A low-voltage power distribution area physical topology identification method and device, electronic equipment and storage medium

Abstract

The application discloses a low-voltage power distribution area physical topology identification method and device, electronic equipment and a storage medium, and belongs to the field of power distribution network automation. The method comprises the following steps: receiving voltage data sequences uploaded by a plurality of low-voltage power distribution network slave nodes, and constructing the low-voltage power distribution network slave nodes into a set of nodes to be processed; repeatedly performing a link construction operation until the number of nodes in the current node set is less than two, and outputting a plurality of link sets; constructing the plurality of link sets into a link group to be processed; repeatedly performing an endpoint merging operation until the current link group only contains one link set, and outputting a low-voltage power distribution area topology connection diagram. Therefore, by implementing the application, the problems of node cross-branch misconnection and chaotic topology structure construction caused by neglecting the line physical extension characteristics when identifying complex multi-level areas in the prior art can be solved.

Description

Technical Field

[0001] This invention relates to the field of power distribution network automation technology, specifically to a method, device, electronic equipment, and storage medium for identifying the physical topology of a low-voltage distribution substation. Background Technology

[0002] With the comprehensive advancement of smart grid construction, the refined management of low-voltage distribution substations places extremely high demands on the accuracy of archival data. Accurate physical topology of low-voltage distribution substations (i.e., clearly defined relationships between transformers and the connection sequence of each level of lines) is fundamental for advanced applications such as refined calculation of substation line losses, precise fault location, and load balancing adjustment of transformers. How to quickly and accurately identify the physical connections of the low-voltage distribution network using existing measurement data has become a critical issue that urgently needs to be addressed to improve the operation and maintenance level of the distribution network.

[0003] However, existing topology identification methods based on voltage data analysis still have certain limitations in practical applications. Most current mainstream technologies calculate the voltage correlation matrix between all nodes and then generate the topology structure all at once through global clustering or setting a uniform threshold. This approach often ignores the inherent "hierarchical extension" physical characteristic of distribution network topology when dealing with long lines, complex branches, or areas with large load fluctuations. This makes it difficult for the algorithm to effectively distinguish between adjacent nodes on the same electrical branch and nodes with similar electrical characteristics on different branches. In multi-level networks, this easily leads to node "misconnections," incorrect branch assignments, or topology breaks, severely impacting the accuracy and practicality of topology identification. Summary of the Invention

[0004] This invention provides a method, apparatus, electronic device, and storage medium for identifying the physical topology of low-voltage distribution transformer areas. It can solve the problems of node misconnection across branches and chaotic topology construction caused by ignoring the physical extension characteristics of lines when identifying complex multi-level transformer areas in the prior art.

[0005] An embodiment of the present invention provides a method for identifying the physical topology of a low-voltage distribution transformer area, applicable to the master node of a low-voltage distribution network. The identification method includes: Receive voltage data sequences uploaded by several low-voltage distribution network slave nodes, and construct the low-voltage distribution network slave nodes into a set of nodes to be processed; Repeat the link construction operation until the number of nodes in the current node set is less than two, and output multiple link sets; wherein, in each link construction operation, calculate the first line correlation coefficient between any two voltage data sequences in the current node set, select the pair of low-voltage distribution network slave nodes with the largest first line correlation coefficient to form a link set, and remove this pair of low-voltage distribution network slave nodes from the current node set; the initial node set is the node set to be processed; The multiple link sets are constructed into a link group to be processed; Repeat the endpoint merging operation until the current link group contains only one link set, and output the low-voltage distribution area topology connection diagram. The endpoint merging operation includes: Calculate the second line correlation coefficient between endpoint nodes of different link sets in the current link group, and select the two endpoint nodes with the largest second line correlation coefficient as target connection endpoints; wherein, the initial link group is the link group to be processed; Establish a topological connection between two target connection endpoints, merge the two link sets to which the two target connection endpoints belong into a new link set, and remove these two link sets from the current link group; Add the new link set to the current link group, and update the two endpoint nodes of the original two link sets, excluding the target connection endpoint, to the endpoint nodes of the new link set.

[0006] Furthermore, the step of receiving voltage data sequences uploaded by several low-voltage distribution network slave nodes and constructing the low-voltage distribution network slave nodes into a set of nodes to be processed includes: Receives several voltage data sequences uploaded from nodes in the low-voltage distribution network; The voltage data sequence is subjected to integrity verification to identify whether there are missing values ​​in the voltage data sequence; For each voltage data sequence with missing values, calculate the arithmetic mean of the valid samples in the current voltage data sequence; calculate the absolute value of the deviation between each valid sample in the current voltage data sequence and the arithmetic mean, and select the valid sample with the smallest absolute value of the deviation as the fill value; use the fill value to fill the missing values ​​in the current voltage data sequence to update the current voltage data sequence; The low-voltage distribution network is constructed from nodes into a set of nodes to be processed.

[0007] Furthermore, the first line correlation coefficient between any two voltage data sequences in the current node set is calculated using the following formula: in, This represents the first line correlation coefficient between any two voltage data sequences; and These represent the two voltage data sequences involved in the calculation; This indicates the total number of voltage sample points contained in the voltage data sequence; Represents voltage data sequence The Middle Voltage sample values ​​at each moment; Represents voltage data sequence The Middle Voltage sample values ​​at each moment; Represents voltage data sequence The arithmetic mean; Represents voltage data sequence The arithmetic mean.

[0008] Furthermore, before repeatedly performing the link building operation until the number of nodes in the current node set is less than two and outputting multiple link sets, the following steps are also included: Obtain the timestamp sequence of all voltage data sequences in the set of nodes to be processed; Perform an intersection operation on all timestamp sequences to determine the common time window shared by all voltage data sequences; According to the common time window, all voltage data sequences in the set of nodes to be processed are truncated, and Z-score normalization is performed on each truncated voltage data sequence to update each voltage data sequence.

[0009] Furthermore, the repeated endpoint merging operation continues until the current link group contains only one link set, outputting a low-voltage distribution area topology connection diagram, including: Repeat the endpoint merging operation until the current link group contains only one set of links; When it is detected that the current link group contains only one link set, the link set contained in the current link group is obtained as the target topology set; Traverse the connection relationships of all low-voltage distribution network slave nodes in the target topology set, and construct a low-voltage distribution network topology tree structure with the low-voltage distribution network master node as the root node; Based on the low-voltage distribution transformer area topology tree structure, a corresponding topology adjacency matrix is ​​generated, and the topology adjacency matrix is ​​visualized and rendered to generate a low-voltage distribution transformer area topology connection diagram.

[0010] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0011] One embodiment of the present invention provides a physical topology identification device for a low-voltage distribution transformer area, comprising: a data acquisition module, a link construction module, and a topology generation module; The data acquisition module is used to receive voltage data sequences uploaded by several low-voltage distribution network slave nodes and construct the low-voltage distribution network slave nodes into a set of nodes to be processed. The link building module is used to repeatedly perform link building operations until the number of nodes in the current node set is less than two, and output multiple link sets. In each link building operation, the first line correlation coefficient between any two voltage data sequences in the current node set is calculated, and a pair of low-voltage distribution network slave nodes with the largest first line correlation coefficient value is selected to form a link set. This pair of low-voltage distribution network slave nodes is then removed from the current node set. The initial node set is the node set to be processed. The topology generation module is used to construct the multiple link sets into a link group to be processed; repeatedly perform the endpoint merging operation until the current link group contains only one link set, and output the low-voltage distribution substation topology connection diagram; the endpoint merging operation includes: calculating the second line correlation coefficient between the endpoint nodes of different link sets in the current link group, and selecting the two endpoint nodes with the largest second line correlation coefficient as target connection endpoints; wherein, the initial link group is the link group to be processed; establishing a topology connection between the two target connection endpoints, merging the two link sets to which the two target connection endpoints belong into a new link set, and removing these two link sets from the current link group; adding the new link set to the current link group, and updating the other two endpoint nodes in the original endpoint nodes of these two link sets, excluding the target connection endpoints, to the endpoint nodes of the new link set.

[0012] Furthermore, the data acquisition module includes: a voltage data sequence receiving unit, a missing value supplementation unit, and a node set construction unit to be processed; The voltage data sequence receiving unit is used to receive several voltage data sequences uploaded from nodes in the low-voltage distribution network; The missing value completion unit is used to perform integrity verification on the voltage data sequence to identify whether there are missing values ​​in the voltage data sequence; for each voltage data sequence with missing values, the arithmetic mean of the valid samples in the current voltage data sequence is calculated; the absolute value of the deviation between each valid sample in the current voltage data sequence and the arithmetic mean is calculated, and the valid sample with the smallest absolute value of the deviation is selected as the filling value; the missing values ​​in the current voltage data sequence are filled using the filling value to update the current voltage data sequence; The node set construction unit is used to construct the low-voltage distribution network from nodes into a node set to be processed.

[0013] Furthermore, the low-voltage distribution area physical topology identification device also includes: a preprocessing module; The preprocessing module is used to obtain the timestamp sequence of all voltage data sequences in the set of nodes to be processed; Perform an intersection operation on all timestamp sequences to determine the common time window shared by all voltage data sequences; According to the common time window, all voltage data sequences in the set of nodes to be processed are truncated, and Z-score normalization is performed on each truncated voltage data sequence to update each voltage data sequence.

[0014] Based on the above method embodiments, the present invention provides corresponding electronic device embodiments.

[0015] An embodiment of the present invention provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the low-voltage distribution substation physical topology identification method described in any of the above-described method embodiments.

[0016] Based on the above method embodiments, the present invention provides corresponding storage medium embodiments.

[0017] One embodiment of the present invention provides a storage medium storing a computer program thereon, wherein, when the computer program is running, it controls the device where the storage medium is located to execute any of the low-voltage distribution substation physical topology identification methods described in the above-described method embodiments.

[0018] Compared with the prior art, the present invention has the following beneficial effects: This invention provides a method, apparatus, electronic device, and storage medium for identifying the physical topology of a low-voltage distribution transformer substation. The method receives voltage data sequences uploaded from several low-voltage distribution network slave nodes and constructs these slave nodes into a set of nodes to be processed. It repeatedly performs link construction operations until the number of nodes in the current node set is less than two, outputting multiple link sets. These multiple link sets are then constructed into a link group to be processed. Finally, it repeatedly performs endpoint merging operations until the current link group contains only one link set, outputting a low-voltage distribution transformer substation topology connection diagram.

[0019] This invention employs a hierarchical processing strategy of "first constructing initial links and then iteratively merging endpoints," leveraging the correlation of voltage data sequences to prioritize locking the most tightly connected node pairs. This effectively overcomes the shortcomings of existing global clustering algorithms that ignore the physical extension characteristics of the distribution network. Simultaneously, by forcibly limiting extension and merging to only "link endpoints," it ensures that the topology grows strictly along the actual route of the physical lines, thereby accurately solving the problems of node cross-branch haphazard connections and incorrect branch attribution caused by the similar electrical characteristics of non-adjacent nodes in complex multi-level distribution areas. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating a method for identifying the physical topology of a low-voltage distribution radio station area according to an embodiment of the present invention.

[0021] Figure 2 This is a schematic diagram of the structure of a low-voltage distribution radio station physical topology identification device provided in an embodiment of the present invention. Detailed Implementation

[0022] 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.

[0023] like Figure 1 As shown, to address the problems of node misconnection across branches and chaotic topology construction caused by ignoring the physical extension characteristics of lines when identifying complex multi-level distribution areas in existing technologies, an embodiment of the present invention provides a physical topology identification method for low-voltage distribution areas, applicable to the main nodes of low-voltage distribution networks. The identification method includes at least the following steps: Step S1: Receive voltage data sequences uploaded by several low-voltage distribution network slave nodes, and construct the low-voltage distribution network slave nodes into a set of nodes to be processed; In a preferred embodiment, receiving voltage data sequences uploaded by a plurality of low-voltage distribution network slave nodes and constructing the low-voltage distribution network slave nodes into a set of nodes to be processed includes: Receives several voltage data sequences uploaded from nodes in the low-voltage distribution network; The voltage data sequence is subjected to integrity verification to identify whether there are missing values ​​in the voltage data sequence; For each voltage data sequence with missing values, calculate the arithmetic mean of the valid samples in the current voltage data sequence; calculate the absolute value of the deviation between each valid sample in the current voltage data sequence and the arithmetic mean, and select the valid sample with the smallest absolute value of the deviation as the fill value; use the fill value to fill the missing values ​​in the current voltage data sequence to update the current voltage data sequence; The low-voltage distribution network is constructed from nodes into a set of nodes to be processed.

[0024] Specifically, in this step, the low-voltage distribution network master node receives voltage data sequences uploaded by several low-voltage distribution network slave nodes via power line carrier communication, low-power wireless communication, or fiber optic communication. The low-voltage distribution network slave nodes specifically refer to smart meters, voltage monitoring units, or sensing terminals with voltage acquisition functions deployed on the user side of the low-voltage distribution transformer area. The voltage data sequence is a sequence of voltage values ​​continuously collected by the low-voltage distribution network slave nodes within a preset time period and at a preset sampling frequency, each with a time stamp. After receiving the voltage data sequence, to ensure the reliability of subsequent physical topology identification calculations, the received raw data needs to be preprocessed, and the processed node objects are uniformly included in management, i.e., the set of nodes to be processed is constructed. The low-voltage distribution network master node is typically deployed on the low-voltage side of the distribution transformer, specifically using a smart transformer terminal unit (TTU) or an HPLC concentrator; the low-voltage distribution network slave nodes, in addition to smart meters, can also be loop status inspection instruments or photovoltaic inverter communication interfaces deployed in branch boxes. The preferred communication method is broadband power line carrier (HPLC), which has microsecond-level high-precision clock synchronization capability, providing high-precision synchronization voltage data support for subsequent correlation analysis.

[0025] It should be understood that in actual low-voltage distribution substation environments, the topology of the communication network often differs from that of the physical distribution network. Furthermore, due to environmental constraints, the communication methods between master and slave nodes are typically heterogeneous and diverse. In addition to the preferred HPLC broadband power line carrier communication described above, in some older substations or basement distribution rooms with severe communication shielding, the low-voltage distribution network slave nodes can also achieve local aggregation with an edge computing gateway deployed in a centralized meter box via low-power wireless (such as LoRa, ZigBee, etc.) protocols. This edge computing gateway then transmits the voltage data sequences of all its subordinate slave nodes to the low-voltage distribution network master node via a 4G / 5G cellular network or Ethernet.

[0026] In such complex hybrid communication networks, voltage data sequences not only face the problem of inconsistent sampling reference clocks among nodes, but are also highly susceptible to factors such as network congestion, out-of-order data packets, or transient interference. The voltage data sequences described in this embodiment of the invention typically have a sampling frequency set to once every 5 minutes or 15 minutes, with a complete data analysis period preferably ranging from 24 to 72 hours. This ensures that the data sequences contain sufficient effective voltage fluctuation characteristics (such as voltage dips caused by morning and evening peak electricity consumption, and voltage rises caused by large-scale distributed photovoltaic power generation at midday). By collecting these long-term series data containing rich electricity consumption behavior characteristics, a more solid data foundation can be provided for this invention, preventing misjudgments in subsequent correlation calculations due to occasional transient disturbances in the power grid (such as the starting impact of large motors). This greatly enhances the fault tolerance and field adaptability of the topology identification algorithm in complex communication and electrical environments.

[0027] In a preferred embodiment, the step of receiving voltage data sequences uploaded by a plurality of low-voltage distribution network slave nodes and constructing the low-voltage distribution network slave nodes into a set of nodes to be processed specifically includes the following process: First, a data receiving operation is performed, whereby the low-voltage distribution network master node receives several voltage data sequences uploaded by the low-voltage distribution network slave nodes. Next, an integrity check is performed on each received voltage data sequence to identify whether any values ​​are missing. A missing value refers to a situation where the voltage data sequence lacks a valid voltage reading at a specific point in time due to communication packet loss, equipment failure, or storage anomalies.

[0028] If the verification result indicates the presence of missing values ​​in the voltage data sequence, a filling strategy based on approximation of real samples is employed for repair. Specifically, for each voltage data sequence containing missing values, all non-empty values ​​in the voltage data sequence are first selected, i.e., valid samples, and the arithmetic mean of all valid samples is calculated. Subsequently, each valid sample in the voltage data sequence is iterated over, and the absolute value of the difference between each valid sample and the aforementioned calculated arithmetic mean is calculated, i.e., the absolute value of the deviation.

[0029] After obtaining the absolute value of the deviation corresponding to each valid sample, all absolute values ​​of deviation are compared numerically, and the valid sample corresponding to the smallest absolute value of deviation is selected as the filler value. The selected filler value is then filled into the position of the missing value in the voltage data sequence, thereby completing the repair of the missing value and updating the current voltage data sequence. Finally, the low-voltage distribution network slave nodes associated with the complete voltage data sequence after integrity verification and repair are uniformly added to the set of nodes to be processed.

[0030] By employing the real sample imputation method based on the minimum absolute value of the deviation, instead of directly using the arithmetic mean, it is ensured that the imputed data not only statistically approximates the central trend of the data but also represents actual voltage values ​​that have actually occurred in the voltage data sequence. This approach preserves the physical characteristics of voltage fluctuations in the low-voltage distribution network to the greatest extent possible, avoiding additional noise interference introduced by virtual values ​​generated during imputation calculations. This effectively improves the accuracy and robustness of subsequent calculations of line correlation coefficients based on the voltage data sequence.

[0031] Step S2: Repeat the link construction operation until the number of nodes in the current node set is less than two, and output multiple link sets; wherein, in each link construction operation, calculate the first line correlation coefficient between any two voltage data sequences in the current node set, select the pair of low-voltage distribution network slave nodes with the largest first line correlation coefficient to form a link set, and remove this pair of low-voltage distribution network slave nodes from the current node set; the initial node set is the node set to be processed; In a preferred embodiment, before repeatedly performing the link building operation until the number of nodes in the current node set is less than two and outputting multiple link sets, the method further includes: Obtain the timestamp sequence of all voltage data sequences in the set of nodes to be processed; Perform an intersection operation on all timestamp sequences to determine the common time window shared by all voltage data sequences; According to the common time window, all voltage data sequences in the set of nodes to be processed are truncated, and Z-score normalization is performed on each truncated voltage data sequence to update each voltage data sequence.

[0032] In a preferred embodiment, the first line correlation coefficient between any two voltage data sequences in the current node set is calculated using the following formula: in, This represents the first line correlation coefficient between any two voltage data sequences; and These represent the two voltage data sequences involved in the calculation; This indicates the total number of voltage sample points contained in the voltage data sequence; Represents voltage data sequence The Middle Voltage sample values ​​at each moment; Represents voltage data sequence The Middle Voltage sample values ​​at each moment; Represents voltage data sequence The arithmetic mean; Represents voltage data sequence The arithmetic mean.

[0033] Specifically, in this step, the core logic lies in pairing the nodes with the closest physical connections through iterative calculation and filtering. Before performing the specific link construction, considering that there may be clock asynchrony, slight deviations in data acquisition time, and inconsistent voltage amplitude references among the slave nodes in the low-voltage distribution network, in order to eliminate the interference of data asynchrony and dimensional differences on correlation calculation, in a preferred embodiment, it is necessary to first perform strict alignment and standardization processing on the data.

[0034] Before repeatedly performing the link construction operation, a data alignment step is first performed. The timestamp sequence carried by each voltage data sequence in the set of nodes to be processed is obtained. By performing a set intersection operation on all the timestamp sequences, a common time window that is included in all the voltage data sequences is determined. Based on the common time window, data is truncated for each voltage data sequence in the set of nodes to be processed, discarding data points outside the common time window to ensure that all voltage data sequences participating in subsequent calculations are strictly aligned in the time dimension and have a completely consistent number of sample points.

[0035] Subsequently, Z-score normalization (StandardScore) is performed on each of the extracted voltage data sequences. The Z-score normalization process aims to map voltage data of different voltage levels or fluctuation amplitudes to the same dimension, and the specific process is based on the following formula: in, This represents the first voltage data sequence after standardization. The standardized voltage value at each moment; This indicates the first voltage data in the original voltage data sequence. Voltage sample values ​​at each moment; This represents the arithmetic mean of the original voltage data sequence; This represents the standard deviation of the original voltage data sequence. Through the above standardization process, each voltage data sequence in the set of nodes to be processed is updated so that the updated voltage data sequences all follow a distribution with a mean of 0 and a standard deviation of 1. This allows the calculation of the correlation coefficient to focus on the morphological similarity of voltage fluctuations rather than differences in numerical magnitude. Furthermore, in long-distance low-voltage power supply lines, due to line impedance, the voltage amplitude at the end of the line is often significantly lower than at the beginning. If the original voltage values ​​are used directly for calculation, the difference in amplitude will mask the synchronicity of the fluctuations. Z-score standardization effectively eliminates the static voltage deviation caused by line impedance, ensuring that nodes at the beginning and end of the same branch exhibit highly consistent fluctuation curves after normalization, thereby significantly improving the robustness of long-distance line topology identification.

[0036] Furthermore, from the perspective of electrical theory analysis of distribution network operation, along the same low-voltage feeder, the actual voltage value at any point is determined by the output voltage of the transformer at the head end, the equivalent impedance of the line, and the instantaneous current of all loads from that point to the end. This means that nodes on the same physical branch, even if their absolute voltage amplitudes differ significantly, have highly coupled relative trends in voltage change over time (i.e., waveform fluctuation characteristics). Without Z-score normalization, or by using only the conventional Min-Max normalization method, the data is often severely affected by a few extreme outliers (such as lightning overvoltage or voltage drops caused by transient short circuits), leading to excessive data compression and obscuring the true fluctuation details.

[0037] The Z-score standardization method employed in this invention cleverly utilizes the mean and standard deviation of the data sequence as global statistics to shift and scale the voltage sequences of each node to the same reference frame. This method not only eliminates the static voltage drop caused by the fixed impedance of the line but also mathematically enhances the same-frequency fluctuation characteristics caused by dynamic load changes. Combined with the aforementioned missing value imputation strategy based on "true samples with the minimum absolute value of the deviation," this invention filters out interference from non-topological factors (such as meter errors, communication failures, and inherent line voltage drops) to the maximum extent during the data preprocessing stage. This provides a highly pure and physically meaningful input environment for subsequent pure mathematical feature extraction based on the first and second line correlation coefficients.

[0038] After completing the above data preprocessing, the link construction iterative process officially begins. First, the preprocessed set of nodes to be processed is set as the initial node set. Then, the link construction operation is repeated, with each operation including the following steps: checking the number of remaining nodes in the current node set; if the number of nodes is greater than or equal to two, then calculating the first line correlation coefficient between any two voltage data sequences in the current node set. The first line correlation coefficient is used to quantify the degree of synchronization of voltage fluctuations between two nodes; the larger the value, the closer the electrical distance and the stronger the physical connection between the two nodes.

[0039] In a preferred embodiment, the first line correlation coefficient between any two voltage data sequences in the current node set is calculated using the Pearson Correlation Coefficient algorithm, specifically based on the following formula: in, The first line correlation coefficient represents the relationship between any two voltage data sequences; and These represent the two voltage data sequences involved in the calculation; This represents the total number of voltage sample points contained in the voltage data sequence; Represents the voltage data sequence The Middle Voltage sample values ​​at each moment; Represents the voltage data sequence The Middle Voltage sample values ​​at each moment; Represents the voltage data sequence The arithmetic mean; Represents the voltage data sequence The arithmetic mean.

[0040] After calculating the first line correlation coefficients for all possible node pairs in the current node set, all calculated coefficient values ​​are sorted and filtered, and the pair of low-voltage distribution network slave nodes with the largest first line correlation coefficient is selected. This selected pair of low-voltage distribution network slave nodes is established as an independent link set, representing the strongest local connection between these two nodes. Subsequently, this pair of low-voltage distribution network slave nodes is removed from the current node set to prevent these two nodes from being repeatedly selected in subsequent iterations. The above process is repeated until the number of nodes in the current node set is less than two (i.e., unable to form a new node pair), at which point the loop stops, and all generated link sets are output.

[0041] It is worth noting that, in a preferred embodiment, before selecting the pair of low-voltage distribution network slave nodes with the largest first-line correlation coefficient to form a link set, a validity determination step is included. The system presets a connection determination threshold (e.g., 0.85). After selecting the first-line correlation coefficient with the largest value, this coefficient is compared with the connection determination threshold. If the coefficient is less than the connection determination threshold, it indicates that there is no significant physical connection between the remaining nodes in the current node set (possibly due to noise interference or being isolated nodes). In this case, link construction should be skipped or the loop should be terminated directly, rather than forcibly establishing a connection. This determination mechanism can effectively avoid topology misconnections caused by false high correlation, further improving the identification accuracy.

[0042] This greedy iterative strategy based on maximum correlation can prioritize the identification of adjacent node pairs with the closest electrical connections in the distribution network, thereby effectively constructing the initial physical link segments of the low-voltage distribution substation and laying an accurate foundation for subsequent topology merging.

[0043] Step S3: Construct the multiple link sets into a link group to be processed; Specifically, after completing the aforementioned link construction operation, the method outputs multiple independent link sets, each of which essentially represents a segment of physical line with the strongest local connectivity in a low-voltage distribution substation. To connect these dispersed segments into a complete topology, this step requires centralized structural management of all output link sets. Specifically, a dynamic list or set container is established to store topology segments. All link sets generated in step S2 are loaded into this dynamic list or set container, and the dynamic list or set container containing all the link sets is defined as the link group to be processed.

[0044] Simultaneously, to meet the retrieval requirements for "connection points" in subsequent endpoint merging operations, the topology attributes of the link sets need to be initialized and defined at this stage. Since each link set initially consists of only one pair of low-voltage distribution network slave nodes, these two low-voltage distribution network slave nodes included in each link set are directly defined as the initial endpoint nodes of the link set. This means that in the initial state of the link group to be processed, both ends of each link set are marked as active ports that can be extended or merged. By constructing the multiple link sets into the link group to be processed and explicitly defining the endpoints, discrete local line segments are effectively integrated into standard topology units that can be traversed by a global algorithm, providing a standardized data input foundation for subsequent topology growth and merging based on endpoint correlation.

[0045] Step S4: Repeat the endpoint merging operation until the current link group contains only one link set, and output the low-voltage distribution area topology connection diagram. In a preferred embodiment, the endpoint merging operation includes: Calculate the second line correlation coefficient between endpoint nodes of different link sets in the current link group, and select the two endpoint nodes with the largest second line correlation coefficient as target connection endpoints; wherein, the initial link group is the link group to be processed; Establish a topological connection between two target connection endpoints, merge the two link sets to which the two target connection endpoints belong into a new link set, and remove these two link sets from the current link group; Add the new link set to the current link group, and update the two endpoint nodes of the original two link sets, excluding the target connection endpoint, to the endpoint nodes of the new link set.

[0046] In a preferred embodiment, the repeated execution of the endpoint merging operation until the current link group contains only one link set, and the output of the low-voltage distribution substation topology map, includes: Repeat the endpoint merging operation until the current link group contains only one set of links; When it is detected that the current link group contains only one link set, the link set contained in the current link group is obtained as the target topology set; Traverse the connection relationships of all low-voltage distribution network slave nodes in the target topology set, and construct a low-voltage distribution network topology tree structure with the low-voltage distribution network master node as the root node; Based on the low-voltage distribution transformer area topology tree structure, a corresponding topology adjacency matrix is ​​generated, and the topology adjacency matrix is ​​visualized and rendered to generate a low-voltage distribution transformer area topology connection diagram.

[0047] Specifically, in this step, the algorithm shifts from local matching to global topology construction. Since low-voltage distribution networks physically resemble radial tree networks, the correct topology growth should involve lines extending towards both ends and merging with other lines. Based on this physical characteristic, this invention employs a restricted endpoint merging strategy, allowing only endpoint nodes located at the edges of the link set to participate in connection calculations, thereby preventing nodes located in the middle of the line from making incorrect cross-branch connections.

[0048] In one embodiment, the endpoint merging operation specifically includes the following steps: First, the correlation calculation and filtering steps are performed. The second line correlation coefficient between endpoint nodes of different link sets in the current link group is calculated. The calculation logic for the second line correlation coefficient is consistent with that of the first line correlation coefficient, both based on the ratio of the covariance to the standard deviation of the voltage data sequence to quantify the electrical distance between nodes. However, the calculation object is strictly limited to the "endpoint nodes" currently marked in each link set. After calculating the second line correlation coefficient for all possible endpoint pairing combinations, all calculated coefficient values ​​are sorted, and the two endpoint nodes with the largest second line correlation coefficient values ​​are selected as the target connection endpoints. This process ensures that each merging operation selects the path with the highest electrical connection probability among all feasible extension paths. Specifically, when this operation is performed for the first time, the initial link group is the link group to be processed constructed in step S3.

[0049] Next, the topology connection and set merging steps are performed. After identifying the two target connection endpoints, a topology connection relationship is established between these two target connection endpoints, and the two link sets to which these two target connection endpoints belong are logically merged into a new link set. Simultaneously, to maintain the updated state of the link group, the two merged old link sets are removed from the current link group.

[0050] Finally, the endpoint update step is performed. This is crucial to ensuring the topology can continue to grow outwards. The newly generated link set is added to the current link group. Since the two link sets before the merge contained a total of four endpoint nodes, two of which were the target connection endpoints with established connections, and the other two were "remote nodes" located at the other end of the line, the two endpoint nodes in the original two link sets, excluding the target connection endpoints, need to be updated to be endpoint nodes of the new link set. Through this update mechanism, the newly generated link set retains its ability to extend outwards, allowing the topology to continue growing along the direction of the physical lines.

[0051] In a preferred embodiment, the repeated execution of the endpoint merging operation until the current link group contains only one link set, and the output of the low-voltage distribution area topology connection diagram, specifically includes the following process: The endpoint merging operation described above is continuously executed in a loop. In each iteration, the number of link sets in the link group decreases by one, while the physical range covered by a single link set expands accordingly. The status of the sets in the current link group is monitored until the current link group contains only one link set, indicating that all local line segments are now connected, forming a complete simply connected component. At this point, the loop terminates, and the link sets contained in the current link group are obtained as the target topology set.

[0052] Subsequently, the target topology set is subjected to structured analysis and visualization. The connection relationships between all low-voltage distribution network slave nodes in the target topology set are traversed. Since the physical topology must have a clear hierarchy and flow direction, the low-voltage distribution network master node (i.e., the monitoring terminal on the distribution transformer side) is used as the root node. All nodes are hierarchically organized according to the extension direction of the connection relationships to construct a low-voltage distribution substation topology tree structure that conforms to the characteristics of a radial network. Specifically, a breadth-first search (BFS) or depth-first search (DFS) algorithm can be used to traverse the target topology set. First, the low-voltage distribution network master node is marked as layer 0 (root node); all slave nodes with direct connections to the layer 0 nodes are retrieved and marked as layer 1 nodes; this process continues, extending the search layer by layer until all nodes are traversed. During the search process, the parent-child relationship and hierarchical depth of each node are recorded, thereby constructing the topology tree structure with a clear hierarchical relationship.

[0053] Based on the constructed low-voltage distribution transformer area topology tree structure, a corresponding topological adjacency matrix is ​​generated. This matrix precisely describes the direct connection status between any two nodes within the transformer area in digital form. The generated topological adjacency matrix is ​​typically a sparse matrix, where an element '1' indicates that two nodes are directly connected, and '0' indicates that they are not connected. In actual storage, adjacency lists or Compressed Sparse Row (CSR) formats are preferred to reduce the memory consumption of electronic devices. This storage optimization strategy has significant engineering value, especially for large transformer areas containing hundreds of nodes. Finally, a graph theory visualization algorithm is used to render the topological adjacency matrix, generating and outputting a low-voltage distribution transformer area topology connection diagram that intuitively displays the hierarchical relationships and connection directions of the transformer areas.

[0054] By implementing the hierarchical processing strategy of "first constructing the initial link and then iteratively merging the endpoints" and in conjunction with mandatory endpoint extension constraints, this invention can effectively overcome the defect of traditional global clustering algorithms that ignore the physical extension characteristics of the distribution network, ensuring that the generated topology grows strictly along the actual direction of the physical line, thereby accurately solving the problem of node cross-branch random connection and branch assignment errors caused by the similar electrical characteristics of non-adjacent nodes in complex multi-level transformer areas.

[0055] To more clearly and intuitively illustrate the implementation process and technical effects of the low-voltage distribution transformer area physical topology identification method provided by this invention, a detailed deduction and explanation are given below in conjunction with a simplified low-voltage distribution transformer area simulation scenario.

[0056] Suppose a low-voltage distribution substation contains one low-voltage distribution network master node (distribution transformer) and six low-voltage distribution network slave nodes (labeled as node A, node B, node C, node D, node E, and node F, respectively). In the actual physical topology, nodes A and B are located on the first branch line (A is closer to the transformer, and B is at the end), nodes C and D are located on the second branch line (C is closer to the transformer, and D is at the end), and nodes E and F are located on the third, shorter branch line.

[0057] During the data acquisition phase (step S1), the master node collected six sets of voltage data sequences from nodes A to F at 15-minute intervals over 24 hours. During integrity verification, a missing value was found in the data sequence of node B at 12:00 noon. The system then calculated the arithmetic mean of all other valid voltage samples from node B (assumed to be 225.4V) and found that the actual voltage reading of node B at 14:30 was 225.5V, with the smallest absolute deviation from the mean (0.1V). Therefore, the system used the actual 225.5V as the filler value to repair the missing value at 12:00 noon, rather than fabricating a value, thus ensuring the data's physical authenticity. Subsequently, these six nodes were added to the "nodes to be processed set".

[0058] Entering the link construction phase (step S2), the system performs time window alignment and Z-score normalization on the voltage data sequences from A to F. Then, the first loop begins, calculating the first-line correlation coefficients (a total of 15 coefficients) between each pair of the six nodes. The calculation results show that since A and B are on the same physical branch, their voltage fluctuation synchronization is the highest, with a calculated correlation coefficient of 0.98 (the largest value). Therefore, the system groups nodes A and B into the first link set (denoted as link L1) and removes them from the node set to be processed. In the second loop, the remaining nodes C, D, E, and F continue to be calculated. It is found that the correlation coefficient between nodes C and D is the largest (0.96), so the system groups them into the second link set (denoted as link L2) and removes them. In the third loop, only nodes E and F remain, forming the third link set (denoted as link L3). At this point, the node set to be processed is empty, the link construction is complete, and three link sets, L1 (AB), L2 (CD), and L3 (EF), are output. This phase successfully identified the most basic and closely connected physical line segments in the transformer substation area.

[0059] Entering the topology generation phase (steps S3 and S4), the system constructs L1, L2, and L3 into a group of links to be processed, and marks A, B, C, D, E, and F as endpoint nodes of their respective link sets. In the first endpoint merging operation, the system calculates the second-line correlation coefficient between endpoints across links. At this time, the algorithm only calculates endpoint combinations (such as AC, AD, BC, etc.), strictly prohibiting merging within links. The calculation reveals that since nodes A, C, and E are physically close to the master node (i.e., the distribution transformer), they are most significantly affected by the transformer's output voltage. Therefore, the second-line correlation coefficient between endpoints A and C is the largest (assumed to be 0.92). The system then establishes a topology connection between endpoints A and C, merging L1 and L2 into a new link set L4. Simultaneously, the system updates the endpoint status: since A and C have been consumed as connection points, the outward extension endpoints of the new link set L4 are automatically updated to two other remote nodes, B and D.

[0060] In subsequent loops, only L4 (endpoints B and D) and L3 (endpoints E and F) remain in the link group. The system recalculates endpoint correlation and finds that there is no direct connection between L4 and L3 that satisfies high correlation (because they belong to different branches). However, since they are both connected to the master node, the algorithm incorporates the voltage data of the master node into the calculation and finds that the intersection area of ​​L4 (i.e., the end closer to A and C) and the endpoint E of L3 are highly correlated with the master node. Through the endpoint merging strategy, all local links eventually converge, leaving only a total link set containing all nodes in the link group.

[0061] Finally, starting from the master node of the low-voltage distribution network, the system uses a breadth-first search algorithm to traverse the connection relationships in the merged total link set, automatically generating a clear topology tree structure: three branches branch out in parallel under the master node, the first branch being master node-AB; the second branch being master node-CD; and the third branch being master node-EF. Subsequently, the corresponding topology adjacency matrix is ​​generated, and the topology connection diagram of the low-voltage distribution substation is rendered.

[0062] The entire simulation clearly demonstrates that by restricting merging to the "link endpoints," the algorithm effectively prevents node B at the end of the first branch from mistakenly connecting to node D at the end of the second branch due to similar voltage drop values. The method of this invention strictly follows the physical growth laws of the distribution network, starting from the strongest local connection and converging step by step, ultimately achieving accurate reconstruction of complex transformer substation topologies.

[0063] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0064] like Figure 2 As shown, an embodiment of the present invention provides a physical topology identification device for a low-voltage distribution transformer area, comprising: a data acquisition module, a link construction module, and a topology generation module; The data acquisition module is used to receive voltage data sequences uploaded by several low-voltage distribution network slave nodes and construct the low-voltage distribution network slave nodes into a set of nodes to be processed. The link building module is used to repeatedly perform link building operations until the number of nodes in the current node set is less than two, and output multiple link sets. In each link building operation, the first line correlation coefficient between any two voltage data sequences in the current node set is calculated, and a pair of low-voltage distribution network slave nodes with the largest first line correlation coefficient value is selected to form a link set. This pair of low-voltage distribution network slave nodes is then removed from the current node set. The initial node set is the node set to be processed. The topology generation module is used to construct the multiple link sets into a link group to be processed; repeatedly perform the endpoint merging operation until the current link group contains only one link set, and output the low-voltage distribution substation topology connection diagram; the endpoint merging operation includes: calculating the second line correlation coefficient between the endpoint nodes of different link sets in the current link group, and selecting the two endpoint nodes with the largest second line correlation coefficient as target connection endpoints; wherein, the initial link group is the link group to be processed; establishing a topology connection between the two target connection endpoints, merging the two link sets to which the two target connection endpoints belong into a new link set, and removing these two link sets from the current link group; adding the new link set to the current link group, and updating the other two endpoint nodes in the original endpoint nodes of these two link sets, excluding the target connection endpoints, to the endpoint nodes of the new link set.

[0065] In a preferred embodiment, the data acquisition module includes: a voltage data sequence receiving unit, a missing value supplementation unit, and a node set construction unit to be processed; The voltage data sequence receiving unit is used to receive several voltage data sequences uploaded from nodes in the low-voltage distribution network; The missing value completion unit is used to perform integrity verification on the voltage data sequence to identify whether there are missing values ​​in the voltage data sequence; for each voltage data sequence with missing values, the arithmetic mean of the valid samples in the current voltage data sequence is calculated; the absolute value of the deviation between each valid sample in the current voltage data sequence and the arithmetic mean is calculated, and the valid sample with the smallest absolute value of the deviation is selected as the filling value; the missing values ​​in the current voltage data sequence are filled using the filling value to update the current voltage data sequence; The node set construction unit is used to construct the low-voltage distribution network from nodes into a node set to be processed.

[0066] In a preferred embodiment, the low-voltage distribution area physical topology identification device further includes: a preprocessing module; The preprocessing module is used to obtain the timestamp sequence of all voltage data sequences in the set of nodes to be processed; Perform an intersection operation on all timestamp sequences to determine the common time window shared by all voltage data sequences; According to the common time window, all voltage data sequences in the set of nodes to be processed are truncated, and Z-score normalization is performed on each truncated voltage data sequence to update each voltage data sequence.

[0067] It should be noted that the embodiments of the device described above correspond to the embodiments of the present invention described above, and can realize the low-voltage distribution substation physical topology identification method described in any one of the above embodiments of the present invention. Furthermore, the embodiments of the device described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between modules indicates that they have a communication connection, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without creative effort.

[0068] Based on the above-described method embodiments of the present invention, a corresponding embodiment of an electronic device is provided.

[0069] An embodiment of the present invention provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the low-voltage distribution substation physical topology identification method according to any one of the present invention, or, when the processor executes the computer program, it implements the functions of each module in the above-described device embodiments.

[0070] For example, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the terminal device.

[0071] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0072] The processor can 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. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0073] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, applications required for at least one function, etc.; the data storage area may store data created based on the use of the mobile phone, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD card), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0074] Furthermore, to meet the high-concurrency reception and real-time processing requirements of massive low-voltage distribution network data from nodes, the processor of the electronic device is preferably an industrial-grade ARM architecture processor or a multi-core DSP processor with powerful parallel computing capabilities. In terms of hardware interface design, the electronic device also integrates multiple high-speed communication bus interfaces (such as gigabit Ethernet ports, RS-485 interfaces, etc.) and hardware encryption chips to ensure the real-time performance and security of voltage data sequences during transmission and processing.

[0075] When the electronic device is specifically instantiated as a smart converged terminal for a distribution network, it not only runs a lightweight IoT operating system (such as an edge computing container environment), but also maintains the dynamic set of nodes to be processed and the container of links to be processed as described in this invention in real time in its local memory. Through this hardware and software collaborative architecture, the electronic device can directly complete the preprocessing of massive voltage data, complex matrix operations of the correlation coefficients of the first / second lines, and adaptive growth of the topology tree structure at the edge of the distribution network, without having to upload all the underlying raw data to the cloud main station. This greatly saves the communication bandwidth and computing power of the main station, and significantly improves the response speed of distribution network topology identification and the overall operating efficiency of the system.

[0076] Based on the above method embodiments, the present invention provides corresponding storage medium embodiments; Another embodiment of the present invention provides a storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the storage medium is located to execute any of the above-described low-voltage distribution substation physical topology identification methods of the present invention.

[0077] The aforementioned storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in the form of source code, object code, executable file, or certain intermediate forms. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0078] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0079] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for identifying the physical topology of a low-voltage distribution transformer area, characterized in that, Applicable to low-voltage distribution network master nodes; the identification method includes: Receive voltage data sequences uploaded by several low-voltage distribution network slave nodes, and construct the low-voltage distribution network slave nodes into a set of nodes to be processed; Repeat the link construction operation until the number of nodes in the current node set is less than two, and output multiple link sets; wherein, in each link construction operation, calculate the first line correlation coefficient between any two voltage data sequences in the current node set, select the pair of low-voltage distribution network slave nodes with the largest first line correlation coefficient to form a link set, and remove this pair of low-voltage distribution network slave nodes from the current node set; the initial node set is the node set to be processed; The multiple link sets are constructed into a link group to be processed; Repeat the endpoint merging operation until the current link group contains only one link set, and output the low-voltage distribution area topology connection diagram. The endpoint merging operation includes: Calculate the second line correlation coefficient between endpoint nodes of different link sets in the current link group, and select the two endpoint nodes with the largest second line correlation coefficient as target connection endpoints; wherein, the initial link group is the link group to be processed; Establish a topological connection between two target connection endpoints, merge the two link sets to which the two target connection endpoints belong into a new link set, and remove these two link sets from the current link group; Add the new link set to the current link group, and update the two endpoint nodes of the original two link sets, excluding the target connection endpoint, to the endpoint nodes of the new link set.

2. The low-voltage distribution transformer area physical topology identification method as described in claim 1, characterized in that, The process of receiving voltage data sequences uploaded by several low-voltage distribution network slave nodes and constructing the low-voltage distribution network slave nodes into a set of nodes to be processed includes: Receives several voltage data sequences uploaded from nodes in the low-voltage distribution network; The voltage data sequence is subjected to integrity verification to identify whether there are missing values ​​in the voltage data sequence; For each voltage data sequence with missing values, calculate the arithmetic mean of the valid samples in the current voltage data sequence; calculate the absolute value of the deviation between each valid sample in the current voltage data sequence and the arithmetic mean, and select the valid sample with the smallest absolute value of the deviation as the fill value; use the fill value to fill the missing values ​​in the current voltage data sequence to update the current voltage data sequence; The low-voltage distribution network is constructed from nodes into a set of nodes to be processed.

3. The low-voltage distribution transformer area physical topology identification method as described in claim 2, characterized in that, The first line correlation coefficient between any two voltage data sequences in the current node set is calculated using the following formula: in, This represents the first line correlation coefficient between any two voltage data sequences; and These represent the two voltage data sequences involved in the calculation; This indicates the total number of voltage sample points contained in the voltage data sequence; Represents voltage data sequence The Middle Voltage sample values ​​at each moment; Represents voltage data sequence The Middle Voltage sample values ​​at each moment; Represents voltage data sequence The arithmetic mean; Represents voltage data sequence The arithmetic mean.

4. The low-voltage distribution transformer area physical topology identification method as described in claim 3, characterized in that, Before repeatedly performing the link building operation until the number of nodes in the current node set is less than two and outputting multiple link sets, the process also includes: Obtain the timestamp sequence of all voltage data sequences in the set of nodes to be processed; Perform an intersection operation on all timestamp sequences to determine the common time window shared by all voltage data sequences; According to the common time window, all voltage data sequences in the set of nodes to be processed are truncated, and Z-score normalization is performed on each truncated voltage data sequence to update each voltage data sequence.

5. The low-voltage distribution transformer area physical topology identification method as described in claim 4, characterized in that, The endpoint merging operation is repeated until the current link group contains only one link set, outputting a low-voltage distribution area topology connection diagram, including: Repeat the endpoint merging operation until the current link group contains only one set of links; When it is detected that the current link group contains only one link set, the link set contained in the current link group is obtained as the target topology set; Traverse the connection relationships of all low-voltage distribution network slave nodes in the target topology set, and construct a low-voltage distribution network topology tree structure with the low-voltage distribution network master node as the root node; Based on the low-voltage distribution transformer area topology tree structure, a corresponding topology adjacency matrix is ​​generated, and the topology adjacency matrix is ​​visualized and rendered to generate a low-voltage distribution transformer area topology connection diagram.

6. A physical topology identification device for a low-voltage distribution transformer area, characterized in that, include: Data acquisition module, link construction module, and topology generation module; The data acquisition module is used to receive voltage data sequences uploaded by several low-voltage distribution network slave nodes and construct the low-voltage distribution network slave nodes into a set of nodes to be processed. The link building module is used to repeatedly perform link building operations until the number of nodes in the current node set is less than two, and output multiple link sets. In each link building operation, the first line correlation coefficient between any two voltage data sequences in the current node set is calculated, and a pair of low-voltage distribution network slave nodes with the largest first line correlation coefficient value is selected to form a link set. This pair of low-voltage distribution network slave nodes is then removed from the current node set. The initial node set is the node set to be processed. The topology generation module is used to construct the multiple link sets into a link group to be processed; repeatedly perform the endpoint merging operation until the current link group contains only one link set, and output the low-voltage distribution substation topology connection diagram; the endpoint merging operation includes: calculating the second line correlation coefficient between the endpoint nodes of different link sets in the current link group, and selecting the two endpoint nodes with the largest second line correlation coefficient as target connection endpoints; wherein, the initial link group is the link group to be processed; establishing a topology connection between the two target connection endpoints, merging the two link sets to which the two target connection endpoints belong into a new link set, and removing these two link sets from the current link group; adding the new link set to the current link group, and updating the other two endpoint nodes in the original endpoint nodes of these two link sets, excluding the target connection endpoints, to the endpoint nodes of the new link set.

7. The low-voltage distribution transformer area physical topology identification device as described in claim 6, characterized in that, The data acquisition module includes: a voltage data sequence receiving unit, a missing value supplementation unit, and a node set construction unit to be processed; The voltage data sequence receiving unit is used to receive several voltage data sequences uploaded from nodes in the low-voltage distribution network; The missing value completion unit is used to perform integrity verification on the voltage data sequence to identify whether there are missing values ​​in the voltage data sequence; for each voltage data sequence with missing values, the arithmetic mean of the valid samples in the current voltage data sequence is calculated; the absolute value of the deviation between each valid sample in the current voltage data sequence and the arithmetic mean is calculated, and the valid sample with the smallest absolute value of the deviation is selected as the filling value; the missing values ​​in the current voltage data sequence are filled using the filling value to update the current voltage data sequence; The node set construction unit is used to construct the low-voltage distribution network from nodes into a node set to be processed.

8. The low-voltage distribution transformer area physical topology identification device as described in claim 7, characterized in that, Also includes: Preprocessing module; The preprocessing module is used to obtain the timestamp sequence of all voltage data sequences in the set of nodes to be processed; Perform an intersection operation on all timestamp sequences to determine the common time window shared by all voltage data sequences; According to the common time window, all voltage data sequences in the set of nodes to be processed are truncated, and Z-score normalization is performed on each truncated voltage data sequence to update each voltage data sequence.

9. An electronic device, characterized in that, The device 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 distribution area physical topology identification method as described in any one of claims 1 to 5.

10. A storage medium, characterized in that, The storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the storage medium is located to perform the low-voltage distribution substation physical topology identification method as described in any one of claims 1 to 5.

Images