An automatic topology identification method and system for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics

Abstract

An automatic topology identification method and system for low-voltage transparent distribution areas based on HPLC+HRF dual-mode communication and electrical characteristics belongs to the field of power system distribution automation technology. The method includes: acquiring the electrical characteristics of each meter's data; clustering the communication delay and signal attenuation characteristics of each meter acquired based on HPLC+HRF dual-mode communication, aggregating all meters in the same meter box into a single meter box node; calculating the coupling impedance between any two meter box nodes using the electrical characteristics, iteratively creating a common parent node as a branch box node for the two nodes corresponding to the maximum coupling impedance, adding it to the node set and removing its child nodes until only one node remains in the node set as the root branch box node connected to the transformer node, establishing a four-level tree-structured topology map of transformer-branch box-meter box node-meter within the distribution area, generating a physical topology map of the low-voltage distribution area. This invention improves the accuracy of identifying hierarchical relationships and physical distances.

Description

Technical Field

[0001] This invention belongs to the field of power system distribution automation technology, and more specifically, relates to an automatic topology identification technology for low-voltage transparent transformer areas based on communication and electrical characteristics. Background Technology

[0002] As the "last mile" connecting the power grid and users, the digital and transparent management of low-voltage distribution networks is fundamental to improving power supply reliability and service quality. However, the current massive and dynamic low-voltage distribution areas lack accurate and real-time physical topology data. Existing management models heavily rely on manual maintenance of ledgers and data entry into systems such as Power Management Systems (PMS), resulting in outdated information and discrepancies between the hierarchical connection relationships of "transformer-branch box-meter box-user" and the actual site conditions. This discrepancy directly leads to a series of operational bottlenecks, such as difficulty in accurately locating power outages within minutes, distorted line loss analysis, and inability to effectively trace the source of power quality anomalies. This hinders the implementation of advanced applications based on digital outage assessment, rapid power restoration, and lean operation and maintenance, becoming a key bottleneck restricting the realization of transparency and fundamental improvement in power supply reliability in distribution networks.

[0003] Existing technical document 1 (CN118861780A) discloses a method, system, device, and storage medium for identifying low-voltage distribution area boxes and meters. Its shortcomings lie in that it only focuses on identifying the ownership of "meter boxes," without addressing the hierarchical topology derivation of "transformers, branch boxes, and meter boxes," thus failing to form a complete four-level topology structure for low-voltage distribution areas. Furthermore, it relies on RSSI feature mapping distance for clustering and does not combine electrical features to achieve joint modeling of topology and impedance, making it difficult to support advanced applications such as power outage location and line loss diagnosis.

[0004] Prior art document 2 (CN121355890A) discloses a non-disruptive topology identification method for low-voltage distribution areas. Its shortcomings are that it relies on RSSI transformation distance to construct an undirected graph, the processing of low-connectivity nodes during clustering can easily lead to insufficient topology integrity, and it only optimizes the topology through the law of energy conservation and Kirchhoff's current law, without introducing a synchronous timestamp calibration and impedance matrix recursion mechanism, thus limiting the accuracy and stability of topology identification.

[0005] Prior art document 3 (CN120691583A) discloses a method for identifying transformer substation topology based on carrier communication networks. Its shortcomings are that it only relies on signal characteristics such as distance and signal-to-noise ratio sequence similarity of carrier communication networks, without integrating electrical parameters for topology verification, and the hierarchical identification depends on the communication time difference between the concentrator and the cluster unit, which requires high stability of the communication environment and is prone to hierarchical misjudgment in complex distribution network environments. Summary of the Invention

[0006] To address the shortcomings of existing low-voltage transformer substation topology identification methods, such as incomplete identification, reliance on numerous monitoring devices, and inability to output line impedance, this invention provides an automatic low-voltage transparent transformer substation topology identification method and system based on HPLC+HRF dual-mode communication and electrical characteristics. It integrates the precise time delay characteristics of HPLC (High-speed Power Line Carrier) with the signal spatial characteristics of HRF (High-speed Radio Frequency), reliably identifying meter-meter box relationships through a fuzzy clustering algorithm. Furthermore, based on the "household-box" relationship, by monitoring the synchronous changes in node current disturbances and the overall network voltage response, it calculates the coupling impedance matrix and designs a maximum element recursive algorithm to automatically discover and deduce a complete transformer-branch box-meter box-meter topology map that does not require monitoring of branch boxes. This invention effectively overcomes the limitations of traditional methods, such as sensitivity to communication environments, reliance on complex verification models, or the need for massive monitoring points. It integrates electrical-spatial dual-channel characteristics and combines the spatial and electrical membership accuracy of meters and meter boxes. This makes the inference based on the physical coupling impedance of the power grid less susceptible to interference from the communication environment, improving the accuracy of the physical meaning of automatic topology identification. It not only identifies topology connection relationships but also synchronously outputs impedance values ​​accurate to each line segment, providing directly usable key data for line loss calculation, fault location, and power supply capacity analysis. This expands the application value of topology data. This invention can dynamically update the topology as equipment changes, ensuring the long-term accuracy of the data.

[0007] The present invention adopts the following technical solution.

[0008] The first aspect of this invention provides an automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics, comprising:

[0009] Data from each electricity meter is acquired, and its electrical characteristics are extracted. Based on HPLC+HRF dual-mode communication, the communication delay and signal attenuation characteristics between the meters are obtained. A clustering algorithm is then used to obtain a list of meter boxes to which the meters belong. The electrical characteristics include voltage and current.

[0010] Calculate the coupling impedance between any two meter box nodes using electrical characteristics; create a common parent node as a branch box node for the two meter box nodes corresponding to the maximum coupling impedance to replace the two meter box nodes corresponding to the maximum coupling impedance in the node set; calculate the coupling impedance between any two nodes in the node set; repeat the above process until only one branch box node remains in the node set; take this branch box node as the root branch box node; take the transformer node as the parent node of the root branch box node; and take all meter box nodes as child nodes of the root branch box node; establish a four-level tree topology diagram of transformer-branch box-meter box node-meter in the transformer substation area.

[0011] Each meter in the meter box list is assigned as a child node and attached to the corresponding meter box node in the four-level tree topology diagram, based on its assigned meter box, to form a four-level tree topology diagram of transformer-branch box-meter box-meter; and a low-voltage distribution area physical topology diagram is generated based on the four-level tree topology diagram.

[0012] Preferably, the electrical characteristics of the meter data include: voltage and current;

[0013] Based on HPLC+HRF dual-mode communication, the communication delay characteristics and signal attenuation characteristics of each meter were obtained, including:

[0014] The electrical characteristics are encapsulated into data frames by a concentrator, and the data frame transmission time and data frames are modulated by HPLC and then sent to each electricity meter.

[0015] The meter acquires the arrival time of the data frame, generates a response frame, and sends the response frame with the transmission time via HPLC back to the concentrator. The concentrator then acquires the reception time of the response frame.

[0016] Based on the frequency deviation of the transceiver crystal oscillator generated by multiple interactions of data frames and response frames between the transceiver and the transceiver, the arrival time of the data frame and the transmission time of the response frame are corrected.

[0017] Calculate the one-way propagation delay from the concentrator to each meter based on the corrected data frame arrival time, the corrected response frame transmission time, the data frame transmission time, and the response frame reception time.

[0018] The communication delay characteristics of each electricity meter are generated based on the device asset ID of each meter and the one-way propagation delay from the concentrator to each meter.

[0019] Based on HRF, the signal attenuation characteristics of each meter are extracted by listening to the signals broadcast by the meters.

[0020] Preferably, the frequency deviation is expressed by the following formula:

[0021]

[0022] In the formula, This represents the frequency deviation of the crystal oscillator of the i-th meter relative to the concentrator crystal oscillator. , Indicates the total number of meters. This represents the total number of interactions. This represents the processing time measured by the i-th meter in the nth interaction using its locally biased clock. This represents the remaining time after deducting the processing time of the i-th meter in the nth interaction from the total round-trip time.

[0023] Preferably, clustering the communication delay characteristics and signal attenuation characteristics of each meter obtained based on HPLC+HRF dual-mode communication includes:

[0024] Based on communication delay characteristics and signal attenuation characteristics, a multi-dimensional meter feature is constructed for each meter.

[0025] Using the multidimensional meter features of all meters as input, a fuzzy clustering algorithm is used for iterative calculation. By optimizing the objective function to minimize the sum of weighted distances from the multidimensional meter features of all meters to each cluster center, the meter-meter box membership matrix is ​​obtained.

[0026] Traverse the meter-box membership matrix row by row, determine that all meters in a row belong to the box corresponding to the maximum membership value, and aggregate all meters that are determined to belong to the same box to obtain a box node.

[0027] Preferably, any one of the meter boxes is acquired in real time. The total current change and the total current change at other meter box nodes within the transformer area The voltage change caused by the above. The ratio of voltage change to total current change is used as the reference value for the two meter box nodes. , The coupling impedance between them is used to construct a coupling impedance matrix.

[0028] Preferably, establishing a four-level tree topology diagram of transformer-branch box-meter box node-meter includes:

[0029] Find the maximum value of the coupling impedance in the current coupling impedance matrix, and determine that the meter box node p corresponding to the maximum value of the coupling impedance is directly connected to the meter box node q.

[0030] Create a common parent node for both table box node p and table box node q, denoted as branch box node H;

[0031] After removing box node p and box node q from node set C, add branch box node H to node set C to obtain a new node set Cnew;

[0032] Calculate the coupling impedance between branch box node H and each meter box node in the new node set Cnew to generate a new coupling impedance matrix. Repeat the above process for the new coupling impedance matrix. If the coupling impedance between branch box node H and meter box node p' is the largest, determine that the branch box node H corresponding to the maximum coupling impedance is directly connected to the meter box node p'. Create a common parent node for branch box node H and meter box node p', denoted as branch box node H'. Remove branch box node H and meter box node p' from the node set Cnew and add branch box node H' to the node set Cnew until only one branch box node remains in the final node set. Set the remaining branch box node as the root branch box node, set the parent node of the root branch box node as the transformer node, point the transformer node to the transformer, divide the iteratively merged branch box nodes into second-level topology nodes, divide the meter box nodes into third-level topology nodes, divide the original meters into fourth-level topology nodes, and output a four-level tree topology diagram of transformer-branch box-meter box node-meter.

[0033] Preferably, the set of nodes in the four-level tree topology graph is the table box node, branch box node, and transformer node in the recursive process, and the set of edges in the four-level tree topology graph is the physical lines between the nodes in the set of nodes.

[0034] Preferably, the final line impedance of each edge in the four-level tree topology is calculated, and the final line impedance is labeled as an attribute on the corresponding connecting edge in the four-level tree topology. The final line impedance of each edge in the four-level tree topology includes:

[0035] The final coupling impedance matrix is ​​generated by taking the coupling impedance matrices of all meter box nodes, branch box nodes, and transformer nodes from the node set of the four-level tree topology diagram of transformer-branch box-meter box node-meter. An environmental attenuation factor is then used. Correct the final coupling impedance matrix to generate the corrected coupling impedance matrix;

[0036] Based on the corrected coupling impedance matrix, a system of linear equations is established for each edge of the four-level tree topology. The least squares method is used to solve the system of linear equations to obtain the line impedance.

[0037] The theoretical coupling impedance is determined based on the line impedance. When the error between the theoretical coupling impedance and the corresponding impedance in the final coupling impedance matrix is ​​greater than a set threshold, the environmental attenuation factor is corrected, and the line impedance is recalculated until the error between the theoretical coupling impedance and the corresponding impedance in the final coupling impedance matrix is ​​less than the set threshold. Finally, the final line impedance is output.

[0038] Preferably, the environmental degradation factor is corrected and expressed by the following formula:

[0039]

[0040] In the formula, This represents the updated environmental degradation factor. This represents the environmental decay factor used in the current iteration. Indicates the learning rate. This represents the ratio of the corresponding impedance in all theoretical coupling impedances and the final coupling impedance matrix. }'s median, Indicates the first The node and the first The theoretical coupling impedance of each node and the corresponding impedance ratio in the final coupling impedance matrix are expressed by the following formula:

[0041]

[0042] In the formula, Indicates the first The node and the first The theoretical coupling impedance of each node, Indicates the first The node and the first The coupling impedance matrix of each node.

[0043] Preferably, each node in the four-level tree topology diagram includes a meter box node, a branch box node, and a transformer node;

[0044] Generating a physical topology map of the low-voltage distribution area based on a four-level tree topology map includes:

[0045] In the equipment ledger, filter the physical equipment attributes corresponding to each node in the four-level tree topology diagram. Use the equipment asset ID of the node in the four-level tree topology diagram as the unique key to match the physical equipment attributes. Write the matched physical equipment attributes into the attribute field of the corresponding node in the four-level tree topology diagram to form a four-level tree topology diagram with bound physical equipment attributes.

[0046] The concentrator extracts BeiDou positioning data from each node in the four-level tree topology diagram. Combining the hierarchical relationship and line impedance in the four-level tree topology diagram, it verifies the spatial rationality of the device connection relationship in the four-level tree topology diagram with bound physical device attributes. It then outputs a spatially verified four-level tree topology diagram to verify the spatial rationality of the positioning information.

[0047] The four-level tree topology diagram that has undergone spatial verification is compared with the configuration information of the transformer area archive to output the transformer-branch-box-household physical topology model.

[0048] The physical topology model of transformer-branch-box-household is visualized and rendered on the GIS map, and its attributes are attached. Finally, the physical topology map of the low-voltage distribution area is output.

[0049] The second aspect of the present invention provides an automatic topology identification system for low-pressure transparent stages based on HPLC+HRF dual-mode communication and electrical characteristics, which, when running the automatic topology identification method for low-pressure transparent stages based on HPLC+HRF dual-mode communication and electrical characteristics described in the first aspect, includes:

[0050] The membership generation module is used to obtain the electrical characteristics of each meter data; it clusters the communication delay characteristics and signal attenuation characteristics of each meter obtained based on HPLC+HRF dual-mode communication, and assigns all meters in the same cluster to the same meter box according to the maximum membership degree, and aggregates all meters in the same meter box into a meter box node.

[0051] The four-level tree topology generation module is used to calculate the coupling impedance between any two meter box nodes using electrical characteristics; it creates a common parent node as a branch box node for the two meter box nodes corresponding to the maximum coupling impedance to replace the two meter box nodes corresponding to the maximum coupling impedance in the node set; it calculates the coupling impedance between any two nodes in the node set; it repeats the above process until only one branch box node remains in the node set; it takes this branch box node as the root branch box node; it takes the transformer node as the parent node of the root branch box node; and all meter box nodes are child nodes of the root branch box node; thus, a four-level tree topology of transformer-branch box-meter box node-meter is established within the transformer substation area.

[0052] The physical topology generation module is used to generate a physical topology map of the low-voltage distribution area based on a four-level tree topology map.

[0053] Compared with the prior art, the beneficial effects of the present invention include at least the following:

[0054] Based on HPLC+HRF dual-mode communication, the communication delay characteristics obtained by HPLC reflect the electrical path length of the signal from the concentrator to the meter. However, relying solely on HPLC may misclassify electrically proximate meters as spatially proximate. The signal attenuation characteristics obtained by HRF reflect the linear spatial distance between the meter and the concentrator. Relying solely on HRF may misclassify spatially proximate meters as electrically proximate. Meters located in the same meter box will exhibit high consistency in both electrical and spatial dimensions, which are completely independent. For meters truly belonging to the same box, they are close to each other both electrically and spatially, thus obtaining a high overall membership degree pointing to the same group. For meters that are misclassified as electrically proximate but spatially distant, or spatially proximate but electrically distant, significant inconsistencies will be observed in the other dimension, leading to a decrease in their overall membership degree and classification into other, more suitable groups. Therefore, this invention utilizes HPLC's susceptibility to electrical topology. To address the issues of confusion and misjudgment due to spatial layout confusion in HRF, this paper integrates communication delay characteristics obtained from HPLC and signal attenuation characteristics obtained from HRF. Through fuzzy clustering fusion, meters are strongly identified as belonging to the same group only when both electrical paths and spatial locations are consistent. This reduces the risk of initial clustering errors caused by interference in a single communication channel (such as power line noise or wireless signal obstruction) or special grid structures (such as long branch lines or complex wiring). This improves the accuracy and reliability of identifying the meter and meter box affiliation relationships for users. Since subsequent electrical deduction heavily relies on the correctness of the initial clustering, the improved robustness of the meter and meter box affiliation relationships directly reduces the systemic risk of the entire topology identification system failing due to front-end errors. This enhances the practicality and stability of the four-level tree topology diagram of transformer-branch box-meter box-meter, and improves and ensures the effectiveness of the final low-voltage distribution area physical topology diagram.

[0055] By utilizing the existing HPLC / HRF dual-mode communication channel of the existing electricity meter to acquire features, there is no need to deploy additional dedicated sensors or acquisition equipment at intermediate nodes such as branch boxes and JP cabinets or at each user's electricity meter. This greatly reduces the complexity and overall cost of hardware procurement, installation and maintenance, making large-scale, universal digital transformation of low-voltage distribution areas economically feasible.

[0056] By using dual-feature fusion, it is ensured that all meters within a meter box node are electrically connected in parallel to the same physical connection point. The minute current changes of each meter are aggregated into a total current change, which is electrically equivalent to a current disturbance occurring at the main incoming line point of the meter box. When a current change is injected into node j, according to circuit theory, the voltage change it causes at node j' is determined by the transfer impedance between the two. In a radial distribution network, this transfer impedance is exactly equal to the sum of the path impedances from the transformer to the nearest common connection point between the two nodes. Therefore, the coupling impedance accurately quantifies the electrical distance between nodes j and j' in the topology tree. Equal coupling impedances indicate the same electrical distance, while maximum coupling impedance indicates the same electrical distance. With the shortest air distance, this invention utilizes the synchronous electrical quantity measurement of the terminal meter box node to reverse deduce the complete intermediate layer topology and line parameters. This significantly reduces the hardware deployment density, complexity, and cost required to achieve complete topology identification, enabling transparency without modifying the intermediate network and simply through terminal intelligence. It reduces the dependence on intermediate layer monitoring hardware and sensitivity to environmental noise, achieving high-reliability topology identification in low-cost and complex environments. If the clustering is incorrect, the total current change becomes a set of disturbances injected from multiple discrete points in the power grid. The voltage disturbance field generated cannot be described by a simple impedance model between pairs of nodes, causing the calculation of the entire coupling impedance matrix to fail and making topology deduction impossible.

[0057] Based on HPLC / HRF dual-mode communication, as long as the current and voltage changes are measured accurately, the coupling impedance matrix is ​​reliable, and the derived topology is accurate. This enables it to handle transformer substations with large differences in line length, complex wiring, and harsh communication environments. By solving the coupling impedance through strict time synchronization of current and voltage changes, the sensitivity of the identification results to complex field communication environments and spatial structures is reduced. This ensures that the coupling impedance has a clear physical meaning and is calculable, enhancing the robustness and universality of the four-level tree topology. The four-level tree topology of transformer-branch box-meter box-meter improves the applicability and reliability of results in complex real-world scenarios. If the current and voltage changes are not strictly time-synchronized, they may originate from two completely unrelated independent electrical events. In this case, the coupling impedance is just a meaningless random ratio, losing all physical meaning.

[0058] The complex problem of network topology inference is transformed into a deterministic process that relies solely on the maximum search and iterative merging of the coupling impedance matrix. It does not require presetting the number of network layers or branch boxes, nor does it require complex graph search or probabilistic models. It automatically generates a complete tree structure from the data, including automatically discovering and creating branch box nodes of all intermediate levels, thus improving the automation level of topology identification. Attached Figure Description

[0059] Figure 1This is a schematic diagram of the method flow provided according to an embodiment of the present invention;

[0060] Figure 2 This is a schematic diagram of the deployment of an integrated sensing and computing module according to an embodiment of the present invention;

[0061] Figure 3 This is a schematic diagram of HPLC carrier time delay ranging provided according to an embodiment of the present invention;

[0062] Figure 4 This is a schematic diagram illustrating the process of deriving a four-level tree topology diagram based on impedance calculation, according to an embodiment of the present invention.

[0063] Figure 5 This is a schematic diagram of a four-level tree topology provided according to an embodiment of the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0065] like Figure 1 As shown, Embodiment 1 of the present invention provides an automatic topology identification method for low-voltage transparent substations based on communication and electrical characteristics, comprising the following steps:

[0066] Step 1: Obtain the electrical characteristics of each meter's data, which include voltage and current.

[0067] In a preferred but non-limiting embodiment of the present invention, step 1 includes:

[0068] like Figure 2 As shown, an integrated sensing and computing module is deployed at the main incoming line of each meter box within the distribution area to collect the electrical characteristics of each meter in real time, including real-time data such as voltage and current. The integrated sensing and computing module includes a low-voltage smart switch and a meter box branch LTU / rail meter, used to uniformly collect the overall electrical characteristics of the meter box, such as voltage, current, active power, and reactive power.

[0069] Low-voltage intelligent switches refer to miniature circuit breakers with remote communication and intelligent control functions. In addition to basic electrical quantity acquisition and communication functions, they integrate the ability to open / close circuits. They can not only monitor but also remotely cut off or restore power to the meter box when necessary (such as in case of overload, undervoltage, or receiving dispatch instructions). Line Terminal Units (LTUs) are terminal devices specifically designed for monitoring power distribution lines. They are used for high-precision data acquisition, monitoring, and communication, and generally do not have direct opening / closing control functions. Rail meters (smart meters) are modular metering devices that can be installed on standard rails for accurate metering and monitoring. They typically do not have opening / closing control functions.

[0070] It is worth noting that this invention deploys the integrated sensing and calculation module uniquely at the meter box entrance. The system does not need to obtain the original data of each individual user's meter, nor does it need to install sensors at intermediate nodes such as JP cabinets and branch boxes. It only needs to install a smart measurement switch or a meter box branch LTU / rail meter sensing module at the main inlet of the meter box. With just this set of aggregated data at the entrance of each meter box, the necessary electrical characteristic input can be provided for subsequent topology identification and impedance calculation, realizing low-cost transformation of the transformer area with "one test per box".

[0071] Step 2, as follows Figure 3 As shown, the communication delay characteristics and signal attenuation characteristics of each meter obtained based on HPLC+HRF dual-mode communication are clustered. All meters in the same cluster are assigned to the same meter box according to the maximum membership degree, and all meters in the same meter box are aggregated into a meter box node.

[0072] Furthermore, the specific steps for step 2 to identify the meter box are as follows:

[0073] Step 2.1: The electrical characteristics of Step 1 are encapsulated into a data frame by a concentrator installed on the transformer side via PLC communication protocol. The transmission time of the data frame is recorded by a network-wide synchronous timestamp counter based on a unified crystal oscillator. The transmission time of the data frame and the data frame are modulated by HPLC and then broadcast to each meter via power line.

[0074] Step 2.2: The meter obtains the data frame transmission time, records the data frame arrival time using the local timestamp, generates a response frame, records the response frame transmission time, modulates the response frame using HPLC, and sends it back to the concentrator. The concentrator receives the response frame and records the response frame reception time. The frequency deviation Δf of the transceiver crystal oscillator generated by multiple exchanges of data frames and response frames between the transceiver ends is expressed by the following formula:

[0075] (1)

[0076] In the formula, This represents the frequency deviation of the crystal oscillator of the i-th meter relative to the concentrator crystal oscillator. , Indicates the total number of meters. This represents the total number of interactions. The processing time measured by the i-th meter in the nth interaction using its locally biased clock is expressed by the following formula:

[0077] (2)

[0078] In the formula, This indicates the transmission time of the response frame of the i-th meter in the nth interaction. This represents the arrival time of the data frame of the i-th meter in the nth interaction. The remaining time after deducting the processing time of the i-th meter in the nth interaction from the total round-trip time is expressed by the following formula:

[0079] (3)

[0080] In the formula, This indicates the time when the concentrator receives the nth response frame from the i-th meter. This indicates the time when the concentrator sends the nth data frame to the i-th meter.

[0081] Based on the frequency deviation, T2 and T3 are corrected to be synchronized with the transmitter's time base, and are denoted as T2' and T3' respectively, as expressed by the following formula:

[0082] (4)

[0083] (5)

[0084] In the formula, This represents the correction value for the time when the i-th meter receives the data frame under the concentrator time base. Indicates the start time of the concentrator sending data frames. This indicates the time when the i-th meter receives the data frame. This represents the correction value of the response frame sent by the i-th meter under the concentrator time base. This indicates the time when the i-th meter sends a response frame.

[0085] Step 2.3: Using the correction value T2' of the received data frame time under the concentrator time reference and the correction value T3' of the transmitted response frame time under the concentrator time reference, calculate the one-way propagation delay characterizing the concentrator to each meter, expressed by the following formula:

[0086] (6)

[0087] In the formula, This represents the one-way propagation delay of the signal from the concentrator to the i-th meter, and characterizes the approximate physical distance between the concentrator and the meter. This represents the time it takes for the i-th meter to receive the response frame. The one-way propagation delay, based on the propagation time of the power line carrier signal, accurately measures the electrical path length from the concentrator to the meter and the total round-trip time. Subtract the meter's processing time at the precise time. Divide this by 2 to get the one-way propagation time. The value is proportional to the distance the signal travels on the power line and is a key quantity characterizing the approximate physical distance between the meter and the concentrator.

[0088] The communication delay characteristics of each electricity meter are generated based on the device asset ID of each meter and the one-way propagation delay from the concentrator to each meter. ,in, This represents the device asset ID of the i-th meter. This represents the total number of electricity meters.

[0089] Step 2.4: Based on HRF, the concentrator listens to the signals broadcast by all electricity meters through the HRF communication channel and extracts the signal attenuation characteristics of each meter. The signal attenuation characteristics include Received Signal Strength Indicator (RSSI) and signal attenuation features. , This represents the signal attenuation characteristic of the i-th meter. The signal attenuation characteristic is based on the attenuation of wireless signals during air propagation and reflects the linear spatial relationship between the meter and the concentrator, but it is affected by multipath effects such as walls and obstacles.

[0090] Step 2.5: Integrate communication delay features and signal attenuation features to generate a meter feature matrix. Use a fuzzy clustering algorithm to cluster the meter feature matrix and calculate the meter-box affiliation relationship to generate a list of meter boxes to which the meters belong.

[0091] More preferably, step 2.5 includes:

[0092] Step 2.5.1: Construct a multi-dimensional meter feature for each meter based on communication delay characteristics and signal attenuation characteristics. All the features of the multidimensional electricity meter constitute the multidimensional electricity meter feature matrix. .

[0093] Step 2.5.2: Set the number of clusters K, where K represents the number of bins; set the fuzzy weight index m, and randomly initialize an S×K membership matrix. ,in, The degree of membership of the i-th meter to the j-th meter box must satisfy the following condition: .

[0094] Step 2.5.3: Using the multidimensional meter features of all meters as input, a fuzzy clustering algorithm is used for iterative calculation to optimize the objective function. The algorithm minimizes the sum of weighted distances from the multidimensional features of all meters to each cluster center until it converges, yielding the final membership matrix. This matrix is ​​then set as the meter-meter box membership matrix, expressed by the following formula:

[0095] (7)

[0096] In the formula, Let S represent the objective function, which describes the sum of weighted distances from all meters to the center of their respective meter boxes. S represents the total number of meters, and K represents the number of meter boxes. Represents Euclidean distance. Represents the fuzzy weight index. , This represents the characteristics of the i-th meter. The center of the cluster corresponding to the j-th bin is represented by the following formula:

[0097] (8)

[0098] The membership degree of the i-th meter to the j-th meter box is expressed by the following formula:

[0099] (9)

[0100] Traverse the meter-box membership matrix row by row, determine which meter in each row belongs to the box corresponding to the highest membership degree, and generate a list of meter boxes to which the meters belong, expressed by the following formula:

[0101] (10)

[0102] In the formula, This indicates the box number to which the i-th meter belongs.

[0103] Meter box nodes are not directly taken from the physical meter boxes in the transformer area, but are logical nodes generated solely based on the clustering results. Specifically, meters within the same cluster (i.e., the same column in the membership matrix) are determined by the algorithm to belong to the same physical meter box because they are highly similar in HPLC delay and HRF attenuation characteristics. All meters determined to belong to the same meter box number are aggregated, thus logically forming a meter box node.

[0104] All meters determined to belong to the same meter box are aggregated to obtain a single meter box node. After performing membership determination on all S meters, a list of meter boxes to which the meters belong is obtained. This clearly identifies the physical meter box to which each meter belongs, thus completing the final identification of the customer-box relationship. The logical entity used to aggregate a physical meter box and all the meters under it is called a meter box node.

[0105] Step 3: Calculate the coupling impedance between any two meter box nodes using electrical characteristics; create a common parent node as a branch box node for the two meter box nodes corresponding to the maximum coupling impedance to replace the two meter box nodes corresponding to the maximum coupling impedance in the node set; calculate the coupling impedance between any two nodes in the node set; repeat the above process until only one branch box node remains in the node set; take this branch box node as the root branch box node, take the transformer node as the parent node of the root branch box node, and all meter box nodes are child nodes of the root branch box node; establish a four-level tree topology diagram of transformer-branch box-meter box node-meter in the transformer substation area.

[0106] In a preferred but non-limiting embodiment of the present invention, step 3 includes:

[0107] Step 3.1: Based on the list of meter boxes to which the meters belong, set all meters belonging to the same meter box as a meter box node; by monitoring the total current change of any meter box node and the voltage change caused by this current change on all other meter box nodes in the transformer area, calculate the coupling impedance between any two meter box nodes and construct the coupling impedance matrix Z.

[0108] More preferably, step 3.1 includes:

[0109] Step 3.1.1: Based on the list of meter boxes to which the meters belong, set all meters belonging to the same meter box as a meter box node, and monitor the total current change of each meter box node j. , The current changes of all its subordinate meters are aggregated, and the total current change at the current meter box node j is simultaneously measured at other meter box nodes within the distribution area. The voltage change caused by the above, j≠ .

[0110] Step 3.1.2, using the ratio of voltage change to total current change as the ratio of the two meter box nodes j, Coupling impedance between , This is represented by meter box node j and all other meter box nodes within the station area. Inter-coupling impedance, which characterizes the distance from meter box node j to all other meter box nodes within the station area. The electrical coupling strength is physically equal to the sum of the line impedances of the common path between the two nodes.

[0111] Step 3.1.3: Traverse all meter box nodes, obtain the coupling impedance of all meter box nodes, and form an N×N coupling impedance matrix, where N is the number of meter boxes, and the number of meter box nodes is the same as the number of meter boxes.

[0112] Step 3.2: As Figure 4 As shown, the maximum value of the coupling impedance is found in the coupling impedance matrix, the node corresponding to the maximum value is aggregated, and the branch box node corresponding to the aggregated maximum value node is created. The four-level tree topology diagram of transformer-branch box-meter box node-meter is recursively derived from bottom to top within the transformer area.

[0113] More preferably, step 3.2 includes:

[0114] Step 3.2.1, set the current node set Elements of the coupling impedance matrix Z Let j be the bin node and all bin nodes except j. Inter-coupling impedance, =0, This represents the self-impedance of node j in the meter box.

[0115] Find the maximum coupling impedance Zmax=Zpq in the current coupling impedance matrix Z. This indicates that among all the meter box nodes whose connection relationship has not yet been determined, the electrical connection between meter box node p and meter box node q is the tightest.

[0116] Step 3.2.2: Determine if the meter box node p corresponding to the maximum coupling impedance is directly connected to meter box node q. Create a common parent node for meter box nodes p and q, denoted as branch box node H, to represent a branch box and record the edges (H, p) and (H, q). The parent node created in this step represents a physically existing branch box that is not directly monitored, and is called a branch box node in the four-level tree topology diagram.

[0117] Step 3.2.3: Remove both table box node p and table box node q from the current node set C, and add branch box node H to the current node set C to obtain a new node set Cnew.

[0118] Step 3.2.4: Calculate the coupling impedance matrix between the branch box node H and the nodes in the new node set Cnew, expressed by the following formula:

[0119] (11)

[0120] in, This represents the branch box node H and the nodes in the new node set Cnew. The coupling impedance matrix, Indicates in At the same moment, synchronous measurement nodes The amount of voltage change. This is expressed as the total change in current at node H of the branch box. , This represents the change in current at node p of the meter box. This represents the change in current at node q of the meter box. The currents under the same branch are superimposed to generate a new coupling impedance matrix.

[0121] Step 3.2.5: Calculate the coupling impedance between branch box node H and each meter box node in the new node set Cnew to generate a new coupling impedance matrix. Repeat the above process for the new coupling impedance matrix. If the coupling impedance between branch box node H and meter box node p' is the largest, determine that the branch box node H corresponding to the maximum coupling impedance is directly connected to the meter box node p'. Create a common parent node for branch box node H and meter box node p', denoted as branch box node H'. Remove branch box node H and meter box node p' from the node set Cnew and add branch box node H' to the node set Cnew until only one branch box node remains in the final node set. Take the remaining branch box node as the root branch box node, set the parent node of the root branch box node as the transformer node, point the transformer node to the transformer, divide the iteratively merged branch box nodes into second-level topology nodes, divide the meter box nodes into third-level topology nodes, divide the original meters into fourth-level topology nodes, and output the four-level tree topology graph G=(V,E) of transformer-branch box-meter box node-meter. In this structure, the meter box nodes, branch box nodes, and transformer nodes in the recursive process form the node set of the fourth-level tree topology, and the physical lines connecting the nodes in the node set form the edge set of the fourth-level tree topology. V is the node set, containing meter box nodes, branch box nodes, and substation nodes; E is the edge set, representing the physical lines connecting the nodes. E = { }, each edge This corresponds to a segment of physical circuitry. This represents the total number of edges in the four-level tree topology graph G, which is the number of lines whose impedance is to be determined.

[0122] The initial data collected by the system comes only from the concentrator on the transformer side and the data acquisition terminal at the main incoming line of each meter box. The branch boxes located in the intermediate layer do not have data acquisition equipment installed. Therefore, the system cannot identify branch boxes by directly reading their asset ID or communication characteristics, as it can identify meters and meter boxes. The existence and connection relationships of branch boxes must be deduced by analyzing the electrical coupling relationships (coupling impedance) between downstream, identified meter box nodes, using the "maximum element recursion" algorithm. The "virtual" branch box node is a logical entity inferred to interpret the observed data, but it corresponds to a real power distribution equipment branch box in the physical world. The four-level tree topology diagram deduced by the algorithm is a causal connection diagram. This diagram is not a simple list of equipment, but accurately describes the actual power supply path of electrical energy from the transformer, through each level of branch box, and finally distributed to each meter box. If two meter boxes share the same upper-level branch box in the topology tree, it means that they are physically connected to different outgoing terminals of the same branch box. The biggest difficulty in the operation and maintenance of low-voltage distribution areas lies in the blind spots below the branch boxes. This invention makes these unmonitored intermediate nodes and their connection relationships visible by back-deriving the end data, thereby realizing full-link topology transparency from the transformer to the user's meter.

[0123] Because the branch box nodes are derived through strict electrical and physical relationships (coupling impedance), this topology map becomes a digital twin model connecting the physical and digital worlds. When a fault occurs downstream of a branch box, the system can immediately locate the specific branch box node and list all the meter boxes and users affected downstream, rather than just knowing a general transformer power supply range. It can refine line loss calculation and location to each segment of the line between "transformer-branch box" and "branch box-meter box", thereby accurately locating high-loss abnormal segments. The topology derived by the algorithm can be compared with the theoretical topology in the operation and maintenance file. When the two are inconsistent (for example, the derivation shows that a meter box is connected to branch box A, but the file records it in branch box B), an alarm can be triggered, driving operation and maintenance personnel to verify on-site, thereby correcting potentially erroneous records and achieving dynamic synchronization of the file.

[0124] The branch box node derived in this invention is a precise digital mapping of physically existing branch box equipment. The final four-level topology structure of "transformer-branch box-meter box-meter" is not a purely logical model, but a complete and accurate map reflecting the actual physical connections of the low-voltage distribution network. Its ultimate physical significance lies in overcoming the industry challenge of invisible blind spots in low-voltage distribution areas through digital derivation with minimal hardware cost (one test box per unit), and constructing a core data foundation to support the lean operation and service of the distribution network.

[0125] Step 3.3: Take the coupling impedance matrix between all nodes in the four-level tree topology diagram to generate the final coupling impedance matrix. Based on the final coupling impedance matrix, establish and solve the linear equation system to generate the final line impedance of each side in the four-level tree topology diagram.

[0126] More preferably, step 3.3 includes:

[0127] Step 3.3.1: Generate the final coupling impedance matrix by taking the coupling impedance matrices of all meter box nodes, branch box nodes, and transformer nodes in the node set of the four-level tree topology, and use the environmental attenuation factor. Correct the final coupling impedance matrix to generate the corrected coupling impedance matrix. . , Indicates the first A corrected coupling impedance Indicates the first The node and the first The coupling impedance matrix of each node.

[0128] Step 3.3.2: Based on the corrected coupling impedance matrix, establish a system of linear equations for each edge of the four-level tree topology graph, expressed by the following formula:

[0129] (12)

[0130] In the formula, Indicates the line impedance variable. ,in, This represents the line impedance of the m-th edge. , The total number of edges in a four-level tree topology graph G is expressed by the following formula:

[0131] (13)

[0132] In the formula, Indicates the first A corrected coupling impedance value Indicates the judgment of the first Is the edge located at the ? The node pair path corresponding to the corrected coupling impedance value. This represents connecting nodes in a four-level tree topology graph G. and The set of edges for a unique path. These represent the two box nodes corresponding to the o-th corrected coupling impedance value. , This indicates the total number of corrected coupling impedance values.

[0133] The path-edge incidence matrix is ​​represented by the following formula:

[0134] (14)

[0135] In the formula, express The The associated element in the m-th column of the nth row. This represents the m-th edge of a four-level tree topology graph G;

[0136] This represents the corrected coupling impedance value. This represents the o-th corrected coupling impedance value.

[0137] Step 3.3.3: Solve the linear equations using the least squares method to obtain the line impedance. ,in, Let represent the initial line impedance of the m-th edge in the four-level tree topology graph G. , This represents the total number of edges in a four-level tree topology graph G.

[0138] Step 3.3.4: Determine the theoretical coupling impedance based on the line impedance, expressed by the following formula:

[0139] (15)

[0140] In the formula, Indicates the first The node and the first Theoretical coupling impedance of each node.

[0141] When the error between the theoretical coupling impedance and the corresponding impedance in the final coupling impedance matrix exceeds a set threshold, the environmental attenuation factor is corrected, expressed by the following formula:

[0142] (16)

[0143] In the formula, This represents the updated environmental degradation factor. This represents the environmental decay factor used in the current iteration. This represents the learning rate, with a value of 0.3. The median of the impedance ratios in the theoretical and final coupling impedance matrices represents the median of the corresponding impedance ratios. These ratios are expressed by the following formula:

[0144] (17)

[0145] In the formula, Indicates the first The node and the first The theoretical coupling impedance of each node and the corresponding impedance ratio in the final coupling impedance matrix. Indicates the first The node and the first The coupling impedance matrix of each node.

[0146] Recalculate the line impedance until the error between the theoretical coupling impedance and the corresponding impedance in the final coupling impedance matrix is ​​less than a set threshold, and then output the final line impedance.

[0147] A four-level tree-structured topology diagram, representing the transformer-branch box-meter box node-meter system, is logically derived using a topology analysis method based on impedance segmentation calculation. Based on electrical characteristic information, topology identification using impedance calculation, and relying on load characteristic quantity transmission relationships, and building upon the identification of user-transformer relationships, a hierarchical recursive strategy is employed to deduce the branch box node of the low-voltage distribution network. This achieves highly reliable topology-impedance joint modeling using only end-user data from the transformer substation, ultimately realizing the four-level tree-structured topology diagram derivation.

[0148] By adopting a minimal data acquisition and digital extrapolation approach, only intelligent data acquisition devices need to be installed at the meter boxes, significantly reducing the cost of digital transformation of low-voltage distribution areas. Through clustering algorithms and impedance-based topology extrapolation, a complete four-level topology analysis is achieved, building a digital foundation for low-voltage distribution areas and providing a basis for refined management. This helps to achieve transparent and digital management and maintenance capabilities for distribution areas. Simultaneously, accurate line impedance provides a data foundation for addressing a series of pain points, such as precise power outage location, abnormal line loss diagnosis, anti-theft analysis, heavy overload warning, and orderly charging. By calibrating the transceiver crystal oscillator deviation through a network-wide synchronous timestamp counter, distance measurement accuracy is improved, and precise line impedance is output synchronously—a feature not found in most existing topology identification methods. Impedance parameters are direct inputs for advanced analytical applications such as power flow calculation, precise line loss analysis, fault location, power supply capacity assessment, and network optimization. Line impedance greatly enhances the back-end application potential and practical value of topology identification results.

[0149] Step 3.4: Assign each meter in the meter box list as a child node to the corresponding meter box node in the four-level tree topology graph output in Step 3.3, based on the meter box to which it belongs.

[0150] More preferably, step 3.4 includes:

[0151] Based on the list of meter boxes to which the meters belong, extract all the sets of meters corresponding to each meter box. Add a field for associated meter sets to the meter box node attributes of the four-level tree topology graph G, and enter all the meters in all the meter sets corresponding to the meter boxes. Embed the final line impedance of each edge in the four-level tree topology graph G into the edge attributes of the four-level tree topology graph G, such as... Figure 5 As shown.

[0152] Based on the meter box to which the meter belongs, extract all the sets of meters corresponding to each meter box, add a field for associated meter set to the meter box node attributes in the four-level tree topology graph G, and enter all the meter asset IDs in all the meter sets corresponding to the meter box to realize the direct association between the meter box affiliation and the meter box node in the tree topology graph.

[0153] Assign a unique topology node code to each transformer node, branch box node (H1, H2…Hn), and meter box node (e.g., “T001” represents the transformer, “B001” represents the branch box, and “M001” represents the meter box), and record the hierarchical relationship of each node (e.g., “T001→B001→M001”).

[0154] Mark the corresponding topology path (e.g., Z1=10mΩ, Z2=40mΩ) for the final line impedance of each edge in the four-level tree topology diagram.

[0155] Based on the meter box to which the meter belongs, extract all the sets of meters corresponding to each meter box. Add a field for associated meter set to the meter box node attributes in the four-level tree topology graph G. Enter all the meter asset IDs of all the meter sets corresponding to the meter box, such as path P1: T001→B001→M001, path P2: T001→B002→M002. Embed the final line impedance of each edge in the four-level tree topology graph into the path attributes of the four-level tree topology graph G. For example, in the “T001→B001” segment of path P1, mark the impedance Z1=10mΩ, and in the “B001→M001” segment, mark the impedance Z2=40mΩ.

[0156] Verify splicing consistency: For meter box node Mj, through the path impedance of its superior branch box Bk, combined with the meter box membership matrix R, verify whether the relationship between the voltage change ΔU and the current change ΔI of all meters in the meter box satisfies ΔU=ΔI×(total path impedance). If the deviation exceeds ±5%, recalibrate the final line impedance and path matching relationship of each side.

[0157] Generate a four-level tree topology diagram of transformer-branch box-meter box-meter.

[0158] Add attribute labels to each connection relationship in the four-level tree topology diagram of transformer-branch box-meter box-meter, including: the final line impedance of each edge, the node associated asset ID, and the membership degree of meter box and meter, to form a complete and traceable four-level tree topology diagram.

[0159] Step 4: Generate the physical topology map of the low-voltage distribution area based on the four-level tree topology map.

[0160] In a preferred but non-limiting embodiment of the present invention, step 4 includes:

[0161] Step 4.1: Based on the type of node (transformer, branch box, meter box, electricity meter), filter the corresponding physical equipment attributes in the equipment ledger. Using the equipment asset ID of the node in the four-level tree topology diagram as the unique key, match the physical equipment attributes and write the matched physical equipment attributes (such as asset ID, model) into the attribute field of the corresponding node in the four-level tree topology diagram to form a four-level tree topology diagram with bound physical equipment attributes.

[0162] More preferably, step 4.1 includes:

[0163] The equipment ledger contains the asset ID, model, specifications, and other attributes of all physical equipment. The asset ID is globally unique. When the concentrator leaves the factory, it has a built-in unique asset ID for itself and all nodes (transformers, branch boxes, meter boxes, meters) within its jurisdiction (such as "T-001-transformer", "B-002-branch", "M-003-box", "E-004-household"). The code of each node in the four-level tree topology diagram is associated with the corresponding equipment asset ID in the equipment ledger to form a node code-equipment asset ID mapping table, which is stored in the concentrator's local database.

[0164] Using the node hierarchy of the four-level tree topology (transformer → branch box → meter box → meter) as the index, the "node code-asset ID" mapping table is called to match the physical equipment asset ID corresponding to each node. The physical equipment entity is locked by the uniqueness of the asset ID (globally unique code), ensuring a one-to-one correspondence between nodes and physical equipment.

[0165] Each node in the four-level tree topology diagram that has been bound to physical device attributes is associated with at least one asset ID and other ledger attributes, establishing a one-to-one correspondence between nodes and physical devices.

[0166] Step 4.2: The concentrator extracts the BeiDou positioning data of the nodes, combines the hierarchical relationship in the four-level tree topology diagram with the physical distance calculated by the line impedance, verifies the spatial rationality of the device connection relationship in the four-level tree topology diagram with bound physical device attributes, and outputs a spatially verified four-level tree topology diagram to verify the spatial rationality of the positioning information.

[0167] More preferably, step 4.2 includes:

[0168] All nodes (transformers, branch boxes, and meter boxes) are equipped with remote BeiDou positioning to collect latitude and longitude (accuracy ±0.1m) and elevation data of the installation location in real time. The data is then uploaded to the concentrator via HPLC+HRF dual channels. The concentrator standardizes the positioning data (e.g., converts it to the WGS84 coordinate system) and binds it to the equipment asset ID.

[0169] The location data of each node (transformer, branch box, meter box) is added as a spatial location attribute to the four-level tree topology diagram that has been bound to the physical device attributes.

[0170] In a GIS system, the actual spatial straight-line distance D between parent and child nodes in a four-level tree topology map with bound physical device attributes is calculated based on the coordinates of the parent and child nodes. 实际 (e.g., the logic path impedance of the branch box and the meter box corresponds to the actual line length).

[0171] Based on the final line impedance of each side and the resistance per unit length of the line determined by the model specifications, the theoretical physical length L of this line segment is estimated. 理论 Set the error range (e.g., [0.8]). L 理论 1.2 L 理论 ]).

[0172] Judge D 实际 Whether it is within the error range, for example, verify whether the actual distance from the branch box to its subordinate meter box basically matches the length calculated by the impedance. If it is within the error range, pass the verification. If D 实际 If the location is outside the error range, the spatial location is determined to be abnormal. In the four-level tree topology diagram with bound physical device attributes, the connection line between the corresponding parent and child nodes is marked as abnormal, and an alarm log is generated, indicating that there may be an error in the device location entry, the actual topology has been changed, or the impedance calculation is abnormal.

[0173] For example, in a four-level tree topology diagram, the physical distance calculated by the path impedance of "branch box B-002 → meter box M-003" is 200m. Therefore, the straight-line distance between the two for BeiDou positioning needs to be within the range of 180m-220m (allowing ±10% error). If it exceeds the threshold, a file correction reminder will be triggered.

[0174] The concentrator synchronizes the distribution area file configuration information (including transformer power supply range, branch box wiring method, meter box ownership, user file and other operating parameters) from the distribution network PMS system through a dedicated power communication link. It then performs a preliminary verification with the locally stored equipment asset ID and location information, removes abnormal data that does not match the file (such as equipment asset ID not existing, location outside the power supply range, etc.), and generates a spatially verified four-level tree topology map.

[0175] The spatially validated four-level tree topology graph contains coordinate information and verifies the spatial rationality of all connections; abnormal connections have been marked.

[0176] Step 4.3: Compare the consistency of the spatially validated four-level tree topology diagram with the configuration information of the transformer area archive, and output the transformer-branch-box-household physical topology model.

[0177] For example, in the four-level tree topology diagram, "meter box M-003" belongs to "branch box B-002", which must be consistent with the configuration information in the transformer area file that "the superior branch box of M-003 is B-002". At the same time, the user file corresponding to the meter asset ID must match the power supply range of the meter box to ensure that the physical affiliation of "meter box-user" is correct.

[0178] If a conflict is detected, the file is marked as mismatched in the spatially validated four-level tree topology diagram, and a conflict report is generated. This conflict is not directly modified; instead, a manual verification work order is triggered, and maintenance personnel verify the issue on-site to determine whether to correct the file or re-identify it, thus forming a closed loop for dynamic file updates.

[0179] Step 4.4: Visualize and render the transformer-branch-box-household physical topology model on the GIS map and attach attributes. Finally, output a visualized physical topology map of the transformer-branch-box-household low-voltage distribution area that includes complete equipment, connection relationships, attributes and anomaly markers.

[0180] More preferably, step 4.4 includes:

[0181] By calling the Geographic Information System (GIS) engine, the location information of each node in the transformer-branch-box-household physical topology model is drawn on the electronic map. According to the connection relationship in the transformer-branch-box-household physical topology model, connection lines are drawn between nodes (different voltage levels or types are represented by different colored lines), realizing the visual mapping of "logical level - physical location".

[0182] Distinguish between icons for different equipment types such as transformers, branch boxes, meter boxes, and meters; clicking on any node icon will bring up an attribute panel that displays all information such as asset ID, model, real-time electrical characteristics (if any), upstream impedance, and branch to which it belongs;

[0183] Devices or connections marked as having spatial anomalies or file conflicts are highlighted (e.g., flashing, red).

[0184] The system integrates application functions such as topology analysis, power supply range analysis, power outage simulation, and line loss calculation into a visual interface.

[0185] When the equipment location changes (such as meter box relocation) or the topology is adjusted, Beidou positioning uploads the new location data in real time. The concentrator automatically triggers the mapping verification process to update the association between asset ID and location information and the physical topology map in the transformer-branch-box-customer physical topology model. If a conflict is detected between the asset ID, location data, and file configuration (such as the location being outside the file's power supply range), an abnormality prompt is pushed through the concentrator's local alarm module. After verification and correction by maintenance personnel, the mapping is re-completed.

[0186] Embodiment 2 of the present invention provides an automatic topology identification system for low-pressure transparent stages based on HPLC+HRF dual-mode communication and electrical characteristics, which runs the automatic topology identification method for low-pressure transparent stages based on HPLC+HRF dual-mode communication and electrical characteristics described in Embodiment 1, including:

[0187] The membership generation module is used to obtain the electrical characteristics of each meter data; it clusters the communication delay characteristics and signal attenuation characteristics of each meter obtained based on HPLC+HRF dual-mode communication, and assigns all meters in the same cluster to the same meter box according to the maximum membership degree, and aggregates all meters in the same meter box into a meter box node.

[0188] The four-level tree topology generation module is used to calculate the coupling impedance between any two meter box nodes using electrical characteristics; it creates a common parent node as a branch box node for the two meter box nodes corresponding to the maximum coupling impedance to replace the two meter box nodes corresponding to the maximum coupling impedance in the node set; it calculates the coupling impedance between any two nodes in the node set; it repeats the above process until only one branch box node remains in the node set; it takes this branch box node as the root branch box node; it takes the transformer node as the parent node of the root branch box node; and all meter box nodes are child nodes of the root branch box node; thus, a four-level tree topology of transformer-branch box-meter box node-meter is established within the transformer substation area.

[0189] The physical topology generation module is used to generate a physical topology map of the low-voltage distribution area based on a four-level tree topology map.

[0190] Embodiment 3 of the present invention takes the analysis of meter box nodes in low-voltage distribution areas as an example to identify a four-level tree topology diagram of transformer-branch box-meter box node-meter, such as... Figure 3 The diagram illustrates the specific process of implementing an impedance-based topology identification method in a concrete example.

[0191] There is a low-voltage distribution area with 7 meter boxes under it. The topology is unknown. The 7 meter box nodes are named 1, 2, 3, 4, 5, 6, and 7.

[0192] Assume all branch lines are resistive, single-phase circuits, and all branch loads are resistive. A user's electricity consumption within node 1 causes... As the current I1 increases, the voltage drop caused by nodes 1, 2, 3, 4, 5, 6, and 7 of the meter box is respectively: U11、 U12 U13 U14 U15 U16 U17; Similarly, the voltage drop at each node caused by the increased current due to user electricity consumption within nodes 2-7 of the meter box can be obtained. Uij(i, j=1, 2, 3, 4, 5, 6, 7), This represents the voltage change synchronously measured at the j-th node when a current change ΔIi occurs at the i-th node. This represents the total current change at the i-th meter box node.

[0193] Define the effect of a change in the current at a certain node on the voltage at other nodes based on the coupling impedance, according to the formula. calculate, Let represent the coupling impedance from node i to node j, which is solved by the sum of all line impedances from the transformer to their nearest common connection point. This impedance represents the voltage change ΔUj at node j caused by a unit change ΔIi in the total current at node i. Therefore, we introduce the following 7×7 coupling impedance matrix, as shown in Table 1:

[0194] Table 1 Coupling Impedance Matrix

[0195]

[0196] Following the maximum element recursion principle, the coupling matrix is ​​decomposed and derived, yielding the lowest-level minimum branch consisting of meter box nodes 6 and 7, and branch box node H4. Further derivation from low to high levels reveals branch box nodes H2 and H3. Branch box node H2 connects to meter box nodes 1, 2, and 3, while branch box node H3 connects to meter box nodes 4, 5, 6, and 7. Both branch box nodes H2 and H3 are connected to the transformer via branch box node H1. This leads to a precise four-level tree topology diagram of transformer-branch box-meter box node-meter.

[0197] The line impedance values ​​on each line in the low-voltage topology can be realized by using the low-voltage topology and the coupling impedance matrix.

[0198] When a user in node 1 consumes electricity, the voltage change at node 2 is related to... Regarding the coupling impedance between node 1 and node 2 The common line impedance between node 1 and node 2 is given by The calculation yields the following result. Similarly, the coupling impedance between node 1 and node 4 is... This leads to the following system of equations:

[0199] (18)

[0200] In the formula, , and These represent the voltage changes measured at nodes 1, 2, and 4 respectively when a current ΔI1 is injected into node 1. , and These represent the impedances of the segments from the transformer to nodes 1, 2, and 4, respectively.

[0201] The line impedance can be obtained by solving the above system of equations. , , Similarly, the impedance values ​​of the remaining lines can be obtained, and the impedance results of each line are shown in Table 2 below:

[0202] Table 2 Line Impedance

[0203]

[0204] Embodiment 4 of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded onto the processor, it implements an automatic topology identification method for low-pressure transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics as described in Embodiment 1.

[0205] Embodiment 5 of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements an automatic topology identification method for low-pressure transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics as described in Embodiment 1.

[0206] Compared with the prior art, the beneficial effects of the present invention include at least the following:

[0207] Based on HPLC+HRF dual-mode communication, the communication delay characteristics obtained by HPLC reflect the electrical path length of the signal from the concentrator to the meter. However, relying solely on HPLC may misclassify electrically proximate meters as spatially proximate. The signal attenuation characteristics obtained by HRF reflect the linear spatial distance between the meter and the concentrator. Relying solely on HRF may misclassify spatially proximate meters as electrically proximate. Meters located in the same meter box will exhibit high consistency in both electrical and spatial dimensions, which are completely independent. For meters truly belonging to the same box, they are close to each other both electrically and spatially, thus obtaining a high overall membership degree pointing to the same group. For meters that are misclassified as electrically proximate but spatially distant, or spatially proximate but electrically distant, significant inconsistencies will be observed in the other dimension, leading to a decrease in their overall membership degree and classification into other, more suitable groups. Therefore, this invention utilizes HPLC's susceptibility to electrical topology. To address the issues of confusion and misjudgment due to spatial layout confusion in HRF, this paper integrates communication delay characteristics obtained from HPLC and signal attenuation characteristics obtained from HRF. Through fuzzy clustering fusion, meters are strongly identified as belonging to the same group only when both electrical paths and spatial locations are consistent. This reduces the risk of initial clustering errors caused by interference in a single communication channel (such as power line noise or wireless signal obstruction) or special grid structures (such as long branch lines or complex wiring). This improves the accuracy and reliability of identifying the meter and meter box affiliation relationships for users. Since subsequent electrical deduction heavily relies on the correctness of the initial clustering, the improved robustness of the meter and meter box affiliation relationships directly reduces the systemic risk of the entire topology identification system failing due to front-end errors. This enhances the practicality and stability of the four-level tree topology diagram of transformer-branch box-meter box-meter, and improves and ensures the effectiveness of the final low-voltage distribution area physical topology diagram.

[0208] By utilizing the existing HPLC / HRF dual-mode communication channel of the existing electricity meter to acquire features, there is no need to deploy additional dedicated sensors or acquisition equipment at intermediate nodes such as branch boxes and JP cabinets or at each user's electricity meter. This greatly reduces the complexity and overall cost of hardware procurement, installation and maintenance, making large-scale, universal digital transformation of low-voltage distribution areas economically feasible.

[0209] By using dual-feature fusion, it is ensured that all meters within a meter box node are electrically connected in parallel to the same physical connection point. The minute current changes of each meter are aggregated into a total current change, which is electrically equivalent to a current disturbance occurring at the main incoming line point of the meter box. When a current change is injected into node j, according to circuit theory, the voltage change it causes at node j' is determined by the transfer impedance between the two. In a radial distribution network, this transfer impedance is exactly equal to the sum of the path impedances from the transformer to the nearest common connection point between the two nodes. Therefore, the coupling impedance accurately quantifies the electrical distance between nodes j and j' in the topology tree. Equal coupling impedances indicate the same electrical distance, while maximum coupling impedance indicates the same electrical distance. With the shortest air distance, this invention utilizes the synchronous electrical quantity measurement of the terminal meter box node to reverse deduce the complete intermediate layer topology and line parameters. This significantly reduces the hardware deployment density, complexity, and cost required to achieve complete topology identification, enabling transparency without modifying the intermediate network and simply through terminal intelligence. It reduces the dependence on intermediate layer monitoring hardware and sensitivity to environmental noise, achieving high-reliability topology identification in low-cost and complex environments. If the clustering is incorrect, the total current change becomes a set of disturbances injected from multiple discrete points in the power grid. The voltage disturbance field generated cannot be described by a simple impedance model between pairs of nodes, causing the calculation of the entire coupling impedance matrix to fail and making topology deduction impossible.

[0210] Based on HPLC / HRF dual-mode communication, as long as the current and voltage changes are measured accurately, the coupling impedance matrix is ​​reliable, and the derived topology is accurate. This enables it to handle transformer substations with large differences in line length, complex wiring, and harsh communication environments. By solving the coupling impedance through strict time synchronization of current and voltage changes, the sensitivity of the identification results to complex field communication environments and spatial structures is reduced. This ensures that the coupling impedance has a clear physical meaning and is calculable, enhancing the robustness and universality of the four-level tree topology. The four-level tree topology of transformer-branch box-meter box-meter improves the applicability and reliability of results in complex real-world scenarios. If the current and voltage changes are not strictly time-synchronized, they may originate from two completely unrelated independent electrical events. In this case, the coupling impedance is just a meaningless random ratio, losing all physical meaning.

[0211] The complex problem of network topology inference is transformed into a deterministic process that relies solely on the maximum search and iterative merging of the coupling impedance matrix. It does not require presetting the number of network layers or branch boxes, nor does it require complex graph search or probabilistic models. It automatically generates a complete tree structure from the data, including automatically discovering and creating branch box nodes of all intermediate levels, thus improving the automation level of topology identification.

[0212] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0213] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. An automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics, characterized in that: Acquire the electrical characteristics of each meter data; cluster the communication delay characteristics and signal attenuation characteristics of each meter acquired based on HPLC+HRF dual-mode communication, and assign all meters in the same cluster to the same meter box according to the maximum membership degree, and aggregate all meters in the same meter box into a meter box node; Calculate the coupling impedance between any two meter box nodes using electrical characteristics; create a common parent node as a branch box node for the two meter box nodes corresponding to the maximum coupling impedance to replace the two meter box nodes corresponding to the maximum coupling impedance in the node set; calculate the coupling impedance between any two nodes in the node set; repeat the above process until only one branch box node remains in the node set; take this branch box node as the root branch box node; take the transformer node as the parent node of the root branch box node; and take all meter box nodes as child nodes of the root branch box node; establish a four-level tree topology diagram of transformer-branch box-meter box node-meter in the transformer substation area. Establishing a four-level tree topology diagram of transformer-branch box-meter box node-meter includes: Find the maximum value of the coupling impedance in the current coupling impedance matrix, and determine that the meter box node p corresponding to the maximum value of the coupling impedance is directly connected to the meter box node q. Create a common parent node for both table box node p and table box node q, denoted as branch box node H; After removing box node p and box node q from node set C, add branch box node H to node set C to obtain a new node set Cnew; Calculate the coupling impedance between branch box node H and each meter box node in the new node set Cnew to generate a new coupling impedance matrix; repeat the above process for the new coupling impedance matrix. If the coupling impedance between branch box node H and meter box node p' is the largest, determine that the branch box node H corresponding to the maximum coupling impedance is directly connected to meter box node p'; ​​create a common parent node for branch box node H and meter box node p', denoted as branch box node H'. After removing branch box node H and meter box node p' from the node set Cnew, add branch box node H' to the node set Cnew until only one branch box node remains in the final node set. Take the remaining branch box node as the root branch box node, set the parent node of the root branch box node as the transformer node, point the transformer node to the transformer, and output the four-level tree topology graph of transformer-branch box-meter box node-meter. The table box node, branch box node, and transformer node in the recursive process are the node set of the fourth-level tree topology graph, and the physical lines between the nodes in the node set are the edge set of the fourth-level tree topology graph. Calculate the final line impedance of each edge in the level 4 tree topology graph, and label the final line impedance as an attribute on the corresponding connecting edge in the level 4 tree topology structure. The final line impedance of each edge in the level 4 tree topology graph includes: The coupling impedance matrix of all meter box nodes, branch box nodes and transformer nodes in the node set of the four-level tree topology graph of transformer-branch box-meter box node-meter is generated to generate a final coupling impedance matrix, and an environmental attenuation factor is adopted The final coupling impedance matrix is corrected to generate a corrected coupling impedance matrix Based on the corrected coupling impedance matrix, a system of linear equations is established for each edge of the four-level tree topology. The least squares method is used to solve the system of linear equations to obtain the line impedance. The theoretical coupling impedance is determined based on the line impedance. When the error between the theoretical coupling impedance and the corresponding impedance in the final coupling impedance matrix is ​​greater than a set threshold, the environmental attenuation factor is corrected and the line impedance is recalculated until the error between the theoretical coupling impedance and the corresponding impedance in the final coupling impedance matrix is ​​less than the set threshold, and the final line impedance is output. A physical topology map of the low-voltage distribution area is generated based on the four-level tree topology map.

2. The automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics according to claim 1, characterized in that: The electrical characteristics of electricity meter data include: voltage and current; Based on HPLC+HRF dual-mode communication, the communication delay characteristics and signal attenuation characteristics of each meter were obtained, including: The electrical characteristics are encapsulated into data frames by a concentrator, and the data frame transmission time and data frames are modulated by HPLC and then sent to each electricity meter. The meter acquires the arrival time of the data frame, generates a response frame, and sends the response frame with the transmission time via HPLC back to the concentrator. The concentrator then acquires the reception time of the response frame. Based on the frequency deviation of the transceiver crystal oscillator generated by multiple interactions of data frames and response frames between the transceiver and the transceiver, the arrival time of the data frame and the transmission time of the response frame are corrected. Calculate the one-way propagation delay from the concentrator to each meter based on the corrected data frame arrival time, the corrected response frame transmission time, the data frame transmission time, and the response frame reception time. The communication delay characteristics of each electricity meter are generated based on the device asset ID of each meter and the one-way propagation delay from the concentrator to each meter. Based on HRF, the signal attenuation characteristics of each meter are extracted by listening to the signals broadcast by the meters.

3. The automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics according to claim 2, characterized in that: Frequency deviation is expressed by the following formula: wherein, represents the frequency deviation of the i-th meter's crystal relative to the concentrator's crystal, , represents the total number of meters, represents the total number of interactions, represents the processing time measured by the i-th meter's locally biased clock in the n-th interaction, represents the remaining time after subtracting the i-th meter's own processing time in the n-th interaction from the total round-trip time.

4. The automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics according to claim 1, characterized in that: Clustering was performed on the communication delay and signal attenuation characteristics of each meter obtained based on HPLC+HRF dual-mode communication, including: Based on communication delay characteristics and signal attenuation characteristics, a multi-dimensional meter feature is constructed for each meter. Using the multidimensional meter features of all meters as input, a fuzzy clustering algorithm is used for iterative calculation. By optimizing the objective function to minimize the sum of weighted distances from the multidimensional meter features of all meters to each cluster center, the meter-meter box membership matrix is ​​obtained. Traverse the meter-box membership matrix row by row, determine that all meters in a row belong to the box corresponding to the maximum membership value, and aggregate all meters that are determined to belong to the same box to obtain a box node.

5. The automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics according to claim 1, characterized in that: Real-time acquisition of any bin node The total current change and the total current change at other meter box nodes within the transformer area The voltage change caused by the above. The ratio of voltage change to total current change is used as the reference value for the two meter box nodes. , The coupling impedance between them is used to construct a coupling impedance matrix.

6. The automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics according to claim 5, characterized in that: The coupling impedance between branch box node H and each meter box node in the new node set Cnew is calculated and expressed by the following formula: in, This represents the branch box node H and the nodes in the new node set Cnew. The coupling impedance matrix, Indicates in At the same moment, synchronous measurement nodes The amount of voltage change. This is expressed as the total change in current at branch box node H; The branch box nodes generated by the iterative merging are divided into secondary topology nodes, the meter box nodes are divided into tertiary topology nodes, and the original meters are divided into quaternary topology nodes.

7. The automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics according to claim 6, characterized in that: The theoretical coupling impedance is determined based on the line impedance, and is expressed by the following formula: In the formula, Indicates the first The node and the first The theoretical coupling impedance of each node, Let represent the initial line impedance of the m-th edge in the four-level tree topology graph G. This represents the m-th edge of a four-level tree topology graph G. This represents connecting nodes in a four-level tree topology graph G. and The set of edges for a unique path.

8. The automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics according to claim 7, characterized in that: The environmental degradation factor is corrected and expressed by the following formula: In the formula, This represents the updated environmental degradation factor. This represents the environmental decay factor used in the current iteration. Indicates the learning rate. This represents the ratio of the corresponding impedance in all theoretical coupling impedances and the final coupling impedance matrix. }'s median, Indicates the first The node and the first The theoretical coupling impedance of each node and the corresponding impedance ratio in the final coupling impedance matrix are expressed by the following formula: In the formula, Indicates the first The node and the first The theoretical coupling impedance of each node, Indicates the first The node and the first The coupling impedance matrix of each node.

9. The automatic topology identification method for low-voltage transparent stations based on HPLC+HRF dual-mode communication and electrical characteristics according to claim 8, characterized in that: Each node in the four-level tree topology diagram includes meter box nodes, branch box nodes, and transformer nodes; Generating a physical topology map of the low-voltage distribution area based on a four-level tree topology map includes: In the equipment ledger, filter the physical equipment attributes corresponding to each node in the four-level tree topology diagram. Use the equipment asset ID of the node in the four-level tree topology diagram as the unique key to match the physical equipment attributes. Write the matched physical equipment attributes into the attribute field of the corresponding node in the four-level tree topology diagram to form a four-level tree topology diagram with bound physical equipment attributes. The concentrator extracts BeiDou positioning data from each node in the four-level tree topology diagram. Combining the hierarchical relationship and line impedance in the four-level tree topology diagram, it verifies the spatial rationality of the device connection relationship in the four-level tree topology diagram with bound physical device attributes. It then outputs a spatially verified four-level tree topology diagram to verify the spatial rationality of the positioning information. The four-level tree topology diagram that has undergone spatial verification is compared with the configuration information of the transformer area archive to output the transformer-branch-box-household physical topology model. The physical topology model of transformer-branch-box-household is visualized and rendered on the GIS map, and its attributes are attached. Finally, the physical topology map of the low-voltage distribution area is output.

10. An automatic topology identification system for low-pressure transparent stages based on HPLC+HRF dual-mode communication and electrical characteristics, comprising the automatic topology identification method for low-pressure transparent stages based on HPLC+HRF dual-mode communication and electrical characteristics as described in any one of claims 1-9, characterized in that: The membership generation module is used to obtain the electrical characteristics of each meter data; it clusters the communication delay characteristics and signal attenuation characteristics of each meter obtained based on HPLC+HRF dual-mode communication, and assigns all meters in the same cluster to the same meter box according to the maximum membership degree, and aggregates all meters in the same meter box into a meter box node. The four-level tree topology generation module is used to calculate the coupling impedance between any two meter box nodes using electrical characteristics; it creates a common parent node as a branch box node for the two meter box nodes corresponding to the maximum coupling impedance to replace the two meter box nodes corresponding to the maximum coupling impedance in the node set; it calculates the coupling impedance between any two nodes in the node set; it repeats the above process until only one branch box node remains in the node set; it takes this branch box node as the root branch box node; it takes the transformer node as the parent node of the root branch box node; and all meter box nodes are child nodes of the root branch box node; thus, a four-level tree topology of transformer-branch box-meter box node-meter is established within the transformer substation area. The physical topology generation module is used to generate a physical topology map of the low-voltage distribution area based on a four-level tree topology map.

Images