, where 0 refers to virtual node 0 and 65535 refers to the transformer.

[0071] b. Terminal type: 0 for transformer, 1 for general node, 2 for virtual node

[0072] c. Wiring type: 1 indicates outlet cabinet, 2 indicates branch box, 3 indicates meter box

[0073] Fitting method: Its implementation includes the following steps (before processing starts, first create T_RESULT (result table):

[0074] The T_RESULT result table is created according to the following principles:

[0075] 1. All data including three sets of data.

[0076] 2. Add data status attributes to indicate whether the data is fitted.

[0077] The T_RESULT (result table) with the following content is obtained:

[0078] Table 4. Initial T_RESULT result table obtained based on known data

[0079]

[0080] Fitting step 1: Handling unrecognized terminals

[0081] Objective: To connect the unidentified terminal to the correct topological node so that the unidentified terminal is no longer unidentified.

[0082] Method: Find the wiring neighbors of the unidentified terminal through the wiring data, find the community neighbors of the unidentified terminal through the community data, and then take the intersection of the wiring neighbors and the community neighbors to get the real neighbor set, and then count the upper terminals of the neighbors, and take the upper terminal with the most appearances as the upper terminal of the unidentified terminal. The flow chart is as follows: Figure 6 , Figure 7 and Figure 8 As shown, Figure 6 This is a process where both wired neighbors and community neighbors exist. Figure 7 This is the processing flow when only community neighbors exist. Figure 8 This is the processing flow for the existence of only wired neighbors.

[0083] Wiring Neighbors: Terminals with the same wiring type and cabinet number in the wiring database are called wiring neighbors.

[0084] Community neighbors: Terminals with the same community number in the community database are called community neighbors.

[0085] Neighbor: A terminal that is both a connection neighbor and a community neighbor is called a neighbor.

[0086] Superior terminal: In the topology of the substation, the terminal that is directly connected to this terminal and closer to the transformer substation is called the superior terminal.

[0087] After processing, T_RESULT (result table) is updated as follows:

[0088] Table 5. T_RESULT result table obtained after processing unrecognized terminals

[0089]

[0090] You can see:

[0091] a. The upper terminal of terminal 2 is fitted as terminal 3.

[0092] b. The upper terminal of terminal 8 is fitted as terminal 3, the wiring type is fitted as 3, and the cabinet number is fitted as 1.

[0093] Fitting step 2: Mapping non-zero virtual nodes

[0094] Goal: Try to map non-zero virtual nodes to real terminals. Note that not all non-zero virtual nodes can be mapped to real terminals.

[0095] Method: Find non-virtual lower-level terminals and non-virtual topological neighbors through algorithm data. Then take the intersection of wiring data and lower-level terminals to get the wiring type of the lower-level terminal, and then subtract 1 to get the wiring type of the virtual node. Then find the non-virtual wiring neighbors of this wiring type in the wiring data. Finally, take the intersection of wiring neighbors and topological neighbors to get the real neighbor set, and take the terminal with the smallest id in the set as the real terminal of the "non-zero virtual node". The flowchart is as follows Fig. 9 shown.

[0096] The order of boxes and cabinets within a substation is generally: 1 outgoing line cabinet -> 2 branch box -> 3 meter box. Therefore, the wiring type of the lower terminal is reduced by 1, and the wiring type of this terminal is obtained.

[0097] Topological neighbors: Terminals with the same parent in the algorithm database are called topological neighbors.

[0098] Subordinate terminal: The terminal under this terminal in the algorithm database is called the subordinate terminal.

[0099] After processing, T_RESULT (result table) is as follows:

[0100] Table 6. T_RESULT result table obtained by mapping non-zero virtual nodes

[0101]

[0102] You can see:

[0103] a. Virtual node terminal 11 is mapped to real terminal 1, and the record of terminal 11 is cleared from the result table.

[0104] b. Terminal 7 and the upper terminal of terminal 3 are fitted into terminal 1.

[0105] Fitting step 3: Processing terminals not included in the algorithm results

[0106] Goal: Add the terminals that are not in the algorithm results to the topology of the station. Note that not all terminals that are not in the algorithm results can be added to the topology of the station.

[0107] Method: Get the wiring neighbors through the wiring data. Then find the upper terminals of these wiring neighbors in the algorithm data, and take the upper terminal with the most occurrences as the upper terminal of the terminal. Then, if the community number of the terminal cannot be found from the community data, then find the community number of the wiring neighbors in the community data, and take the community number with the most occurrences as the community number of the terminal. The flowchart is as follows: Fig.10 shown.

[0108] After processing, T_RESULT (result table) is as follows:

[0109] Table 6. T_RESULT result table obtained by processing terminals not included in the algorithm results

[0110]

[0111]

[0112] You can see:

[0113] a. The upper terminal of terminal 10 is fitted to 7, and the community number is fitted to 4.

[0114] Step 5: Draw a topology diagram

[0115] Objective: Use the fitted topological data to draw the topological map of the current area.

[0116] Method: Based on the characteristics of the tree structure, start from the root node (transformer ID is 65535) and draw downwards layer by layer. According to the connection type and cabinet number, use the corresponding icon and number to hang the terminal into the corresponding cabinet.

[0117] Result: The topological diagram drawn is as follows Fig.11 As shown, the terminals marked in red are "to be checked" terminals:

[0118] Terminal 2: Need to check the power usage.

[0119] Terminal 8: Need to check the actual wiring information.

[0120] Terminal 10: needs to check power usage and location information.

Claims

1. An enhanced low-voltage topology identification method based on multi-dimensional data, characterized in that: The following steps are involved: Step 1: Collect the physical wiring information of the terminal to obtain a wiring database; collect the terminal location information and use the community division algorithm to obtain a community database; collect power consumption feature data and use the topology recognition algorithm to generate an algorithm result database based on the power consumption feature data; Step 2: Fit the connection database, the community database and the algorithm result database to obtain the topological data of the current substation area; Step 3: Use the topological data of the current area obtained in step 2 to draw a topological map of the current area.

2. The enhanced low-voltage topology identification method based on multi-dimensional data combination according to claim 1 is characterized in that: The data in the wiring database include terminal number, wiring type and cabinet number, the data in the community database include terminal number and community number, and the data in the algorithm result database include terminal number, terminal type and superior number.

3. The enhanced low-voltage topology identification method based on multi-dimensional data combination according to claim 2 is characterized in that: The wiring types include outgoing line cabinets, branch boxes and meter boxes, and the terminal types include transformers, general nodes and virtual nodes.

4. The enhanced low-voltage topology identification method based on multi-dimensional data combination according to claim 2 is characterized in that: In the terminal number, 0 represents virtual node 0, and 65535 represents a transformer.

5. The enhanced low voltage topology identification method based on multi-dimensional data combination according to claim 1 is characterized in that: In step 2, fitting the wiring database, the community database and the algorithm result database includes the following steps: Step 21: Create a topology result table. Step 22, processing unidentified terminals: finding the wiring neighbors of the unidentified terminals through the wiring database, finding the community neighbors of the unidentified terminals through the community database, taking the intersection of the wiring neighbors and the community neighbors to obtain the real first neighbor set, counting the upper terminals of the first neighbor set according to the algorithm result database, taking the upper terminal with the largest number of appearances as the upper terminal of the unidentified terminal, and adding the result to the topology result table; Step 23, mapping the non-zero virtual node to the real terminal: searching the algorithm result database to obtain the non-virtual subordinate terminal and the non-virtual topological neighbor, taking the intersection of the wiring database and the non-virtual subordinate terminal to obtain the first wiring type of the non-virtual subordinate terminal, subtracting 1 from the first wiring type to obtain the second wiring type of the virtual node, then searching the wiring database for the non-virtual wiring neighbor of the second wiring type, and finally taking the intersection of the non-virtual wiring neighbor and the non-virtual topological neighbor to obtain the real second neighbor set, taking the terminal with the smallest number in the second neighbor set as the real terminal of the non-zero virtual node, and adding the result to the topological result table; Step 24, processing the first terminal that is not in the algorithm result database: obtain the first wiring neighbor of the first terminal through the wiring database, and then search the upper terminal of the first wiring neighbor in the algorithm result database, and take the upper terminal with the largest number of appearances as the upper terminal of the first terminal; if the community number of the first terminal cannot be found from the community database, then search the community number of the first wiring neighbor in the community database, take the community number with the largest number of appearances as the community number of the first terminal, and add the result to the topology result table to obtain the topology data of the current substation.

6. The enhanced low-voltage topology identification method based on multi-dimensional data combination according to claim 5, characterized in that: The topology result table includes all data of the wiring database, the community database and the algorithm result database as well as data status attributes, and the data status attributes are used to indicate whether the data is fitted.

7. The enhanced low-voltage topology identification method based on multi-dimensional data combination according to claim 5, characterized in that: In step 22, when the wiring neighbor of the unidentified terminal is not found through the wiring database, the wiring type and cabinet number of the community neighbor are found by using the wiring database; when the community neighbor of the unidentified terminal is not found through the community database, the community number of the wiring neighbor is found by using the community database.

Images