A cloud-edge collaborative transformer area line loss lean management method

Abstract

This invention discloses a cloud-edge collaborative method for lean management of transformer substation line loss, comprising: deploying intelligent fusion terminals at the edge of the transformer substation to collect power consumption data of key nodes; automatically identifying the physical topology of the transformer substation using pulse analysis and characteristic current injection methods; calculating graded line loss rates according to a four-level hierarchy of "transformer-outgoing line-branch box-meter box-household meter"; uploading the topology and line loss data to a cloud platform; verifying the line loss data on the cloud based on horizontal comparison of multiple transformer substations and vertical trend analysis of a single transformer substation, identifying abnormal line loss points; diagnosing the causes of anomalies by combining weather and load data, and generating remediation strategies; issuing the strategies to the intelligent fusion terminals and maintenance personnel terminals in the form of defect elimination work orders; verifying the effect by recalculating the line loss rate after defect elimination, forming a closed-loop management system; this invention achieves automatic identification of transformer substation topology, precise location of anomalies, intelligent diagnosis of causes, and closed-loop defect elimination through cloud-edge collaboration, significantly improving the real-time performance, accuracy, and efficiency of line loss management.

Description

Technical Field

[0001] This invention belongs to the field of power system automation and line loss management technology, specifically relating to a cloud-edge collaborative method for lean management of transformer area line losses. Background Technology

[0002] With the advancement of new power system construction, a large number of distributed power sources, energy storage devices, and charging piles are being connected to the distribution network, making the topology of transformer substations increasingly complex and the amount of data growing rapidly. Traditional centralized line loss management models face challenges such as poor real-time performance, slow response, and heavy data processing pressure. Currently, line loss management in transformer substations generally suffers from unclear relationships between customers and transformers, difficulty in locating line loss anomalies, and reliance on manual experience for diagnosing anomalies. The level of lean line loss management urgently needs to be improved.

[0003] Existing technologies include cloud-edge collaborative line loss analysis systems, such as the Chinese invention patent application CN114204674A entitled "A Cloud-Edge Collaborative Distribution Station Classified Line Loss Analysis System," which includes a cloud-based line loss analysis platform, an edge-side intelligent fusion terminal, and an energy data acquisition terminal. The energy data acquisition terminal is responsible for collecting electricity consumption information at various key nodes in the distribution station area and transmitting the collected information to the edge-side intelligent fusion terminal. The intelligent fusion terminal is responsible for calculating the distribution station topology on the edge side and calculating line losses in a graded manner based on the topology information. The cloud-based line loss analysis platform is used to verify the topology information on the edge side and, based on the graded line loss calculation values ​​on the edge side, to perform lean line loss management such as regional line loss calculation, abnormal line loss point location, and electricity theft analysis. This achieves line loss accounting at the transformer-low-voltage outgoing line and branch box-meter box levels. However, this method focuses more on system architecture and data acquisition and still has shortcomings in terms of the accuracy of automatic topology identification, precise location of abnormal line loss points, intelligent diagnosis of abnormal causes, and closed-loop management.

[0004] To address the aforementioned issues, it is essential to develop a cloud-edge collaborative lean management method for transformer substation line loss. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a cloud-edge collaborative method for lean management of distribution transformer area line losses. Through cloud-edge collaboration, it realizes automatic identification of transformer area topology, accurate location of anomalies, intelligent diagnosis of causes, and closed-loop defect elimination, which significantly improves the real-time performance, accuracy, and efficiency of line loss management. It is applicable to lean management of line losses in low-voltage distribution transformer areas under new power systems.

[0006] The objective of this invention is achieved as follows: a cloud-edge collaborative method for lean management of line loss in transformer substations, comprising the following steps: S1, deploy intelligent fusion terminals on the edge of the distribution area, collect raw power consumption data of each key node in the distribution area through the power data acquisition terminal of the distribution area. Key nodes include the low-voltage side of the distribution transformer, the incoming line of the branch box, the incoming line of the meter box and the user's power meter; S2, the intelligent fusion terminal, based on the original power consumption data, uses a combination of pulse analysis and characteristic current injection methods to automatically identify the physical topology of the low-voltage distribution network in the distribution area and generate the distribution area topology information. S3, the intelligent fusion terminal calculates the graded line loss rate of each level according to the transformer area topology information, according to the four levels of transformer-low voltage outgoing line, low voltage outgoing line-branch box, branch box-meter box, and meter box-household meter; S4, the intelligent fusion terminal uploads the transformer area topology information and graded line loss rate to the cloud-based line loss analysis platform; S5, the cloud-based line loss analysis platform receives and stores data from multiple intelligent converged terminals in multiple distribution areas, performs horizontal comparative analysis and vertical historical trend analysis based on data from multiple distribution areas, verifies the graded line loss rate, and identifies abnormal line loss points; S6, the cloud-based line loss analysis platform, diagnoses the causes of abnormalities and generates line loss mitigation strategies based on the location results of abnormal line loss points, combined with weather and load data; The S7 cloud-based line loss analysis platform distributes line loss management strategies to all integrated terminals and maintenance personnel terminals to implement closed-loop defect elimination management.

[0007] Preferably, the method also includes a data preprocessing step. After collecting the raw electricity consumption data, the intelligent fusion terminal first uses a Markov process missing data completion algorithm to fill in the missing values, and then uses a mean drift clustering algorithm to identify and correct abnormal data. Finally, the preprocessed data is used for line loss calculation.

[0008] Preferably, the process also includes a model update step, whereby the cloud-based line loss analysis platform periodically aggregates and optimizes the global line loss analysis model based on the received data from multiple data zones, and then distributes the updated model parameters to each intelligent fusion terminal to replace the local model.

[0009] Preferably, in step S1, the power data acquisition terminal communicates with the intelligent fusion terminal via HPLC high-speed power line carrier communication, with a transmission frequency of once every 15 minutes.

[0010] Preferably, in step S2, the specific steps for the intelligent fusion terminal to identify the transformer area topology include: When the topology of the transformer area changes, the intelligent fusion terminal controls the characteristic current generator at the total outlet of the transformer area to inject a coded current signal of a specific frequency into the line. The detection device at each branch node receives and decodes the coded current signal to determine the signal transmission path; The intelligent fusion terminal generates updated transformer topology information based on the signal transmission path and uploads it to the cloud platform for verification.

[0011] Preferably, in step S3, the formula for calculating the line loss rate of a certain level is: the power supply of that level minus the power sales of that level, and then divided by the power supply of that level; wherein, the power sales of the previous level node are used as the power supply of the next level node.

[0012] Preferably, in step S3, the intelligent fusion terminal performs distributed collaborative computing with other intelligent fusion terminals in the same area to jointly establish and train a lightweight line loss analysis model, and the terminals transmit model parameters instead of raw data.

[0013] Preferably, in step S5, the horizontal comparison analysis involves comparing the same-period graded line loss rates of different transformer substations within the same power supply area during the same time period, and marking substations with line loss rates significantly higher than the average as abnormal candidates; the vertical historical trend analysis involves comparing the current graded line loss rate of a single transformer substation with the graded line loss rate of that substation in the same historical period, and marking substations with fluctuations exceeding a preset threshold as abnormal candidates.

[0014] Preferably, in step S6, the step of diagnosing the cause of the abnormality specifically includes: When an abnormal line loss is detected at the branch box level, the cloud platform combines weather data and load data of the area covered by the branch box to determine whether the abnormality is due to technical line loss or management line loss. If the abnormal line loss is determined to be abnormal, further analysis based on the characteristics of electricity consumption data will be conducted to determine whether there is any electricity theft.

[0015] Preferably, in step S7, the closed-loop defect elimination management step specifically includes: The cloud platform generates a defect elimination work order that includes the location of the anomaly, the type of anomaly, and handling suggestions; Push the defect elimination work order to the maintenance personnel's terminal; Receive the troubleshooting results from the maintenance personnel's terminals, and recalculate the line loss rate based on the data after troubleshooting to verify the governance effect; If the verification passes, the work order is closed; if the verification fails, the defect elimination strategy is regenerated and issued.

[0016] Due to the adoption of the above technical solution, the beneficial effects of the present invention are: (1) The present invention performs local topology identification and hierarchical line loss calculation through edge-side intelligent fusion terminal, which reduces data transmission pressure and improves real-time performance and response speed; (2) This invention combines pulse analysis and characteristic current injection method to automatically identify topology, which solves the problems of unclear household transformer relationships and untimely file updates; (3) The cloud platform of this invention uses data from multiple stations to perform horizontal comparison and vertical trend analysis, which can accurately locate abnormal line loss points and combine multi-source data to diagnose the causes of abnormalities, providing accurate decision-making basis for line loss management; (4) This invention achieves a complete closed loop from “discovering problems - analyzing problems - solving problems - verifying effects” through a defect elimination work order closed-loop management mechanism, thereby improving the efficiency of line loss management. Attached Figure Description

[0017] Figure 1 This is a flowchart of the method steps of the present invention.

[0018] Figure 2 This is a logical relationship diagram of the graded line loss calculation and cloud anomaly location of the present invention. Detailed Implementation

[0019] The technical solution of the present invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0020] like Figure 1 and Figure 2 As shown, this invention provides a cloud-edge collaborative method for lean management of line loss in transformer substations, comprising the following steps: S1, Data Acquisition and Edge Deployment: Intelligent converged terminals are deployed at the edge of the distribution area to collect raw electricity consumption data from key nodes within the area. These key nodes include the low-voltage side of the distribution transformer, the incoming line of the branch box, the incoming line of the meter box, and the user's electricity meter. The electricity data acquisition terminal communicates with the intelligent converged terminal via HPLC high-speed power line carrier communication, transmitting data once every 15 minutes.

[0021] Data preprocessing: After the intelligent fusion terminal collects data, it uses a Markov process missing data completion algorithm to fill in missing values, and a mean-shift clustering algorithm to identify and correct abnormal data. The preprocessed data is then used for line loss calculation.

[0022] S2, Automatic Distribution Network Topology Identification: Based on raw electricity consumption data, the intelligent fusion terminal automatically identifies the physical topology of the low-voltage distribution network in the distribution area using a combination of pulse analysis and characteristic current injection methods, generating distribution area topology information. When the distribution area topology changes, the intelligent fusion terminal controls the characteristic current generator at the main outlet of the distribution area to inject a coded current signal of a specific frequency into the line; the detection devices at each branch node receive and decode the signal to determine the transmission path; the intelligent fusion terminal then generates updated topology information and uploads it to the cloud for verification.

[0023] S3, Tiered Line Loss Calculation: Based on the transformer substation topology information, the intelligent fusion terminal calculates the tiered line loss rate for each of the four levels: transformer-low-voltage outgoing line, low-voltage outgoing line-branch box, branch box-meter box, and meter box-household meter. The formula for calculating the line loss rate of a certain level is: power supply of that level minus power sales of that level, then divided by the power supply of that level; the power sales of the higher-level node are used as the power supply of the lower-level node. The intelligent fusion terminal performs distributed collaborative computing with other terminals within the same transformer substation to jointly establish and train a lightweight line loss analysis model, transmitting model parameters between terminals instead of raw data.

[0024] S4, Data Upload: The intelligent fusion terminal uploads the transformer area topology information and the graded line loss rate to the cloud-based line loss analysis platform.

[0025] S5, Cloud-based Verification and Anomaly Location: The cloud-based line loss analysis platform receives and stores data from multiple intelligent fusion terminals in various distribution areas. Based on the data from multiple distribution areas, it performs horizontal comparative analysis and vertical historical trend analysis to verify the graded line loss rate and identify abnormal line loss points. Horizontal comparison involves comparing the same-period graded line loss rates of different distribution areas within the same region, marking those that are significantly higher. Vertical trend analysis involves comparing the current line loss rate of a single distribution area with its historical data for the same period, marking those whose fluctuations exceed the threshold.

[0026] S6, Anomaly Diagnosis and Strategy Generation: The cloud-based line loss analysis platform, based on the location results of abnormal line loss points and combined with weather and load data, diagnoses the causes of anomalies and generates line loss mitigation strategies. During diagnosis, if the anomaly is at the branch box level, it determines whether it is a technical or administrative line loss anomaly based on weather and load data; if it is determined to be an administrative line loss anomaly, it further analyzes the characteristics of electricity consumption data to determine whether there is electricity theft.

[0027] S7, Closed-Loop Defect Elimination Management: The cloud-based line loss analysis platform distributes line loss management strategies to intelligent converged terminals and maintenance personnel terminals to execute closed-loop defect elimination management. Specifically, this includes: generating defect elimination work orders containing the location, type, and handling suggestions of anomalies; pushing these work orders to maintenance personnel terminals; receiving feedback on defect elimination results; recalculating the line loss rate based on the eliminated data to verify the effectiveness; closing the work order if verification is successful, otherwise regenerating and distributing the strategy.

[0028] In practice, a cloud platform can be set up to periodically aggregate and optimize the global line loss analysis model, and the updated model parameters can be sent to each intelligent fusion terminal to replace the local model. Example

[0029] In this embodiment, intelligent converged terminals are deployed as edge computing nodes in the distribution area. Power data acquisition terminals are installed on the low-voltage side of the distribution transformer, branch boxes, meter boxes, and at user power meters, uploading voltage, current, and power data to the intelligent converged terminals every 15 minutes via HPLC communication.

[0030] After receiving data, the intelligent fusion terminal first performs data preprocessing: a Markov process missing data completion algorithm is used to fill in missing values, and a mean-shift clustering algorithm is used to identify and correct abnormal data. Following preprocessing, the intelligent fusion terminal identifies the transformer area topology based on time-series voltage correlation analysis and characteristic current signal injection. Specifically, when a new power connection or topology change is detected, a characteristic current generator is triggered to inject an coded signal into the line. Each node decodes the signal and feeds back path information, achieving automatic topology update and identification.

[0031] The intelligent fusion terminal calculates the graded line loss according to the identified topology information in four levels. Taking the branch box-meter box level as an example, the electricity sales of the branch box node is the sum of the power supply of each subordinate meter box. The line loss rate of this level = (power supply of branch box - sum of power consumption of each meter box) / power supply of branch box × 100%.

[0032] The intelligent fusion terminal packages and uploads the graded line loss calculation results and topology information to the cloud-based line loss analysis platform. The cloud platform aggregates data from multiple transformer substations and performs horizontal comparative analysis: it sorts and compares the line loss rates of different substations within the same region, marking substations with significantly higher line loss rates. Simultaneously, it performs vertical trend analysis: it compares the current line loss rate of each substation with its historical average over the past three months, marking abnormal substations with fluctuations exceeding 20%.

[0033] The cloud platform locates anomalies based on cross-comparison results. When an anomaly is detected at the branch box level, the cloud platform retrieves data such as temperature, humidity, and load curves for the area covered by that branch box. For example, if increased technical line loss is caused by increased line impedance under high temperatures, it is determined to be an anomaly in technical line loss, and a mitigation suggestion of "increasing reactive power compensation or replacing with larger cross-section conductors" is generated. If the investigation finds that a user's electricity consumption suddenly drops and does not match the load curve, it is determined to be an anomaly in management line loss, and further marked as a suspected area for electricity theft.

[0034] The cloud platform generates a defect elimination work order containing the location, type, and handling suggestions of the anomaly, and pushes it to the mobile terminal of the maintenance personnel. After handling the issue on-site, the maintenance personnel report the results on their terminal. Upon receiving the feedback, the cloud platform adjusts the power consumption data after defect elimination and recalculates the line loss rate for that level. If the loss rate returns to the normal range, the work order is closed in a closed loop; if the situation remains unresolved, a new strategy is generated and issued. Example

[0035] In this embodiment, multiple intelligent fusion terminals are configured within the same area. These terminals jointly establish a line loss analysis model through distributed collaborative computing. Each terminal trains a sub-model based on local data, and iteratively optimizes the model by passing model parameters to each other. With topology identification accuracy and line loss calculation accuracy as joint optimization objectives, the model parameters are continuously updated to ultimately form a complete line loss analysis model.

[0036] The cloud platform regularly receives model parameters and operational data uploaded by each distribution area, aggregates and optimizes the global line loss analysis model, and distributes the updated global model parameters to the intelligent fusion terminals of each distribution area to replace the old local model, thereby achieving continuous model evolution and performance improvement.

[0037] 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 scope of the claims of the present invention.

Claims

1. A cloud-edge collaborative method for lean management of transformer area line loss, characterized in that, Includes the following steps: S1, deploy intelligent fusion terminals on the edge of the distribution area, collect raw power consumption data of each key node in the distribution area through the power data acquisition terminal of the distribution area. Key nodes include the low-voltage side of the distribution transformer, the incoming line of the branch box, the incoming line of the meter box and the user's power meter; S2, the intelligent fusion terminal, based on the original power consumption data, uses a combination of pulse analysis and characteristic current injection methods to automatically identify the physical topology of the low-voltage distribution network in the distribution area and generate the distribution area topology information. S3, the intelligent fusion terminal calculates the graded line loss rate of each level according to the transformer area topology information, according to the four levels of transformer-low voltage outgoing line, low voltage outgoing line-branch box, branch box-meter box, and meter box-household meter; S4, the intelligent fusion terminal uploads the transformer area topology information and graded line loss rate to the cloud-based line loss analysis platform; S5, the cloud-based line loss analysis platform receives and stores data from multiple intelligent converged terminals in multiple distribution areas, performs horizontal comparative analysis and vertical historical trend analysis based on data from multiple distribution areas, verifies the graded line loss rate, and identifies abnormal line loss points; S6, the cloud-based line loss analysis platform, diagnoses the causes of abnormalities and generates line loss mitigation strategies based on the location results of abnormal line loss points, combined with weather and load data; The S7 cloud-based line loss analysis platform distributes line loss management strategies to all integrated terminals and maintenance personnel terminals to implement closed-loop defect elimination management.

2. The cloud-edge collaborative lean management method for transformer area line loss according to claim 1, characterized in that: It also includes the data preprocessing step. After collecting the raw electricity consumption data, the intelligent fusion terminal first uses the Markov process missing data completion algorithm to fill in the missing values, and then uses the mean drift clustering algorithm to identify and correct abnormal data. Finally, the preprocessed data is used for line loss calculation.

3. The cloud-edge collaborative lean management method for transformer area line loss according to claim 1, characterized in that: It also includes a model update step. The cloud-based line loss analysis platform regularly aggregates and optimizes the global line loss analysis model based on the received data from multiple stations, and then sends the updated model parameters to each intelligent fusion terminal to replace the local model.

4. The cloud-edge collaborative lean management method for transformer area line loss according to claim 1, characterized in that: In step S1, the power data acquisition terminal communicates with the intelligent fusion terminal via HPLC high-speed power line carrier communication, with a transmission frequency of once every 15 minutes.

5. The cloud-edge collaborative lean management method for transformer area line loss according to claim 1, characterized in that: In step S2, the specific steps for the intelligent fusion terminal to identify the transformer area topology include: When the topology of the transformer area changes, the intelligent fusion terminal controls the characteristic current generator at the total outlet of the transformer area to inject a coded current signal of a specific frequency into the line. The detection device at each branch node receives and decodes the coded current signal to determine the signal transmission path; The intelligent fusion terminal generates updated transformer topology information based on the signal transmission path and uploads it to the cloud platform for verification.

6. The cloud-edge collaborative lean management method for transformer area line loss according to claim 1, characterized in that: In step S3, the formula for calculating the line loss rate of a certain level is: the power supply of that level minus the power sales of that level, and then divided by the power supply of that level; wherein, the power sales of the previous level node are used as the power supply of the next level node.

7. The cloud-edge collaborative lean management method for transformer area line loss according to claim 1, characterized in that: In step S3, the intelligent fusion terminal performs distributed collaborative computing with other intelligent fusion terminals in the same area to jointly establish and train a lightweight line loss analysis model, and the terminals transmit model parameters instead of raw data.

8. The cloud-edge collaborative lean management method for transformer area line loss according to claim 1, characterized in that: In step S5, the horizontal comparison analysis involves comparing the same-period graded line loss rates of different transformer areas within the same power supply area during the same time period, and marking transformer areas with line loss rates significantly higher than the average as abnormal candidates. The longitudinal historical trend analysis compares the current graded line loss rate of a single transformer area with the graded line loss rate of the same period in history, and marks transformer areas whose fluctuations exceed a preset threshold as abnormal candidates.

9. The cloud-edge collaborative lean management method for transformer area line loss according to claim 1, characterized in that: In step S6, the step of diagnosing the cause of the abnormality specifically includes: When an abnormal line loss is detected at the branch box level, the cloud platform combines weather data and load data of the area covered by the branch box to determine whether the abnormality is due to technical line loss or management line loss. If the abnormal line loss is determined to be abnormal, further analysis based on the characteristics of electricity consumption data will be conducted to determine whether there is any electricity theft.

10. The cloud-edge collaborative lean management method for transformer area line loss according to claim 1, characterized in that: In step S7, the closed-loop defect elimination management step specifically includes: The cloud platform generates a defect elimination work order that includes the location of the anomaly, the type of anomaly, and handling suggestions; Push the defect elimination work order to the maintenance personnel's terminal; Receive the troubleshooting results from the maintenance personnel's terminals, and recalculate the line loss rate based on the data after troubleshooting to verify the governance effect; If the verification passes, the work order is closed; if the verification fails, the defect elimination strategy is regenerated and issued.

Images