A power line topology identification method integrating dual-mode communication and Bluetooth technology

Abstract

This invention relates to the field of power grids and discloses a power line topology identification method integrating dual-mode communication and Bluetooth technology. The method includes the following steps: The communication master node (CCO) and communication sub-nodes (STAs) are powered on and initialized, automatically forming a high-speed power line carrier and low-power wireless hybrid communication network, establishing a dual-mode communication link; the CCO collects outage event information recorded by each STA through this link, and filters out the set of STA nodes that experienced outages within the same time window based on a timestamp matching algorithm, generating a preliminary electrical topology relationship; each node's Bluetooth module periodically broadcasts and records RSSI values, which the CCO then aggregates and spatially groups the STA nodes using a density clustering algorithm; further, the physical distance between key node pairs is obtained, optimizing the physical location grouping; finally, the CCO integrates electrical topology and physical location information and uses a graph theory algorithm to generate a complete power line topology map; this method achieves effective identification of the topology structure of low-voltage distribution substations, possessing both electrical and spatial dimensions of information.

Description

Technical Field

[0001] This invention relates to the field of power grids, and more specifically to a power line topology identification method that integrates dual-mode communication with Bluetooth technology. Background Technology

[0002] As the final link in the smart grid, the clarity of the topology of low-voltage distribution transformer areas directly affects the accurate implementation of core operations such as line loss analysis, fault location, load management, and remote fee control. However, due to factors such as changes in user information, equipment replacement, and line upgrades, the topology of distribution transformer areas often changes dynamically. Traditional manual investigation methods are inefficient and cannot guarantee accuracy. Therefore, there is an urgent need for an automatic, accurate, and efficient topology identification method.

[0003] Currently, existing low-voltage transformer area topology identification technologies can be mainly classified into the following categories:

[0004] The first category is methods based on the correlation of transformer area information, such as analyzing the correlation of power frequency zero-crossing sequences, power outage records, or hourly voltage curves. This method relies on the coverage quality and clock synchronization accuracy of communication equipment. The identification process involves a large amount of communication data and a long cycle, and is easily affected by load fluctuations. It is not effective for identification in complex scenarios such as common neutral lines or strongly coupled lines.

[0005] The second category is based on the method of injecting electrical signal distortion, which enhances the characteristics of the transformer area by artificially creating power frequency voltage or current distortion for identification. Although this method has high accuracy under certain conditions, actively interfering with the operation of the power grid poses risks to power supply quality, user equipment safety, and distribution network stability, violating the principle of "non-sensory" monitoring. Furthermore, it is costly to implement and the equipment is difficult to miniaturize.

[0006] The third category is based on big data analysis, which mainly utilizes data such as voltage, current, and power collected from electricity meters to solve topological relationships through various optimization algorithms. This method requires no additional hardware, but its recognition accuracy is highly dependent on the synchronization, precision, and completeness of the basic data. In real-world environments, it is easily affected by abnormal data, making its reliability difficult to guarantee and its universality poor.

[0007] Furthermore, the existing methods mentioned above mainly focus on the identification of electrical connections, and lack effective technical means to identify and present the important dimension of the actual location distribution of equipment in physical space. Summary of the Invention

[0008] The purpose of this invention is to provide a power line topology identification method that integrates dual-mode communication with Bluetooth technology, thereby solving the above-mentioned technical problems.

[0009] The objective of this invention can be achieved through the following technical solutions:

[0010] A power line topology identification method integrating dual-mode communication and Bluetooth technology, the method comprising the following steps:

[0011] S1. The communication master node CCO and the communication sub-node STA are powered on and initialized, automatically forming a high-speed power line carrier and low-power wireless hybrid communication network and establishing a dual-mode communication link;

[0012] S2 and CCO collect power outage event information recorded by all STA nodes through a dual-mode communication link, and filter out the set of STA nodes that experienced power outages within the same time window based on a timestamp matching algorithm, generating a preliminary electrical topology relationship of the line branches;

[0013] S3. Each node's Bluetooth module periodically broadcasts beacon frames and listens for beacon frames broadcast by surrounding nodes, recording the Received Signal Strength Indicator (RSSI) value. The CCO summarizes the RSSI information of neighboring nodes reported by all STA nodes, spatially groups the STA nodes, and obtains preliminary physical location grouping information.

[0014] S4. Obtain the physical distance between key node pairs. Combine the preliminary physical location grouping information obtained in step S3 with the optimization algorithm to obtain optimized physical location grouping information. The optimization algorithm used to calibrate the grouping boundary is the weighted least squares method.

[0015] S5 and CCO periodically summarize the electrical topology relationships and optimized physical location grouping information, and use graph theory algorithms to fuse the information to generate a complete power line topology map that includes both electrical connection relationships and physical spatial location relationships.

[0016] As a further technical solution, in step S2, a density clustering algorithm is used to spatially group the STA nodes. The density clustering algorithm is a noise-based density clustering (DBSCAN) algorithm. In step S4, the physical distance between key node pairs is obtained through Bluetooth time-of-flight (ToF) ranging technology.

[0017] As a further technical solution, the communication master node (CCO) is installed in the data acquisition terminal of the distribution area, and the communication sub-node (STA) is installed in the user's electricity meter; both the CCO and STA integrate a high-speed power line carrier HPLC communication module, a low-power wireless HRF communication module, and a Bluetooth module; the HPLC communication module operates in the frequency band of 0.7MHz to 12MHz, and the HRF communication module operates in the frequency band of 470MHz to 510MHz.

[0018] As a further technical solution, the power outage event information includes a precise timestamp. The timestamp matching algorithm determines power outage events whose timestamps fall within the same time window as related events by setting a time window threshold.

[0019] As a further technical solution, the method for obtaining the time window threshold includes:

[0020] Based on historical power outage event data or network performance data, obtain the power outage timestamp difference within the transformer area, which is formed by the combined effect of switch action time differences and communication transmission delays. standard deviation of distribution ;

[0021] The time window threshold According to the formula Calculated;

[0022] in This is a preset confidence factor, with a value range of 3-4.

[0023] As a further technical solution, the distribution standard deviation The methods for obtaining it include:

[0024] Step 1: Collect the outage timestamp differences of all STA node pairs that are determined to belong to the same related event in multiple historical power outage events within the transformer area. ;

[0025] Step 2: Calculate the difference between all timestamps The absolute value of the sequence is obtained. ,in The number of node pairs;

[0026] Step 3: Calculate the arithmetic mean of the absolute value sequence, denoted as . ;

[0027] Step 4: According to the formula Calculate the standard deviation of the distribution .

[0028] As a further technical solution, the distribution standard deviation is obtained based on the network performance data. Specific methods include:

[0029] The communication delay between the master node (CCO) and each child node (STA) is measured periodically or triggered using a dual-mode communication link. , obtain a containing A sample set of delayed data {T1,T2,..., };

[0030] Calculate the arithmetic mean of the communication delay sample set, denoted as . ;

[0031] According to the formula Calculate the standard deviation of communication delay and will As the difference of the power outage timestamp standard deviation of distribution Valuation.

[0032] As a further technical solution, when historical power outage event data within the transformer area is insufficient or missing, a method based on network performance data is used to obtain the distribution standard deviation. Valuation;

[0033] The determination criteria for insufficient or missing historical power outage event data are any of the following:

[0034] Within a preset statistical time period T, the total number of valid historical power outage events collected by the transformer substation is less than a preset threshold A;

[0035] Within a preset statistical time period T, the number of STA node pairs that can be identified as belonging to the same related event is less than a preset threshold B.

[0036] The sample size N of the calculated absolute value sequence of power outage timestamp differences is insufficient for effective standard deviation estimation, i.e., N < Nmin.

[0037] As a further technical solution, the method also includes a topology verification step:

[0038] CCO sends a targeted wake-up or status query command to the STA node of the designated branch line through the dual-mode communication link.

[0039] Based on the response status and response delay of the STA node, verify whether the electrical topology of this branch line matches the generated topology. Figure 1 To;

[0040] If there is a discrepancy, a topology re-identification process is triggered or the abnormal node is marked. The beneficial effects of this invention are:

[0041] (1) The communication of the present invention is highly reliable. It integrates HPLC and HRF dual-mode communication. HPLC uses power lines to provide stable and high-bandwidth backbone communication, while HRF makes up for the communication blind spots of HPLC under specific line conditions such as switch disconnection and high impedance. The two complement each other, improve the overall network coverage and communication success rate, and lay the foundation for reliable acquisition of topology data.

[0042] (2) The present invention has high recognition accuracy. By matching the power outage records with the precise timestamps of multiple energy meters, it can accurately capture simultaneous power outage events caused by the same switch action, such as the branch fuse blowing, thereby identifying the set of lines that are electrically belonging to the same branch with high accuracy. At the same time, Bluetooth-assisted positioning is introduced. Combined with Bluetooth signal strength RSSI clustering and Bluetooth time-of-flight (ToF) precise ranging, multi-dimensional calculation of physical distance between nodes is realized. DBSCAN clustering can initially group nearby devices by processing RSSI data. ToF ranging provides a precise distance benchmark for key nodes. Then, by weighted optimization, such as least squares calibration of clustering results, the influence of environmental interference on RSSI is effectively overcome, and the accuracy of physical location relationship judgment is greatly improved.

[0043] (3) The present invention is completely passive or active but harmless in the entire process of identifying power outage events. It does not require injecting any interference signals into the power grid or creating distortion, thus completely avoiding the impact of existing distortion injection methods on power grid safety and power supply quality, and ensuring the stable operation of user equipment and distribution network. Furthermore, it creatively integrates the electrical connection relationship from the power outage event with the physical spatial location relationship from Bluetooth positioning. The generated topology map not only includes the line connection logic, but also the physical distance information between devices, providing a richer and more three-dimensional view of the transformer area, meeting diverse needs such as line loss analysis, fault location, and operation and maintenance management. It does not rely on perfect clock synchronization or high-precision electrical quantity sampling, and has a strong adaptability to load fluctuations, common zero / coupled lines, and dynamic changes in transformer areas. The introduction of dual-mode communication and Bluetooth also reduces the dependence on a single communication method. Attached Figure Description

[0044] The invention will now be further described with reference to the accompanying drawings.

[0045] Figure 1 This is a schematic diagram of the topology of the low-voltage distribution area;

[0046] Figure 2 This is a schematic diagram of the disordered network configuration of HPLC and HRF.

[0047] Figure 3 Implementation process for topology;

[0048] Figure 4 This describes the topology information acquisition process. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] Please see Figures 1-4 As shown, this invention is a power line topology identification method integrating dual-mode communication and Bluetooth technology. The method includes the following steps:

[0051] S1. The communication master node CCO and the communication sub-node STA are powered on and initialized, automatically forming a high-speed power line carrier and low-power wireless hybrid communication network and establishing a dual-mode communication link;

[0052] S2 and CCO collect power outage event information recorded by all STA nodes through a dual-mode communication link, and filter out the set of STA nodes that experienced power outages within the same time window based on a timestamp matching algorithm, generating a preliminary electrical topology relationship of the line branches;

[0053] S3. Each node's Bluetooth module periodically broadcasts beacon frames and listens for beacon frames broadcast by surrounding nodes, recording the Received Signal Strength Indicator (RSSI) value. The CCO summarizes the RSSI information of neighboring nodes reported by all STA nodes, spatially groups the STA nodes, and obtains preliminary physical location grouping information.

[0054] S4. Obtain the physical distance between key node pairs. Combine the preliminary physical location grouping information obtained in step S3 with the optimization algorithm to obtain optimized physical location grouping information. The optimization algorithm used to calibrate the grouping boundary is the weighted least squares method.

[0055] S5 and CCO periodically summarize the electrical topology relationships and optimized physical location grouping information, and use graph theory algorithms to fuse the information to generate a complete power line topology map that includes both electrical connection relationships and physical spatial location relationships.

[0056] As one implementation method, the networking and switching logic of dual-mode communication HPLC+HRF is as follows:

[0057] First, network priority and initial link establishment:

[0058] The priority rule is: by default, HPLC communication links are established first, and HRF links are only enabled when HPLC links cannot be established or the communication quality does not meet the requirements;

[0059] The basis for this is that HPLC uses power line transmission, which does not require additional wiring, and its bandwidth (usually ≥2Mbps) is higher than that of HRF (usually ≤500kbps), making it suitable for transmitting medium-volume information such as power outage event timestamps and topology data; HRF is wireless communication, which can make up for the communication blind spots of HPLC in scenarios such as when the line switch is off, high impedance lines (such as old cables), and strong electromagnetic interference.

[0060] Initial link establishment process:

[0061] After the CCO is powered on, it first sends a network request frame to all STAs in the area through the HPLC module. The frame format conforms to the "Technical Requirements for Low Voltage Power Line Carrier Communication" DL / T1485-2015.

[0062] After receiving the HPLC networking request, if the STA can return an acknowledgment frame, that is, the received signal strength is ≥-80dBm, then an HPLC link is established with the CCO.

[0063] For STAs that do not respond to HPLC requests, the CCO switches to the 470-510MHz band of the HRF module, with channel selection following the "Technical Requirements for Low-Power Short-Range Radio Transmitting Equipment" GB / T37978-2019, and resends the networking request until an HRF link is established.

[0064] Secondly, communication quality assessment and link switching triggering conditions;

[0065] The HPLC communication quality assessment criteria are as follows: Communication quality is considered substandard if any of the following conditions are met within three consecutive communication cycles:

[0066] Communication success rate <90%, communication success rate is the number of successfully received / sent frames ÷ total number of frames;

[0067] Received signal strength RSSI consistently < -85 dBm;

[0068] Data transmission delay > 500ms. Data transmission delay is the time difference between when the CCO sends a request and when the STA receives a response.

[0069] The switching logic is as follows:

[0070] When the HPLC communication quality fails to meet the standard, the CCO immediately triggers dual-mode switching, automatically switching to the HRF link to transmit data; once the HPLC link recovers to the standard state, if the communication quality requirements are met for two consecutive cycles, it switches back to the HPLC link to ensure a balance between communication stability and bandwidth efficiency.

[0071] In this embodiment, an HPLC+HRF dual-mode communication network is established through power-on initialization of the CCO and STA. This fully combines the advantages of HPLC relying on power lines for stable backbone transmission and HRF filling communication blind spots, completely solving the problems of incomplete coverage and easy data transmission interruption under single communication methods, and providing a highly reliable communication foundation for subsequent topology data acquisition. In the electrical topology identification stage, related nodes are screened by matching the timestamps of power outage events. There is no need to inject interference signals into the power grid, completely avoiding the potential threats to power quality and equipment safety posed by traditional distortion injection methods. At the same time, it avoids the defects of correlation methods that rely on a large amount of communication data and are susceptible to load fluctuations, achieving safe and accurate preliminary electrical topology generation. In terms of physical location identification, the RSSI clustering and critical node ranging optimization of the Bluetooth module overcome the limitations of existing technologies that only focus on electrical connections and lack physical spatial dimensions. This supplements the topology map with actual device location information. Finally, the graph theory algorithm fuses the two types of information, generating a complete topology map that simultaneously covers electrical logic and physical distribution. This not only supports accurate location of high-loss branches in line loss analysis, but also helps to quickly locate physical locations during faults. Furthermore, it provides equipment layout basis for load management, comprehensively improving the efficiency of transformer area operation and maintenance and the reliability of remote services, and meeting the diverse and high-precision needs of low-voltage distribution transformer areas for topology information.

[0072] In step S2, a density clustering algorithm is used to spatially group the STA nodes. The density clustering algorithm is the noise-based density clustering DBSCAN algorithm. In step S4, the physical distance between key node pairs is obtained through Bluetooth Time-of-Flight (ToF) ranging technology.

[0073] In this embodiment, the DBSCAN density clustering algorithm is used to spatially group STA nodes, which can effectively filter neighboring nodes based on signal strength association, eliminate environmental noise interference, and achieve preliminary accurate grouping of physical locations. Bluetooth ToF ranging technology is used to obtain the physical distance between key nodes, and weighted least squares method is used to optimize the grouping boundary. This can overcome the problem that RSSI is easily affected by environmental factors such as occlusion and reflection, and significantly improve the accuracy of physical location grouping. The combination of the two not only solves the defect of fuzzy physical topology identification in the existing technology, but also provides highly reliable physical location data for subsequent topology fusion, further ensuring the accuracy of the final topology map and meeting the needs of the transformer area for physical spatial topology.

[0074] As one implementation method, the triggering timing and execution process triggering timing for Bluetooth ToF ranging are as follows:

[0075] To avoid unnecessary ranging overhead, only the following key nodes are targeted for triggering ToF ranging:

[0076] Boundary node pairs in DBSCAN clustering, i.e., STAs that simultaneously belong to the candidate ranges of two clusters.

[0077] For node pairs with large RSSI fluctuations, such as those with an RSSI difference of >10dBm from 5 consecutive data collections;

[0078] In an electrical topology, nodes on the same branch but with an RSSI indication distance greater than 10 meters may be misjudged due to signal reflection / obstruction.

[0079] The execution process is as follows:

[0080] The CCO sends a ToF ranging command to the target node pair (such as STA1 and STA2) via a dual-mode communication link. The command includes the ranging start time.

[0081] Two STAs simultaneously activate the Bluetooth ToF module at a specified time. STA1 sends a ranging request frame, and STA2 immediately returns a ranging response frame upon receiving it.

[0082] The two STAs record the time to send the request t1, the time to receive the response t2, the time to send the response t3, and the time to receive the request t4, and report the four timestamps to the CCO.

[0083] CCO calculates physical distance using a formula. : Where c is the speed of light, taken as 3 × 10⁻⁶. 8 m / s, need to deduct the Bluetooth module processing delay of 10μs.

[0084] As one implementation method, the logic associated with the setting method of the ε neighborhood, i.e. the distance threshold, in the density clustering DBSCAN algorithm is as follows: the ε neighborhood needs to correspond to the mapping relationship between Bluetooth RSSI and physical distance. The RSSI-distance model needs to be established through on-site calibration first, and then the ε value is calculated in reverse.

[0085] Calibration and value acquisition steps:

[0086] Select 3-5 typical locations within the transformer area, such as near the transformer, at the end of a branch, or in a middle building, and place STA simulation nodes at known physical distances, such as 1m, 5m, 10m, and 20m.

[0087] Collect the average RSSI value at different distances and establish a fitting model (such as a linear model: RSSI=−10nlgL+A, where n is the path loss exponent and L is the RSSI value at 1m).

[0088] Based on the actual layout of the transformer substation, such as the maximum meter spacing ≤20m, the physical distance corresponding to ε is set to 5m. The RSSI threshold of ε is inversely calculated by fitting the model. For example, when n=2.5 and L=-40dBm, the RSSI corresponding to 5m is -59dBm, so ε is set to -59dBm.

[0089] The method for setting the minimum number of samples, MinPts, is based on the following criteria:

[0090] MinPts must match the node density within the transformer area. Node density = total number of STAs in the transformer area ÷ transformer area coverage; for example, for a high-density transformer area, the node density should be ≥ 5 nodes / 100m. 2 For example, in urban residential areas: MinPts=4; in medium-density areas, the node density is 2-5 per 100m. 2 For example, in urban-rural fringe areas: MinPts=3; low-density areas, node density <2 / 100m 2 For example, in rural transformer substations: MinPts=2; the basis is that MinPts needs to ensure that truly adjacent nodes can form clusters, while excluding interference from isolated nodes such as faulty offline STAs. The above value can cover more than 90% of low-voltage transformer substation scenarios.

[0091] The communication master node (CCO) is installed in the data acquisition terminal of the distribution area, and the communication sub-node (STA) is installed in the user's electricity meter. Both the CCO and STA integrate a high-speed power line carrier (HPLC) communication module, a low-power wireless (HRF) communication module, and a Bluetooth module. The HPLC communication module operates in the frequency band of 0.7MHz to 12MHz, and the HRF communication module operates in the frequency band of 470MHz to 510MHz.

[0092] In this embodiment, the CCO and STA are installed at the data acquisition terminal and the energy meter in the distribution area, respectively, which conforms to the actual layout of low-voltage distribution area equipment. This eliminates the need for a large number of additional dedicated devices, reducing deployment costs. The integrated HPLC module can achieve stable backbone communication via power lines, while the HRF module can supplement and cover HPLC communication blind spots. Together, they improve communication network coverage and data transmission reliability. The Bluetooth module supports near-field node discovery, signal strength measurement, and precise ranging, providing hardware support for physical location identification. This architecture solves the problems of poor communication reliability and lack of hardware foundation for physical location identification in existing technologies, laying the equipment foundation for efficient topology identification and ensuring the smooth progress of subsequent data acquisition and analysis.

[0093] The power outage event information includes a precise timestamp. The timestamp matching algorithm determines power outage events whose timestamps fall within the same time window as related events by setting a time window threshold.

[0094] In this embodiment, by using precise timestamps and time window thresholds to determine power outage events within the same time window, the set of STA nodes belonging to the same power supply branch can be accurately screened, effectively avoiding interference from factors such as load fluctuations and communication delays on electrical topology identification. Compared with methods based on transformer area information correlation, it does not require a large amount of communication data transmission and is not affected by clock synchronization accuracy, coupled lines, or other scenarios. The generated preliminary electrical topology relationship of line branches is more accurate, providing a reliable electrical connection foundation for subsequent topology fusion. It can support accurate location of high-loss points in line loss analysis and rapid identification of faulty branches in case of faults, solving the problems of low accuracy and susceptibility to interference in existing electrical topology identification methods.

[0095] The method for obtaining the time window threshold includes:

[0096] Based on historical power outage event data or network performance data, obtain the power outage timestamp difference within the transformer area, which is formed by the combined effect of switch action time differences and communication transmission delays. standard deviation of distribution ;

[0097] The time window threshold According to the formula Calculated;

[0098] in This is a preset confidence factor, with a value range of 3-4.

[0099] In this embodiment, the time window threshold is calculated based on historical power outage event data or network performance data. This ensures that the time window setting no longer relies on subjective experience but conforms to the actual power outage patterns and communication transmission characteristics of the transformer area, thus having a scientific basis. The introduction of confidence factors can adapt to the stability requirements of different transformer areas, ensuring that the time window can accurately encompass nodes of the same power outage event while excluding interference from unrelated power outage nodes. This method solves the problems of vague time window judgment criteria and poor applicability in existing technologies, improves the accuracy of screening for the same power outage event, and thus ensures the reliability of the initial electrical topology relationship, providing high-quality electrical data support for subsequent topology information fusion.

[0100] The standard deviation of the distribution The methods for obtaining it include:

[0101] Step 1: Collect the outage timestamp differences of all STA node pairs that are determined to belong to the same related event in multiple historical power outage events within the transformer area. ;

[0102] Step 2: Calculate the difference between all timestamps The absolute value of the sequence is obtained. ,in The number of node pairs;

[0103] Step 3: Calculate the arithmetic mean of the absolute value sequence, denoted as . ;

[0104] Step 4: According to the formula Calculate the standard deviation of the distribution .

[0105] In this embodiment, by collecting the timestamp differences of STA node pairs for the same related event in multiple historical power outage events, the mean and standard deviation of the absolute value sequence are calculated, so that... The method acquires power outage data covering multiple scenarios, accurately reflecting the timestamp difference pattern under the combined effects of switch action time differences and communication transmission delays within the transformer area. The calculation process is rigorous, avoiding deviations caused by single data or extreme values, providing a reliable basis for the accurate calculation of time window thresholds. This method solves the problems of the lack of a clear process for calculating the distribution standard deviation and the low reliability of the results in existing technologies, further improving the effectiveness of time window threshold setting, ensuring the accuracy of the judgment of the same power outage event, and strengthening the rigor of electrical topology identification.

[0106] The standard deviation of the distribution is obtained based on the network performance data. Specific methods include:

[0107] The communication delay between the master node (CCO) and each child node (STA) is measured periodically or triggered using a dual-mode communication link. , obtain a containing A sample set of delayed data {T1,T2,..., };

[0108] Calculate the arithmetic mean of the communication delay sample set, denoted as . ;

[0109] According to the formula Calculate the standard deviation of communication delay and will As the difference of the power outage timestamp standard deviation of distribution Valuation.

[0110] In this embodiment, when historical outage event data is lacking, the communication delay between the CCO and STA is measured through a dual-mode communication link. The standard deviation of the outage timestamp difference is estimated using the standard deviation of the communication delay, effectively solving the problem of threshold calculation stagnation caused by insufficient historical data. This estimation method relies on real-time available network performance data and does not depend on occasional outage events, making the time window threshold calculation continuous and feasible. It is suitable for newly commissioned transformer substations or transformer substations with extremely low outage frequency, avoiding the shortcomings of existing technologies that cannot perform topology identification when data is missing. It ensures that time windows can be accurately set for transformer substations under different data conditions, maintaining the accuracy and stability of electrical topology identification.

[0111] When historical power outage event data within the aforementioned distribution area is insufficient or missing, a method based on network performance data is used to obtain the distribution standard deviation. Valuation;

[0112] The determination criteria for insufficient or missing historical power outage event data are any of the following:

[0113] Within a preset statistical time period T, the total number of valid historical power outage events collected by the transformer substation is less than a preset threshold A. The preset threshold A, i.e., the threshold for the total number of valid historical power outage events, is the core threshold for determining whether historical power outage event data is sufficient; that is, the lower limit of the number of power outage events with complete timestamps collected within the statistical period T. As one implementation method, it is set to 5 times. This is because if there are fewer than 5 valid power outage events, it cannot cover different load conditions such as peak and off-peak periods, and different fault types such as switch tripping and branch fuse failure, which would lead to differences in the subsequently calculated power outage timestamps. Standard deviation of distribution The deviation is too large; 5 or more times can preliminarily reflect the time distribution pattern of power outage events in the transformer area and meet the statistical validity requirement.

[0114] Within a preset statistical time period T, the number of STA node pairs that can be identified as belonging to the same associated event is less than a preset threshold B. The preset threshold B is the threshold for the number of STA node pairs belonging to the same associated event; it is also the threshold for determining whether the node pair sample is sufficient in a single power outage event, i.e., the lower limit of the number of node pairs formed between STA nodes identified as belonging to the same branch in a power outage event, such as STA1 and STA2, STA1 and STA3. As one implementation method, it is set to 10 pairs; because if the number of node pairs is too small, such as less than 10 pairs, it will lead to… The sample size of the absolute value sequence is insufficient to accurately calculate the mean. and standard deviation Ten or more pairs can cover STAs at different locations within a branch, such as the beginning, middle, and end of the branch, reducing the impact of local bias on the statistical results.

[0115] The sample size N of the calculated absolute value sequence of power outage timestamp differences is insufficient for effective standard deviation estimation, i.e., N < Nmin. Nmin is... The absolute value sequence sample size threshold is used to determine... Does the absolute value sequence meet the lower limit of the sample size requirement for standard deviation calculation, i.e., the cumulative sample size across all historical power outage events? The lower limit for the total number of absolute value data; as one implementation method, it can be set to 30; because according to the large sample theory in statistics, when the sample size is ≥30, the data distribution can approximate a normal distribution, and the standard deviation calculated by the formula in this case... It has statistical properties; if N < 30, the standard deviation is susceptible to extreme values, such as those caused by an abnormal communication delay. An excessively large impact cannot reflect the true distribution pattern of timestamp differences.

[0116] This embodiment clarifies the criteria for determining insufficient or missing historical power outage event data. It specifically employs a method based on network performance data to obtain the estimated standard deviation of the distribution, ensuring that time window threshold calculations can still be performed normally in data-scarce scenarios such as new transformer substations and areas with fewer outages. This breaks the strong dependence of existing technologies on historical power outage data. Through clear criteria and corresponding solutions, the technical solution is applicable under different data conditions, avoiding interruptions in topology identification due to data issues. This effectively expands the application scope of the technology, enhances its adaptability to complex transformer substation environments, ensures the continuity and reliability of topology identification services, and solves the problem of existing methods being difficult to implement when data is insufficient.

[0117] The method also includes a topology verification step:

[0118] CCO sends a targeted wake-up or status query command to the STA node of the designated branch line through the dual-mode communication link.

[0119] Based on the response status and response delay of the STA node, verify whether the electrical topology of this branch line matches the generated topology. Figure 1 To;

[0120] If there is a discrepancy, the topology re-identification process will be triggered or the abnormal node will be marked.

[0121] Targeted wake-up target branch and STA selection rule branch selection:

[0122] Prioritize selecting the following branches based on electrical topology level to ensure verification coverage:

[0123] Main branch, that is, the line from the distribution transformer to the first branch switch;

[0124] The main branches with a load share of ≥20% are those branches that are prone to overload;

[0125] Branches with a history of ≥3 failures are prone to failure;

[0126] STA selection: Three representative STAs must be selected for each target branch:

[0127] The first meter at the branch start point (STA) is the first meter closest to the branch switch.

[0128] The meter at the midpoint of the branch (STA), which is located at 1 / 2 of the branch length;

[0129] The STA at the end of the branch refers to the meter at the furthest point of the branch.

[0130] The basis for this is that STAs in three locations can comprehensively reflect the communication connectivity of branches, avoiding misjudgments caused by local node failures.

[0131] Criteria for determining response status and response delay: Response status determination:

[0132] Normal response: The STA returns a status query response frame within 10 seconds, in which the power supply status and communication module status are both normal;

[0133] An abnormal response is defined as one that meets any of the following conditions:

[0134] If there is no response within 10 seconds, it may be due to a node being offline or a misjudgment of the topology.

[0135] The power supply status is abnormal in the response frame, but the electrical topology shows that the branch is not out of power;

[0136] The communication module status in the response frame is abnormal, such as a Bluetooth module malfunction.

[0137] Response delay determination:

[0138] Delay calculation: The time difference from when the CCO sends the status query command to when the STA receives the response, minus the fixed delay of dual-mode communication, such as the fixed delay of 20ms for HPLC;

[0139] Anomaly threshold: based on the historical average response latency of this branch. Calculate, threshold = +3× ,in The standard deviation of historical lag;

[0140] If the delay exceeds the threshold, it is determined that the topology does not match the actual communication path, and there may be branching connections.

[0141] Exception handling procedure if a single STA fails:

[0142] If the CCO resends the query command twice and the result is still abnormal, the STA is marked as a node to be investigated, and the overall topology re-identification is not triggered.

[0143] If ≥2 STAs are abnormal within the same branch: immediately trigger the local topology re-identification process, only re-identify the electrical topology and physical location of that branch, without needing to re-identify the entire station area, thus reducing resource consumption;

[0144] If the partial re-identification fails three times in a row, a maintenance work order will be generated, prompting on-site troubleshooting, such as line wiring or equipment failure.

[0145] In this embodiment, the CCO sends targeted wake-up or status query commands to the STAs of a designated branch, which can actively detect the consistency between the topology map and the actual line. Targeted wake-up is for a specific branch and can accurately verify the electrical connection relationship of the branch. If the topology map shows that a certain STA belongs to the branch, but it cannot be woken up or the response delay is abnormal, it indicates that there may be a topology misjudgment. Status query can obtain the real-time power supply and communication status of the STA and further verify whether the electrical connection is normal. Based on the judgment criteria of response status and delay, abnormal nodes or branches can be quickly identified, avoiding the use of incorrect topology information for business decisions.

[0146] The anomaly handling mechanism, which triggers re-identification or anomaly marking, can promptly correct deviations. For example, if multiple STAs on a branch are found to be unresponsive, the branch topology can be immediately re-identified to ensure that the topology information is consistent with reality. Marking a single abnormal node can remind maintenance personnel to conduct on-site troubleshooting, preventing the overall topology reliability from being affected by local node problems. This verification process makes topology identification no longer a one-time operation, but a dynamic process that continuously ensures accuracy. It avoids information becoming outdated due to static topology and can promptly eliminate identification errors, ensuring that the generated topology map always matches the actual situation of the transformer area. This not only enhances the support capability of topology information for core business, such as quickly locating fault points based on the verified topology during fault location and accurately locating high-loss branches based on accurate topology during line loss analysis, but also reduces the waste of maintenance costs caused by topology errors, improving transformer area management efficiency and business reliability.

[0147] It should be noted that the calculation formulas and all parameters involved in the calculations in this invention have been dimensionless beforehand. The process of dimensionless processing is well known in the industry and will not be described here.

[0148] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A power line topology identification method integrating dual-mode communication and Bluetooth technology, characterized in that, The method includes the following steps: S1. The communication master node CCO and the communication sub-node STA are powered on and initialized, automatically forming a high-speed power line carrier and low-power wireless hybrid communication network and establishing a dual-mode communication link; S2 and CCO collect power outage event information recorded by all STA nodes through a dual-mode communication link, and filter out the set of STA nodes that experienced power outages within the same time window based on a timestamp matching algorithm, generating a preliminary electrical topology relationship of the line branches; S3. Each node's Bluetooth module periodically broadcasts beacon frames and listens for beacon frames broadcast by surrounding nodes, recording the Received Signal Strength Indicator (RSSI) value. The CCO summarizes the RSSI information of neighboring nodes reported by all STA nodes, spatially groups the STA nodes, and obtains preliminary physical location grouping information. S4. Obtain the physical distance between key node pairs, and combine it with the preliminary physical location grouping information obtained in step S3 to obtain optimized physical location grouping information through an optimization algorithm; S5 and CCO periodically summarize the electrical topology relationships and optimized physical location grouping information, and use graph theory algorithms to fuse the information to generate a complete power line topology map that includes both electrical connection relationships and physical spatial location relationships. The power outage event information includes a precise timestamp. The timestamp matching algorithm determines power outage events whose timestamps fall within the same time window as related events by setting a time window threshold. The method for obtaining the time window threshold includes: Based on historical power outage event data or network performance data, obtain the power outage timestamp difference within the transformer area, which is formed by the combined effect of switch action time differences and communication transmission delays. standard deviation of distribution ; The time window threshold According to the formula Calculated; in This is a preset confidence factor, with a value range of 3-4; The standard deviation of the distribution The methods for obtaining it include: Step 1: Collect the outage timestamp differences of all STA node pairs that are determined to belong to the same related event in multiple historical power outage events within the transformer area. ; Step 2: Calculate the difference between all timestamps The absolute value of the sequence is obtained. ,in The number of node pairs; Step 3: Calculate the arithmetic mean of the absolute value sequence, denoted as . ; Step 4: According to the formula Calculate the standard deviation of the distribution ; The standard deviation of the distribution is obtained based on the network performance data. Specific methods include: The communication delay between the master node (CCO) and each child node (STA) is measured periodically or triggered using a dual-mode communication link. , obtain a containing A sample set of delayed data {T1,T2,..., }; Calculate the arithmetic mean of the communication delay sample set, denoted as . ; According to the formula Calculate the standard deviation of communication delay and will As the difference of the power outage timestamp standard deviation of distribution Valuation.

2. The power line topology identification method based on dual-mode communication and Bluetooth technology according to claim 1, characterized in that, In step S2, density clustering algorithm is used to spatially group STA nodes; in step S4, Bluetooth Time-of-Flight (ToF) ranging technology is used to obtain the physical distance between key node pairs.

3. The power line topology identification method based on dual-mode communication and Bluetooth technology according to claim 2, characterized in that, The communication master node (CCO) is installed in the data acquisition terminal of the distribution area, and the communication sub-node (STA) is installed in the user's electricity meter. Both the CCO and STA integrate a high-speed power line carrier (HPLC) communication module, a low-power wireless (HRF) communication module, and a Bluetooth module. The HPLC communication module operates in the frequency band of 0.7MHz to 12MHz, and the HRF communication module operates in the frequency band of 470MHz to 510MHz.

4. The power line topology identification method based on dual-mode communication and Bluetooth technology according to claim 1, characterized in that, When historical power outage event data within the aforementioned distribution area is insufficient or missing, a method based on network performance data is used to obtain the distribution standard deviation. Valuation; The criteria for determining insufficient or missing historical power outage data are any of the following: Within a preset statistical time period T, the total number of valid historical power outage events collected by the transformer substation is less than a preset threshold A; Within a preset statistical time period T, the number of STA node pairs that can be identified as belonging to the same related event is less than a preset threshold B. The sample size N of the calculated absolute value sequence of power outage timestamp differences is insufficient for effective standard deviation estimation, i.e., N < Nmin, where Nmin is... Absolute value sequence sample size threshold.

5. The power line topology identification method based on dual-mode communication fused with Bluetooth technology according to claim 1 or 4, characterized in that, The method also includes a topology verification step: CCO sends a targeted wake-up or status query command to the STA node of the designated branch line through the dual-mode communication link. Based on the response status and response delay of the STA node, verify whether the electrical topology of the branch line is consistent with the generated topology diagram; If there is a discrepancy, the topology re-identification process will be triggered or the abnormal node will be marked.

Images