Topology identification system for power distribution network maintenance
By constructing a high-precision distribution network topology identification system, the problems of low efficiency and large errors in existing technologies have been solved, and the accuracy and security of distribution network topology identification have been improved, ensuring the accuracy and safety of the maintenance process.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANGQIU POWER SUPPLY CO OF STATE GRID HANAN ELECTRIC POWER CO
- Filing Date
- 2025-12-04
- Publication Date
- 2026-05-12
AI Technical Summary
Existing methods for identifying the topology of power distribution networks rely on manual inspections and traditional algorithms, which are inefficient, have large errors, and poor dynamic adaptability, leading to misjudgments and safety hazards during maintenance.
The system employs a data acquisition module, a data preprocessing module, a feature extraction module, a topology reasoning module, an optimization module, and a visualization output module. Combined with wireless sensors, intelligent monitoring units, and time stamping units, it constructs a high-precision distribution network topology identification model and generates a work boundary topology map through data cleaning, filtering, normalization, fusion, and transfer learning techniques.
This has improved the accuracy and efficiency of distribution network topology identification, eliminated misjudgments caused by inconsistencies between diagrams and models, and enhanced the safety and convenience of the maintenance process.
Smart Images
Figure CN122026596A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of power distribution network maintenance technology, and specifically relates to a topology identification system for power distribution network maintenance. Background Technology
[0002] As a crucial component of the power system, the distribution network is complex in structure and contains numerous nodes. Its topology serves as the foundation for grid dispatching, fault handling, maintenance work, and load transfer. Therefore, accurate identification of the distribution network topology is a prerequisite for ensuring the safe and efficient conduct of maintenance work. Currently, existing methods for distribution network topology identification primarily rely on manual inspection records or SCADA system data analysis. These methods have the following drawbacks: 1. Manual inspections are inefficient and prone to errors, making it difficult to capture real-time changes in the topology; 2. Traditional topology identification algorithms are based on static data to build models, resulting in poor dynamic adaptability and an inability to respond promptly to changes in the state of distribution network nodes; 3. The lack of precise location of associated nodes during maintenance easily leads to inconsistencies between the diagram and the model, causing maintenance personnel to misjudge the outage area, resulting in accidental power restoration or missed power outages, posing significant safety hazards. Therefore, to address these issues, it is essential to develop an accurate, efficient, safe, and convenient distribution network maintenance topology identification system. Summary of the Invention
[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide a precise, efficient, safe and convenient topology identification system for power distribution network maintenance.
[0004] The objective of this invention is achieved as follows: A topology identification system for power distribution network maintenance includes a data acquisition module, a data preprocessing module, a feature extraction module, a topology reasoning module, an optimization module, a topology generation module, and a visualization output module. The data acquisition module consists of multiple wireless sensors, an intelligent monitoring unit, and a time stamping unit. The wireless sensors are deployed at power distribution network lines and transformer nodes to collect voltage, current, power factor, and phase data of power distribution network equipment. The intelligent monitoring unit is used to acquire the opening and closing status data of circuit breakers and disconnect switches in real time. The time stamping unit is used to time stamp the data collected by each data acquisition module, thereby ensuring the consistency of data timestamps. The data preprocessing module is connected to the data acquisition module via a wireless communication component. It is used to clean, filter, normalize, and fuse the data acquired by the data acquisition module to form a high-quality dataset. The feature extraction module is used to extract the characteristic parameters of the operation of the distribution network nodes; The topology reasoning module can construct an initial topology model of the distribution network based on the operating feature parameters of the distribution network nodes extracted by the feature extraction module. Subsequently, it receives sensing data from the field and uses it as the highest priority topology verification benchmark to train the initial topology model of the distribution network, thereby obtaining the distribution network topology identification model. The optimization module can conduct an in-depth analysis of the mechanism of the distribution network topology identification model based on transfer learning technology, propose a highly available parameter identification algorithm that adapts to complex sampling conditions, and then construct an algorithm confidence evaluation system through validation set performance monitoring, model correction and hidden pattern extraction technology during supervised training, thereby enhancing the accuracy and reliability of the distribution network topology identification model. The topology generation module can input high-quality datasets into the distribution network topology recognition model to generate the current real topology of the distribution network; and then, based on the maintenance work plan ticket and the current real topology, automatically generate a work boundary topology map related to the maintenance area. The visualization output module can be used to display the topology map of the work boundary, and can also be used to plan and mark the access paths to the maintenance area.
[0005] Furthermore, in the data preprocessing process: data cleaning is to remove outliers and null values; data filtering is to remove the influence of electromagnetic interference and impulse noise on the data; data normalization is to eliminate the incommensurability between data features due to differences in dimensions and sizes; and data fusion is to use a feature fusion algorithm based on an attention mechanism to achieve multi-source data fusion.
[0006] Furthermore, the work boundary topology map can clearly mark the boundaries of maintenance equipment, isolation points, grounding wire connection points, and energized and de-energized areas.
[0007] Furthermore, the visualization output module is equipped with an editing unit, which can be used to zoom and pan the work boundary topology map.
[0008] Furthermore, the data preprocessing module is equipped with a storage unit, which can be used to store and back up the high-quality dataset formed after preprocessing.
[0009] Furthermore, the visualization output module is equipped with a calibration unit, which can be used to calibrate and verify the location information of the power distribution network equipment within the work boundary topology map.
[0010] The beneficial effects of this invention are as follows: By combining a data acquisition module and a data preprocessing module, this invention can collect and preprocess the operating data of the power distribution network, thereby forming a high-quality dataset, which effectively solves the problems of low efficiency and large error in existing manual inspection records. By setting up the topology reasoning module, a distribution network topology identification model can be obtained through training. By setting up the optimization module, the mechanism of the distribution network topology identification model can be analyzed in depth, and the accuracy and reliability of the distribution network topology identification model can be enhanced by using the algorithm confidence evaluation system. By setting the topology generation module, a work boundary topology map related to the maintenance area can be generated based on a high-quality dataset and a distribution network topology recognition model. This allows for the unification of the "map-model-site" information, fundamentally eliminating misjudgments caused by inconsistencies between the map and the model, and increasing the safety of the maintenance process. Overall, this invention has the advantages of being accurate, efficient, safe, and convenient. Attached Figure Description
[0011] Figure 1 This is a flowchart illustrating the structure of the present invention. Detailed Implementation
[0012] The present invention will now be further described with reference to the accompanying drawings.
[0013] Example: Figure 1 As shown, a topology identification system for power distribution network maintenance includes a data acquisition module, a data preprocessing module, a feature extraction module, a topology reasoning module, an optimization module, a topology generation module, and a visualization output module. The data acquisition module consists of multiple wireless sensors, an intelligent monitoring unit, and a time stamping unit. The wireless sensors are deployed at power distribution network lines and transformer nodes to collect voltage, current, power factor, and phase data of power distribution network equipment. The intelligent monitoring unit is used to acquire the opening and closing status data of circuit breakers and disconnectors in real time. The time stamping unit is used to time stamp the data collected by each data acquisition module to ensure the consistency of data timestamps. The data preprocessing module is connected to the data acquisition module via a wireless communication component. It performs cleaning, filtering, normalization, and fusion processing on the data acquired by the data acquisition module to form a high-quality dataset. Specifically, data cleaning removes outliers and null values; data filtering removes the influence of electromagnetic interference and impulse noise; data normalization eliminates incommensurability caused by differences in dimensions and sizes between data features; and data fusion uses an attention-based feature fusion algorithm to fuse multi-source data. After preprocessing, the resulting high-quality dataset is stored and backed up using a storage unit within the data preprocessing module. The feature extraction module is used to extract the characteristic parameters of the operation of the distribution network nodes; The topology reasoning module can construct an initial topology model of the distribution network based on the operating feature parameters of the distribution network nodes extracted by the feature extraction module. Subsequently, it receives sensing data from the field and uses it as the highest priority topology verification benchmark to train the initial topology model of the distribution network, thereby obtaining the distribution network topology identification model. The optimization module can conduct an in-depth analysis of the mechanism of the distribution network topology identification model based on transfer learning technology, propose a highly available parameter identification algorithm that adapts to complex sampling conditions, and then construct an algorithm confidence evaluation system through validation set performance monitoring, model correction and hidden pattern extraction technology during supervised training, thereby enhancing the accuracy and reliability of the distribution network topology identification model. The topology generation module can input high-quality datasets into the distribution network topology recognition model to generate the current real topology of the distribution network; and then, based on the maintenance work plan ticket and the current real topology, automatically generate a work boundary topology map related to the maintenance area; wherein, the work boundary topology map can clearly mark the boundaries of maintenance equipment, isolation points, grounding wire connection points, and energized and de-energized areas; The visualization output module can be used to display the work boundary topology map, and can also be used to plan and mark the access paths to the maintenance area. Furthermore, the visualization output module is equipped with an editing unit and a calibration unit. The editing unit can be used to zoom and pan the work boundary topology map, and the calibration unit can be used to calibrate and verify the location information of the power distribution network equipment inside the work boundary topology map.
[0014] In use, this invention first collects voltage, current, power factor, and phase data of the distribution network equipment using wireless sensors deployed at distribution network lines and transformer nodes. An intelligent monitoring unit acquires real-time data on the opening and closing status of circuit breakers and disconnectors. Then, a time stamping unit timestamps the data collected by each acquisition module to ensure data timestamp consistency. Next, a data preprocessing module cleans, filters, normalizes, and fuses the collected data to form a high-quality dataset. Data cleaning removes outliers and null values; data filtering removes the influence of electromagnetic interference and impulse noise; data normalization eliminates incommensurability caused by differences in dimensions and sizes between data features; and data fusion uses an attention-based feature fusion algorithm to fuse multi-source data. Then, a feature extraction module extracts the characteristic parameters of the distribution network node operation. Following this, based on the extracted characteristic parameters, a topology reasoning module constructs an initial distribution network topology model. The system receives sensing data from the field and uses it as the highest priority topology verification benchmark to train the initial topology model of the distribution network, thereby obtaining the distribution network topology identification model. Finally, based on transfer learning technology, an optimization module is used to deeply analyze the mechanism of the distribution network topology identification model, proposing a highly available parameter identification algorithm adapted to complex sampling conditions. Subsequently, through supervised training, validation set performance monitoring, model correction, and hidden pattern extraction techniques, an algorithm confidence evaluation system is constructed to enhance the accuracy and reliability of the distribution network topology identification model. After completing the above operations, a high-quality dataset is input into the distribution network topology identification model to generate the current real topology of the distribution network. Then, based on the maintenance work plan and the current real topology, a work boundary topology map related to the maintenance area is automatically generated. This work boundary topology map clearly marks the boundaries of maintenance equipment, isolation points, grounding wire connection points, and energized and de-energized areas. Finally, a visualization output module is used to display the work boundary topology map and to plan and mark the entry and exit paths to the maintenance area within the work boundary topology map.
[0015] This invention, through the coordinated setup of a data acquisition module and a data preprocessing module, can collect and preprocess operational data of the distribution network, thereby forming a high-quality dataset. This effectively solves the problems of low efficiency and large errors in existing manual inspection records. The topology reasoning module allows for the training of a distribution network topology recognition model. The optimization module enables in-depth analysis of the mechanism of the distribution network topology recognition model, and the use of an algorithm confidence evaluation system enhances the accuracy and reliability of the model. The topology generation module generates a work boundary topology map related to the maintenance area based on the high-quality dataset and the distribution network topology recognition model. This work boundary topology map achieves consistency between the map, model, and site, fundamentally eliminating misjudgments caused by inconsistencies between the map and model, and increasing the safety of the maintenance process. In summary, this invention has the advantages of accuracy, efficiency, safety, and convenience.
[0016] 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 topology identification system for distribution network maintenance, comprising a data acquisition module, a data preprocessing module, a feature extraction module, a topology reasoning module, an optimization module, a topology generation module, and a visualization output module, characterized in that: The data acquisition module consists of multiple wireless sensors, an intelligent monitoring unit, and a time stamping unit. The wireless sensors are deployed at the distribution network lines and transformer nodes to collect voltage, current, power factor, and phase data of the distribution network equipment. The intelligent monitoring unit is used to acquire the opening and closing status data of circuit breakers and disconnect switches in real time. The time stamping unit is used to time stamp the data collected by each data acquisition module to ensure the consistency of data timestamps. The data preprocessing module is connected to the data acquisition module via a wireless communication component. It is used to clean, filter, normalize, and fuse the data acquired by the data acquisition module to form a high-quality dataset. The feature extraction module is used to extract the characteristic parameters of the operation of the distribution network nodes; The topology reasoning module can construct an initial topology model of the distribution network based on the operating feature parameters of the distribution network nodes extracted by the feature extraction module. Subsequently, it receives sensing data from the field and uses it as the highest priority topology verification benchmark to train the initial topology model of the distribution network, thereby obtaining the distribution network topology identification model. The optimization module can conduct an in-depth analysis of the mechanism of the distribution network topology identification model based on transfer learning technology, propose a highly available parameter identification algorithm that adapts to complex sampling conditions, and then construct an algorithm confidence evaluation system through validation set performance monitoring, model correction and hidden pattern extraction technology during supervised training, thereby enhancing the accuracy and reliability of the distribution network topology identification model. The topology generation module can input high-quality datasets into the distribution network topology identification model to generate the current real topology of the distribution network. Then, based on the maintenance work plan ticket and the current actual topology, an operation boundary topology map related to the maintenance area is automatically generated. The visualization output module can be used to display the topology map of the work boundary, and can also be used to plan and mark the access paths to the maintenance area.
2. The topology identification system for distribution network maintenance as described in claim 1, characterized in that: In the data preprocessing process: data cleaning is to remove outliers and null values; data filtering is to remove the influence of electromagnetic interference and impulse noise on the data; data normalization is to eliminate the incommensurability between data features due to differences in units and sizes; and data fusion is to use a feature fusion algorithm based on an attention mechanism to achieve multi-source data fusion.
3. The topology identification system for distribution network maintenance as described in claim 1, characterized in that: The operation boundary topology map can clearly mark the boundaries of maintenance equipment, isolation points, grounding wire connection points, and energized and de-energized areas.
4. The topology identification system for distribution network maintenance as described in claim 1, characterized in that: The visualization output module has an internal editing unit, which can be used to zoom and pan the work boundary topology map.
5. A topology identification system for distribution network maintenance as described in claim 1, characterized in that: The data preprocessing module is equipped with a storage unit, which can be used to store and back up the high-quality dataset formed after preprocessing.
6. The topology identification system for distribution network maintenance as described in claim 1, characterized in that: The visualization output module is equipped with a calibration unit, which can be used to calibrate and verify the location information of power distribution network equipment within the work boundary topology map.