A low-voltage transformer area electrical topology dynamic correction method

Abstract

The application discloses a low-voltage transformer area electrical topology dynamic correction method, comprising the following steps: multi-source data acquisition, collecting link signal quality data, electrical parameters and ranging data through the coordination of the main node, the relay node and the terminal node; multi-source data preprocessing, denoising and standardizing the link signal quality data, the electrical parameters and the ranging data; multi-dimensional fusion topology correction model construction, fusing the multi-source data based on the weighted D-S evidence theory, and combining a long short-term memory network time series prediction model to construct a topology correction model; topology dynamic updating and verification, setting a condition trigger update, after triggering, the edge node executes correction according to the topology correction model output, and verifies the correction result. The application solves the technical problems of traditional topology identification relying on a single data source, low static identification accuracy and poor dynamic load / interference adaptability.

Description

Technical Field

[0001] This invention relates to the field of power system low-voltage distribution area topology identification technology, and in particular to a method for dynamic correction of low-voltage distribution area electrical topology. Background Technology

[0002] With the continuous development of power systems, accurate identification and dynamic adaptation of electrical topology in low-voltage distribution areas has become a core requirement. However, traditional topology identification technologies are unable to meet these requirements due to several key shortcomings. First, traditional methods rely on a single data source, primarily on PLC / RF communication signal quality indicators such as SNR and RSSI, while neglecting electrical parameter data (such as current / voltage) and physical distance measurement information. This can easily lead to misjudgments of topology associations, such as misjudging different meter boxes that overlap signals only due to physical proximity as belonging to the same branch. Second, existing technologies mostly employ static update mechanisms with fixed cycles (such as every 3 hours), which cannot achieve real-time updates. The response time is insufficient to handle topology changes caused by sudden load changes such as charging pile start-up and shutdown or link interference such as power line noise, which is seriously inconsistent with the requirements of dynamic networking. Furthermore, traditional ranging technology relies on single communication delay information and does not effectively deduct power line noise interference, resulting in an accuracy of ≤30 meters, which is difficult to support the fine-level topology correction of the "transformer-branch box-meter box-user" in the distribution area. Finally, the identification results lack an effective verification loop and are not consistent with physical laws such as power balance of electrical parameters, which can easily lead to discrepancies between the topology map and the actual wiring, thereby affecting the accuracy of subsequent business applications such as line loss calculation and fault location.

[0003] Patent document CN202210416651.3 discloses a method and system for identifying low-voltage distribution network topology based on frozen data. This method involves obtaining the distribution of branch measurement terminals within the target distribution area; acquiring a set of frozen current data using current acquisition equipment; obtaining a set of desired classification values; obtaining target frozen current data information through numerical statistical solving; analyzing the target frozen current data information to generate line topology branch relationships; and optimizing based on a time-series development. Therefore, it is necessary to break through the static, single identification mode and construct a dynamic correction mechanism by integrating multi-dimensional data to promote the evolution of low-voltage distribution area topology identification technology towards a dynamic, multi-source direction. This aims to meet the requirements of high-quality communication and accurate topology management under the new power system, and proposes a dynamic correction method for low-voltage distribution area electrical topology. Summary of the Invention

[0004] The main objective of this invention is to propose a dynamic correction method for the electrical topology of low-voltage distribution areas, aiming to solve the technical problems of traditional topology identification relying on a single data source, low static identification accuracy, and poor dynamic load / interference adaptability.

[0005] To achieve the above objectives, the present invention provides a method for dynamic correction of the electrical topology of a low-voltage distribution area, wherein the method includes the following steps:

[0006] S1. Multi-source data acquisition: Through the coordination of relay nodes and terminal nodes by the main node of the transformer area, link signal quality data, electrical parameters and ranging data are acquired.

[0007] S2. Multi-source data preprocessing: Denoising and standardizing link signal quality data, electrical parameters and ranging data.

[0008] S3. Construction of a multi-dimensional fusion topology correction model: Based on the weighted DS evidence theory, multi-source data are fused, and a topology correction model is constructed by combining a long short-term memory network time series prediction model.

[0009] S4. Dynamic topology update and verification: Set conditions to trigger the update. After triggering, the edge nodes perform correction according to the topology correction model output and verify the correction results.

[0010] One preferred embodiment is that step S1 involves collecting link signal quality data, electrical parameters, and ranging data, specifically as follows:

[0011] Signal quality data actively reported at a first time interval, including the SNR, packet loss rate, and communication rate of the PLC link, as well as the received signal strength indication, transmit power, and channel occupancy rate of the RF link;

[0012] Electrical parameters reported at a second time interval, including the effective values ​​of current / voltage, phase difference, and active power at each node connection point;

[0013] Distance measurement data queried at the third time interval.

[0014] One preferred embodiment is that the ranging data acquisition is based on high-precision clock synchronization technology, specifically: the master node and each node are synchronized with GPS / BeiDou, the transmitting end injects ranging signal and records timestamp T1, the receiving end records timestamp T2, the propagation delay |T2-T1| is calculated, and the link processing delay is deducted to calculate the physical distance.

[0015] One preferred embodiment is that the noise reduction process in step S2 specifically includes:

[0016] Kalman filtering is used to filter the ranging data to suppress power line impulse noise, signal quality data is smoothed to reduce RF channel fluctuations, and electrical parameters are processed using 3D filtering. Outliers are eliminated based on accurate measurements.

[0017] One preferred embodiment is that step S2, the standardization process, specifically involves:

[0018] Z-score normalization is uniformly applied to signal quality data, electrical parameters, and ranging data; the signal quality data is then mapped to... The interval maps electrical parameters to The interval maps the ranging data to Interval.

[0019] One preferred embodiment, step S3, specifically includes:

[0020] S31. Weight allocation: Initial weights are assigned to signal quality data, electrical parameters and ranging data based on the analytic hierarchy process, and a dynamic adjustment mechanism is designed to make the weights adaptively updated according to the real-time link status.

[0021] S32. Multi-source fusion: Weighted DS evidence theory is used to fuse signal quality data, electrical parameters and ranging data to generate a topological correlation matrix between nodes.

[0022] S33. Dynamic correction: Introducing a long short-term memory network time-series prediction model, learning historical topological correlation data, and predicting topological change trends, and initiating a correction process for node pairs whose correlation fluctuations exceed a threshold in the prediction results.

[0023] One preferred embodiment is that the fusion rule of the weighted DS evidence theory is as follows:

[0024] Set the signal quality correlation to The weight is The correlation of electrical parameters is The weight is The ranging correlation is The weight is ;

[0025] Fusion correlation after fusion When the degree of fusion correlation If the correlation is ≥0.8, then the nodes are considered to be strongly correlated. If the value is less than 0.3, it is considered as having no correlation.

[0026] One preferred embodiment is that step S4 sets conditions to trigger the update, specifically as follows:

[0027] Setting conditions: signal quality fluctuations exceeding 10%, electrical parameter phase difference changes exceeding 5°, ranging data deviations exceeding 3 meters, or periodic timed triggering;

[0028] An update is triggered when any of the above conditions are met.

[0029] In one preferred embodiment, after step S4 is triggered, the edge node performs correction according to the topology correction model output, including adjusting the topology affiliation of the node and synchronously updating the topology layer in the GIS system.

[0030] One preferred embodiment is that step S4 verifies the correction result, specifically as follows:

[0031] Structural verification is performed through physical wiring checks of the transformer substations;

[0032] Data consistency verification is performed based on the principle of power balance of electrical parameters.

[0033] If the verification finds that the topological distance deviation exceeds the distance threshold, return to step S3 and re-execute the multi-dimensional fusion topological correction.

[0034] The above-described technical solution of this invention provides a method for dynamic correction of the electrical topology of low-voltage distribution areas, comprising the following steps: multi-source data acquisition, coordinating relay nodes and terminal nodes through the main node of the distribution area to acquire link signal quality data, electrical parameters, and ranging data; multi-source data preprocessing, performing noise reduction and standardization on the link signal quality data, electrical parameters, and ranging data; multi-dimensional fusion topology correction model construction, fusing multi-source data based on weighted DS evidence theory and combining it with a long short-term memory network time-series prediction model to construct a topology correction model; and dynamic topology update and verification, setting conditions to trigger updates, after which edge nodes execute corrections according to the output of the topology correction model and verify the correction results. This invention solves the technical problems of traditional topology identification relying on a single data source, low static identification accuracy, and poor dynamic load / interference adaptability.

[0035] In this invention, signal quality data, electrical parameters, and ranging data are integrated, overcoming the shortcomings of traditional methods that rely on a single data source. This avoids the problem of different meter boxes with overlapping signals due to their close physical location being misjudged as the same branch, thus improving the accuracy of topology identification.

[0036] In this invention, a dynamic update mechanism is adopted, which can trigger updates when parameter fluctuations exceed 10% or every 5 minutes, and respond in real time to topology changes caused by load mutations or link interference. This solves the problem that traditional static update mechanisms cannot adapt to dynamic environments and meets the needs of dynamic networking.

[0037] In this invention, high-precision ranging data acquisition based on high-precision clock synchronization technology has a ranging error of ≤2 meters, which can be used to make fine-level topology corrections such as transformer substation-branch box-meter box-user, which is a significant improvement compared to the traditional ranging accuracy of ≤30 meters.

[0038] In this invention, a closed-loop control is formed by verifying the actual wiring of the transformer area and verifying the power balance of electrical parameters, which ensures the accuracy of the topology relationship, avoids discrepancies between the topology diagram and the actual wiring, and guarantees the accuracy of subsequent business applications such as line loss calculation and fault location. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0040] Figure 1 This is a first schematic diagram of a dynamic correction method for the electrical topology of a low-voltage distribution area according to an embodiment of the present invention;

[0041] Figure 2 This is a schematic diagram of multi-source data acquisition according to an embodiment of the present invention;

[0042] Figure 3 This is a schematic diagram of multi-source data preprocessing according to an embodiment of the present invention;

[0043] Figure 4 This is a schematic diagram illustrating the construction of a multi-dimensional fusion topology correction model according to an embodiment of the present invention;

[0044] Figure 5 This is a schematic diagram illustrating the dynamic updating and verification of topology in an embodiment of the present invention.

[0045] The realization of the objective, functional characteristics and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0046] 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 a part of the embodiments of the present invention, and not all of them. 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.

[0047] Furthermore, in this invention, descriptions involving "first," "second," etc., are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature.

[0048] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0049] See Figures 1-5According to one aspect of the present invention, the present invention provides a method for dynamic correction of the electrical topology of a low-voltage distribution area, wherein the method for dynamic correction of the electrical topology of a low-voltage distribution area includes the following steps:

[0050] S1. Multi-source data acquisition: Through the coordination of relay nodes and terminal nodes by the main node of the transformer area, link signal quality data, electrical parameters and ranging data are acquired.

[0051] S2. Multi-source data preprocessing: Denoising and standardizing link signal quality data, electrical parameters and ranging data.

[0052] S3. Construction of a multi-dimensional fusion topology correction model: Based on the weighted DS evidence theory, multi-source data are fused, and a topology correction model is constructed by combining a long short-term memory network time series prediction model.

[0053] S4. Dynamic topology update and verification: Set conditions to trigger the update. After triggering, the edge nodes perform correction according to the topology correction model output and verify the correction results.

[0054] Specifically, in this embodiment, see Figure 2 Step S1, which involves collecting link signal quality data, electrical parameters, and ranging data, specifically includes:

[0055] Signal quality data is actively reported at a first time interval. The signal quality data includes the SNR, packet loss rate, and communication rate of the PLC link, as well as the received signal strength indication, transmit power, and channel occupancy rate of the RF link. Based on the asynchronous frequency hopping networking mechanism, node signal quality data is synchronously collected in the signaling time slot, and the received signal strength indication of the RF frequency hopping channel is collected in the service time slot and interference channels are eliminated. In this invention, the first time interval is 1 second, and link fluctuations are captured in real time at a frequency of 1 second.

[0056] The electrical parameters reported at a second time interval include the effective values ​​of current / voltage, phase difference, and active power at each node connection point; the second time interval is 5s, and the parameters are reported at a frequency of 5s to balance real-time performance and power consumption.

[0057] Ranging data queried at a third time interval; the third time interval is 10s; in this invention, the first time interval, the second time interval, and the third time interval do not need to be specifically limited, and can be set as needed; all data are uploaded to the edge computing node through a DTLS secure channel that meets the requirements of secure communication.

[0058] Specifically, in this embodiment, the ranging data acquisition is based on high-precision clock synchronization technology (time deviation ≤ 10ns). Specifically, the master node and each node are synchronized with GPS / BeiDou. The transmitting end injects a ranging signal and records a timestamp T1, and the receiving end records a timestamp T2. The propagation delay |T2-T1| is calculated, and the link processing delay is deducted. The link processing delay is 1ms. The physical distance is calculated. Based on the formula: physical distance = propagation speed * propagation delay / 2, a ranging result with an accuracy better than 2m is obtained. Data is collected every 10 seconds to avoid bandwidth occupation.

[0059] Specifically, in this embodiment, see Figure 3 The noise reduction process in step S2 specifically involves: using a Kalman filter to filter the ranging data and suppress power line impulse noise, ensuring that the SNR fluctuation is controlled within 3dB and the ranging data fluctuation does not exceed 0.5m; smoothing the signal quality data to reduce RF channel fluctuation; and applying a 3D filter to the electrical parameters. The system eliminates abnormal values ​​such as current peaks caused by sudden load changes, retaining current and voltage data that reflect normal load characteristics.

[0060] Specifically, in this embodiment, step S2, the standardization process, involves: uniformly normalizing the signal quality data, electrical parameters, and ranging data using Z-score, ensuring a mean of 0 and a standard deviation of 1; and mapping the signal quality data to... Ranges, such as mapping SNR (0-100dB) and RSSI (-110—-50dBm) to Map electrical parameters to An interval, such as mapping the voltage phase difference (0°-180°) to... The closer the value is to -1, the stronger the topological correlation, and the more the power (0-10kW) is mapped to... Mapping ranging data to Interval; through the above processing, all data are unified to the same dimension, providing standard input for subsequent fusion calculations.

[0061] Specifically, in this embodiment, see Figure 4 Step S3 specifically includes:

[0062] S31. Weight Allocation: Initial weights are assigned to signal quality data, electrical parameters, and ranging data based on the analytic hierarchy process (AHP), and a dynamic adjustment mechanism is designed to adaptively update the weights according to the real-time link status. In this invention, the initial weights of the signal quality data, electrical parameters, and ranging data are 0.35, 0.35, and 0.3, respectively. This invention does not impose specific limitations, and the weights can be dynamically adjusted according to the link status. For example, when the SNR of the PLC link is lower than 60dB, the weight of the ranging data is increased to 0.4, and the weight of the signal quality data is reduced to 0.3; when the phase difference of the electrical parameters fluctuates by more than 5°, the weight of the electrical parameters is increased to 0.4.

[0063] S32. Multi-source fusion: The weighted DS evidence theory is used to fuse signal quality data, electrical parameters and ranging data to generate a topological correlation matrix between nodes. Each element takes a value between 0 and 1 to quantify the correlation strength of node pairs. 1 indicates strong correlation, such as belonging to the same branch or bin.

[0064] S33. Dynamic correction: A long short-term memory network time-series prediction model is introduced to learn historical topology correlation data and predict topology change trends. For node pairs whose correlation fluctuations exceed a threshold in the prediction results, a correction process is initiated. The threshold is 10%. When the deviation between the predicted value and the current value exceeds 10%, the pre-correction mechanism is triggered in advance. The prediction error of the long short-term memory network time-series prediction model does not exceed 5%, which can effectively adapt to dynamic changes caused by load fluctuations and link interference.

[0065] Specifically, in this embodiment, the fusion rule of the weighted DS evidence theory is as follows:

[0066] Set the signal quality correlation to The weight is The correlation of electrical parameters is The weight is The ranging correlation is The weight is ;

[0067] Fusion correlation after fusion When the degree of fusion correlation If the correlation is ≥0.8, then the nodes are considered to be strongly correlated. If <0.3, it is considered as unrelated; if 0.3≤ If the value is less than 0.8, then the nodes are considered to be weakly related.

[0068] Specifically, in this embodiment, step S4, setting conditions to trigger an update, specifically involves:

[0069] Setting conditions: signal quality fluctuations exceeding 10%, electrical parameter phase difference changes exceeding 5°, ranging data deviations exceeding 3 meters, or periodic timed triggering; the periodic timed triggering means the system performs a forced update every 5 minutes to ensure the timeliness of topology data;

[0070] An update is triggered when any of the above conditions are met.

[0071] Specifically, in this embodiment, after step S4 is triggered, the edge node performs correction according to the topology correction model output, including adjusting the topology affiliation of the node and updating the topology layer in the GIS system simultaneously; for example, correcting three terminal nodes that were mistakenly identified as belonging to meter box C5 to meter box C6, and updating the GIS topology layer simultaneously, clearly marking the transformer-branch box-meter box-user hierarchical relationship and the node position deviation before and after correction, and uploading the correction results to the system in real time.

[0072] Specifically, in this embodiment, step S4, verifying the correction result, specifically involves:

[0073] Structural verification is performed by checking the physical wiring of the transformer substation; physical verification is performed by comparing the corrected topology with the actual wiring diagram, with a required correction accuracy of ≤5 meters.

[0074] Data consistency verification was performed based on the principle of power balance of electrical parameters; it was confirmed that the error between the total power of the branch box and the sum of the power of the lower meter box was within ±2%.

[0075] If the verification finds that the topological distance deviation exceeds the distance threshold, return to step S3 and re-execute the multi-dimensional fusion topological correction.

[0076] Specifically, in this embodiment, the power balance verification of electrical parameters in step S4 must meet the following condition: total power of the branch box P = ∑ power of the meter box Pi * (1 ± 2%), where 2% is the allowable range of metering error. If this condition is not met, the electrical parameters will be re-acquired and corrected.

[0077] Specifically, in this embodiment, the low-voltage distribution area electrical topology dynamic correction method further includes: step S5, an abnormal alarm mechanism, when the deviation of any node's distance measurement data exceeds 5 meters for 3 consecutive times or the signal quality SNR < 50dB, the edge node sends an early warning to the system, and the corresponding delay of the early warning is ≤ 1s.

[0078] Specifically, in this embodiment, the signal quality data acquisition of the RF link in step S1 is based on channel monitoring of asynchronous frequency hopping network. The received signal strength indication of each frequency hopping channel is collected in the service time slot, and the data of interfering channels (RSSI < -90dBm) is removed. RSSI is the received signal strength indication. The average value of the effective channels is taken as the final RSSI value, and the acquisition error is ≤2dBm.

[0079] Specifically, in this embodiment, taking a low-voltage distribution area (including 1 main node, 20 relay nodes, 480 terminal nodes, covering 500 households, including 10 charging piles and 5 photovoltaic inverters, adapted to the asynchronous frequency hopping network in Document 1) as an example, the implementation steps are as follows:

[0080] For multi-source data acquisition, in terms of signal quality acquisition, the master node is configured with a PLC link SNR acquisition threshold of ≥50dB and an RF link RSSI acquisition threshold of ≥-85dBm. It also synchronously acquires signal quality data from all terminal nodes in the first time slot (signaling time slot) of each superframe, with an acquisition frequency of 1 second / time. For electrical parameter acquisition, each terminal node integrates an ADE9078 metering chip to acquire parameters such as current (range 0-100A), voltage (220V±10%), and phase difference (0-180°) in real time, and uploads the data to the relay node at a frequency of 5 seconds / time. For ranging acquisition, the master node and each node achieve high-precision clock synchronization based on GPS timing (synchronization deviation ≤8ns). By sending a 100kHz ranging signal and accurately recording the sending timestamp T1 and receiving timestamp T2, the signal propagation delay is calculated, and high-precision ranging data is generated at a frequency of 10 seconds / time.

[0081] In the multi-source data preprocessing stage, the system performs denoising and standardization operations on the node data of terminal node 15 and terminal node 20 respectively. In the denoising process, Kalman filtering is used to smooth the ranging data of terminal node 15, reducing its fluctuation range from the original ±3 meters to ±0.4 meters. At the same time, the 3σ criterion is used to identify and remove abnormal current values ​​(such as abnormal peak values ​​of 200A) caused by the start and stop of charging piles in terminal node 20. In the standardization stage, the SNR (75dB) of terminal node 15 is normalized to 0.75, the voltage phase difference (8°) is mapped to -0.91, and the ranging data (8 meters) is calibrated to 8, thereby unifying the multi-source data to the same dimension and laying the foundation for subsequent fusion calculation.

[0082] A multi-dimensional fusion topology correction model was constructed. During the topology correction process, the system first performed weight allocation: since the current SNR of the PLC link is 72dB (good condition), the weights of signal quality, electrical parameters, and ranging data were set to 0.35, 0.35, and 0.3, respectively. Then, fusion calculations were performed: regarding the correlation between terminal node 15 and meter box C5, the signal quality correlation degree m1 = 0.8 (high SNR), the electrical parameter correlation degree m2 = 0.85 (small phase difference), and the ranging correlation degree m3 = 0.9 (physical distance only 8 meters). After weighted fusion, the comprehensive correlation degree m = 0.35 × 0.8 + 0.35 × 0.85 + 0.3 × 0.9 = 0.84, which was determined to be a strong correlation. Therefore, STA15 was correctly assigned to meter box C5. Finally, time series prediction was used for verification: the Long Short-Term Memory Network time series prediction model predicted based on the historical correlation sequence of terminal node 15 over the past 10 minutes, and found that its correlation in the next 5 minutes was 0.82, which was only 1.2% different from the current value and did not exceed the 10% pre-correction threshold, so no correction operation was required;

[0083] Topology dynamic update and verification: When the SNR of terminal node 30 drops from 70dB to 58dB (fluctuation of 17%) due to power line noise, the system triggers the event update mechanism. During the correction execution phase, the fusion model recalculates the topology correlation of terminal node 30: signal quality correlation m1=0.58 (low SNR), electrical parameter correlation m2=0.83 (phase difference remains stable), and ranging correlation m3=0.88 (distance remains unchanged). Due to the decrease in SNR, the system dynamically adjusts the weight to 0 for signal quality. 3. With electrical parameters of 0.35 and ranging of 0.35, the overall correlation after fusion is m=0.76, which falls into the category of weak correlation. Therefore, its assignment is corrected from the original meter box to the adjacent meter box C6. In the verification stage, the physical wiring check confirmed that the terminal node 30 does indeed belong to meter box C6, and the correction deviation is 3 meters (meeting the accuracy requirement of ≤5 meters). At the same time, the power balance verification of electrical parameters shows that the ratio of the total power of meter box C6 to the sum of the power of each STA connected to it is 1.01, and the error is within the allowable range of 2%. The verification result is qualified, and the correction is confirmed to be effective.

[0084] The above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. All equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method for dynamic correction of electrical topology in low-voltage distribution areas, characterized in that, Includes the following steps: S1. Multi-source data acquisition: Through the coordination of relay nodes and terminal nodes by the main node of the transformer area, link signal quality data, electrical parameters and ranging data are acquired. S2. Multi-source data preprocessing: Denoising and standardizing link signal quality data, electrical parameters and ranging data. S3. Construction of a multi-dimensional fusion topology correction model: Based on the weighted DS evidence theory, multi-source data are fused, and a topology correction model is constructed by combining a long short-term memory network time series prediction model. S4. Dynamic topology update and verification: Set conditions to trigger the update. After triggering, the edge nodes perform correction according to the topology correction model output and verify the correction results.

2. The method for dynamic correction of electrical topology in low-voltage distribution areas according to claim 1, characterized in that, Step S1, which involves collecting link signal quality data, electrical parameters, and ranging data, specifically includes: Signal quality data actively reported at a first time interval, including the SNR, packet loss rate, and communication rate of the PLC link, as well as the received signal strength indication, transmit power, and channel occupancy rate of the RF link; Electrical parameters reported at a second time interval, including the effective values ​​of current / voltage, phase difference, and active power at each node connection point; Distance measurement data queried at the third time interval.

3. A method for dynamic correction of electrical topology in low-voltage distribution areas according to any one of claims 1-2, characterized in that, The ranging data acquisition is based on high-precision clock synchronization technology. Specifically, the master node and each node are synchronized with GPS / BeiDou. The transmitting end injects ranging signals and records timestamp T1, the receiving end records timestamp T2, the propagation delay |T2-T1| is calculated, and the link processing delay is deducted to calculate the physical distance.

4. A method for dynamic correction of electrical topology in low-voltage distribution areas according to any one of claims 1-2, characterized in that, The noise reduction process in step S2 is specifically as follows: Kalman filtering is used to filter the ranging data to suppress power line impulse noise, signal quality data is smoothed to reduce RF channel fluctuations, and electrical parameters are processed using 3D filtering. Outliers are eliminated based on accurate measurements.

5. A method for dynamic correction of electrical topology in low-voltage distribution areas according to any one of claims 1-2, characterized in that, The standardization process in step S2 is as follows: Z-score normalization is uniformly applied to signal quality data, electrical parameters, and ranging data; the signal quality data is then mapped to... The interval maps electrical parameters to The interval maps the ranging data to Interval.

6. A method for dynamic correction of electrical topology in low-voltage distribution areas according to any one of claims 1-2, characterized in that, Step S3 specifically includes: S31. Weight allocation: Initial weights are assigned to signal quality data, electrical parameters and ranging data based on the analytic hierarchy process, and a dynamic adjustment mechanism is designed to make the weights adaptively updated according to the real-time link status. S32. Multi-source fusion: Weighted DS evidence theory is used to fuse signal quality data, electrical parameters and ranging data to generate a topological correlation matrix between nodes. S33. Dynamic correction: Introducing a long short-term memory network time-series prediction model, learning historical topological correlation data, and predicting topological change trends, and initiating a correction process for node pairs whose correlation fluctuations exceed a threshold in the prediction results.

7. The method for dynamic correction of electrical topology in low-voltage distribution areas according to claim 6, characterized in that, The fusion rule of the weighted DS evidence theory is as follows: Set the signal quality correlation to The weight is The correlation of electrical parameters is The weight is The ranging correlation is The weight is ; Fusion correlation after fusion When the degree of fusion correlation If the correlation is ≥0.8, then the nodes are considered to be strongly correlated. If the value is less than 0.3, it is considered as having no correlation.

8. A method for dynamic correction of electrical topology in low-voltage distribution areas according to any one of claims 1-2, characterized in that, Step S4, setting conditions to trigger the update, specifically involves: Setting conditions: signal quality fluctuations exceeding 10%, electrical parameter phase difference changes exceeding 5°, ranging data deviations exceeding 3 meters, or periodic timed triggering; An update is triggered when any of the above conditions are met.

9. A method for dynamic correction of electrical topology in low-voltage distribution areas according to any one of claims 1-2, characterized in that, After step S4 is triggered, the edge nodes perform corrections based on the topology correction model output, including adjusting the topology affiliation of the nodes and synchronously updating the topology layer in the GIS system.

10. A method for dynamic correction of electrical topology in low-voltage distribution areas according to any one of claims 1-2, characterized in that, Step S4, verifying the correction result, specifically involves: Structural verification is performed through physical wiring checks of the transformer substations; Data consistency verification is performed based on the principle of power balance of electrical parameters. If the verification finds that the topological distance deviation exceeds the distance threshold, return to step S3 and re-execute the multi-dimensional fusion topological correction.

Images