Power distribution network line abnormal state identification method and system

Abstract

The invention provides a power distribution network line abnormal state identification method and system, and relates to the technical field of power distribution network abnormal state identification. According to the method, the distributed network based on topology analysis and edge calculation is constructed, so that localization and low-delay response of data processing are realized, and the real-time performance of anomaly recognition is remarkably improved; meanwhile, a key node set is dynamically corrected by introducing an environment sensing mechanism, and a personalized threshold value is set, so that the system can adapt to the operation environment and load change of the power distribution network, and the false alarm rate is effectively reduced; besides, multi-source data such as voltage, current, temperature and environment are fused, and a space-time aggregation analysis technology is combined to carry out association research and judgment on discrete abnormal information, so that the comprehensiveness and anti-interference capability of fault diagnosis are enhanced, a fault propagation path and progressive hidden dangers can be accurately identified, and the fault diagnosis accuracy is improved. Finally, a power distribution network active defense system which is optimized in resource configuration, quick in response and high in confidence coefficient is formed, and the reliability and the operation and maintenance efficiency of power grid operation are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of power distribution network anomaly identification technology, and in particular to a method and system for identifying abnormal states of power distribution lines. Background Technology

[0002] As a crucial component of the power system, the power distribution network directly serves a vast number of users, and its safe and stable operation is vital for ensuring the normal functioning of the social economy and the quality of life for the people. Traditional methods for identifying abnormal conditions in power distribution network lines have revealed numerous limitations when dealing with these complex situations, making it difficult to meet the demands of modern power distribution networks for efficient, accurate, and real-time monitoring and identification.

[0003] First, traditional monitoring methods often employ a terminal-based data acquisition and cloud-based centralized processing architecture. Uploading massive amounts of high-frequency sensor data to the main station can lead to communication bandwidth congestion and transmission delays, making it difficult to meet the millisecond-level rapid response requirements for power distribution network faults. Second, the selection of key monitoring nodes is often based on static topology or fixed experience, lacking the ability to dynamically perceive the real-time operating environment. This results in the inability to adjust monitoring priorities in a timely manner when the environment changes abruptly, and fixed, uniform thresholds are difficult to adapt to the historical fluctuation characteristics of different nodes, easily leading to missed or false alarms. Finally, existing methods mostly rely on single electrical quantity indicators for independent judgment, lacking the ability to deeply integrate multi-source heterogeneous data and perform cross-time and cross-space correlation analysis. This makes it difficult to effectively distinguish between transient interference and real faults, and also fails to accurately identify highly concealed, progressive hidden dangers, resulting in insufficient accuracy and reliability in anomaly identification.

[0004] Therefore, it is necessary to provide a method and system for identifying abnormal states of power distribution lines to solve the above-mentioned technical problems. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method and system for identifying abnormal states of distribution network lines. This addresses the issues in existing technologies, such as the difficulty in meeting the need for millisecond-level rapid response to distribution network faults and the lack of dynamic perception capabilities of key monitoring nodes in the selection of real-time operating environments.

[0006] The present invention provides a method for identifying abnormal states of power distribution lines, comprising the following steps: S1. Based on the analysis of the power distribution network topology, identify and determine the initial set of key nodes, allocate edge computing nodes to the initial key nodes, and build a distributed edge computing network. S2. Based on a multi-sensor intelligent sensing system and historical data, an environmental perception mechanism is established to monitor the operating environment data of the power distribution network in real time, and the initial set of key nodes is dynamically corrected to form an updated set of key nodes while setting their corresponding operating anomaly judgment thresholds. S3. During the current monitoring cycle, the line operation data is collected through the intelligent sensing system to form the line information for the current cycle, which includes voltage, current, temperature and environmental parameters. S4. Input the current cycle line information to the corresponding edge computing node, use the generated anomaly judgment threshold to compare and analyze the running data of each updated key node, and output the node anomaly status identifier of each updated key node in the current cycle. S5. Collect the node anomaly status identifiers of each updated key node under different characteristic frequencies within the current period, and combine them with the preset fluctuation deviation threshold to perform spatiotemporal aggregation analysis on the discrete node anomaly information to generate identification and analysis results.

[0007] Preferably, step S1 is as follows: S101. Read the topology data of the distribution network and organize the connection relationships of line nodes and load distribution characteristics; S102. Based on the electrical distance centerness and load importance, select the matching line nodes from the line node connection relationships and organize them into an initial set of key nodes. S103. Assign a corresponding edge computing node to each line node in the initial set of key nodes, establish a one-to-one or one-to-many mapping relationship, and build a distributed edge computing network.

[0008] Preferably, step S2 is as follows: S201. Establish an environmental perception mechanism, the steps of which include: collecting historical sensing data and operating parameters collected by the intelligent sensing system, and arranging the historical sensing data and operating parameter data in chronological order to form time series data, wherein the intelligent sensing system includes multiple types of sensors; based on the time series data, analyzing the changing trend of the data using the moving average method to obtain time series analysis results; according to the time series analysis results, setting environmental data deviation thresholds corresponding to multiple time periods within the same period, and when the operating environmental data in the current time period exceeds its environmental data deviation threshold, it is included in the set of initial key nodes.

[0009] S202. Monitor the distribution network operation environment data in real time, and adjust the initial set of key nodes based on the established environment perception mechanism, and correct it to the updated set of key nodes. S203. Collect the historical normal fluctuation range of each key node in the updated key node set, and set a corresponding operation anomaly judgment threshold for each key node in the updated key node set, wherein the operation anomaly judgment threshold is within the historical normal fluctuation range.

[0010] Preferably, step S3 is as follows: S301. Set a fixed time interval as the monitoring cycle, and within the current monitoring cycle, collect the operating data of the distribution network lines through the intelligent sensing system, including voltage, current, temperature and environmental parameters. S302. The collected voltage, current, temperature and environmental parameters are processed to form the current cycle line information.

[0011] Preferably, step S4 is as follows: S401. Transmit the current cycle line information to the corresponding edge computing node. In the edge computing node, call the generated anomaly judgment threshold of the corresponding updated key node, and compare the voltage, current and temperature in the current cycle line information with the anomaly judgment threshold one by one to obtain the comparison analysis results. S402. Based on the comparison and analysis results, output the node abnormal status identifier for each updated key node in the current cycle. If any one of the voltage, current, or temperature indicators exceeds the abnormal judgment threshold, it is marked as an abnormal state.

[0012] Preferably, step S5 is as follows: S501, Aggregate the node abnormal status identifiers of all nodes in the current period under different characteristic frequencies, i.e., different sampling frequencies in the current period. S502. Based on the preset fluctuation deviation threshold and the abnormal status identifiers of all nodes in the current cycle, analyze the correlation of abnormal status of adjacent line nodes in the same time period, as well as the abnormal status change trend of the same line node in different time periods, and finally compile and generate the identification and analysis results.

[0013] A power distribution network line abnormality identification system, the identification system comprising: The node allocation module is used to identify and determine the initial set of key nodes based on the topology analysis of the power distribution network, allocate edge computing nodes to the initial key nodes, and build a distributed edge computing network. The dynamic update module is used to establish an environmental perception mechanism based on the intelligent sensing system with multiple sensors and historical data, monitor the operating environment data of the power distribution network in real time, and dynamically correct the initial set of key nodes to form updated key nodes while setting their corresponding operating anomaly judgment thresholds. The data acquisition module is used to collect line operation data through the intelligent sensing system during the current monitoring period to form the line information for the current period, which includes voltage, current, temperature and environmental parameters. The comparison and analysis module is used to input the current cycle line information to the corresponding edge computing node, and use the generated anomaly judgment threshold to compare and analyze the running data of each updated key node, and output the node anomaly status identifier of each updated key node in the current cycle. The results generation module is used to collect the node anomaly status identifiers of each updated key node under different characteristic frequencies within the current period. Combined with the preset fluctuation deviation threshold, it performs spatiotemporal aggregation analysis on discrete node anomaly information to generate identification and analysis results.

[0014] Compared with related technologies, the method and system for identifying abnormal states of power distribution lines provided by this invention have the following advantages: This invention achieves localized data processing and low-latency response by constructing a distributed network based on topology analysis and edge computing, significantly improving the real-time performance of anomaly identification. Simultaneously, by introducing an environmental awareness mechanism to dynamically correct the set of key nodes and set personalized thresholds, the system can adapt to the distribution network operating environment and load changes, effectively reducing the false alarm rate. Furthermore, by integrating multi-source data such as voltage, current, temperature, and environmental data, and combining spatiotemporal aggregation analysis technology to correlate and analyze discrete anomaly information, it not only enhances the comprehensiveness and anti-interference capability of fault diagnosis but also accurately identifies fault propagation paths and progressive hidden dangers. Ultimately, this forms a distribution network proactive defense system with optimized resource allocation, rapid response, and high confidence, significantly improving the reliability and operational efficiency of the power grid. Attached Figure Description

[0015] Figure 1 This is a flowchart of a method for identifying abnormal states of power distribution lines according to the present invention; Figure 2 This is a system block diagram of a power distribution line abnormality identification system according to the present invention. Detailed Implementation

[0016] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0017] Example 1

[0018] like Figure 1 As shown, a method for identifying abnormal states of power distribution lines includes the following steps: S1. Based on the analysis of the power distribution network topology, identify and determine the initial set of key nodes, allocate edge computing nodes to the initial key nodes, and build a distributed edge computing network. S2. Based on a multi-sensor intelligent sensing system and historical data, an environmental perception mechanism is established to monitor the operating environment data of the power distribution network in real time, and the initial set of key nodes is dynamically corrected to form an updated set of key nodes while setting their corresponding operating anomaly judgment thresholds. S3. During the current monitoring cycle, the line operation data is collected through the intelligent sensing system to form the line information for the current cycle, which includes voltage, current, temperature and environmental parameters. S4. Input the current cycle line information to the corresponding edge computing node, use the generated anomaly judgment threshold to compare and analyze the running data of each updated key node, and output the node anomaly status identifier of each updated key node in the current cycle. S5. Collect the node anomaly status identifiers of each updated key node under different characteristic frequencies within the current period, and combine them with the preset fluctuation deviation threshold to perform spatiotemporal aggregation analysis on the discrete node anomaly information to generate identification and analysis results.

[0019] In the specific implementation process, step S1 is as follows: S101. Read the topology data of the distribution network and organize the connection relationships of line nodes and load distribution characteristics.

[0020] Specifically, the topology data of the distribution network is obtained from the distribution network management system or relevant databases. This data includes detailed information about each line node in the distribution network, such as the connection methods between nodes and the route of the lines. This allows for a clear understanding of the connection relationships between line nodes, identifying which nodes are directly connected and which are indirectly connected through other nodes. Simultaneously, the load distribution characteristics of each node must be extracted from the data, including the size of the electrical load borne by each node, the load type (e.g., industrial load, commercial load, residential load), and the load variation patterns over time.

[0021] In this embodiment, there is a simple power distribution network containing four nodes: A, B, C, and D. Topology data retrieved from the database shows that node A is directly connected to node B, node B is directly connected to node C, and node C is directly connected to node D. Furthermore, data analysis reveals that node A primarily handles residential loads, with a relatively stable load and an average load of 100kW; node B, in addition to residential loads, also handles some commercial loads, with significant load fluctuations, reaching a peak of 200kW; node C is a concentrated area of ​​industrial loads, with a large and highly variable load, averaging 300kW and peaking at 500kW; and node D mainly handles residential and small commercial loads, with a relatively small and stable load, averaging 80kW.

[0022] S102. Based on the electrical distance centerness and load importance, select the matching line nodes from the line node connection relationships and organize them into an initial set of key nodes.

[0023] Specifically, electrical distance centrality is an indicator that measures the importance of a node's electrical location in a distribution network, while load importance is determined based on the nature and size of the load borne by the node and its impact on the overall operation of the distribution network. When selecting nodes, thresholds for electrical distance centrality and load importance need to be preset. Then, based on the line node connection relationships and load distribution characteristics, the electrical distance centrality and load importance of each node are calculated. The calculated electrical distance centrality and load importance are compared with the preset thresholds. Only when both the electrical distance centrality and load importance of a node meet the corresponding threshold requirements will the node be selected. The calculation of load importance can be based on load size and load type, using a weighted summation method. The formula is: Load Importance = Load Size Weight × Load Size Coefficient + Load Type Weight × Load Type Coefficient. It should be noted that different coefficients are set according to the load size. To simplify the calculation, a piecewise linear function method can be used. For example, when the load active power is less than 1000kW, the load size coefficient is 0.2; when the load active power is between 1000kW and 3000kW, the load size coefficient is 0.5; when the load active power is greater than 3000kW, the load size coefficient is 0.8. Load type coefficient: The coefficient is set according to the load type. (Industrial load) Class I load coefficient is 0.9, (Commercial load) Class II load coefficient is 0.6, and (Residential load) Class III load coefficient is 0.3.

[0024] In this embodiment, the preset electrical distance centrality threshold is 0.5, and the load importance threshold is 0.6. Calculations show that node A has an electrical distance centrality of 0.3 and a load importance of 0.4; node B has an electrical distance centrality of 0.6 and a load importance of 0.7; node C has an electrical distance centrality of 0.8 and a load importance of 0.9; and node D has an electrical distance centrality of 0.2 and a load importance of 0.3. Based on the threshold comparison, nodes B and C both meet the requirements for electrical distance centrality and load importance; therefore, nodes B and C are grouped into the initial set of critical nodes.

[0025] S103. Assign a corresponding edge computing node to each line node in the initial set of key nodes, establish a one-to-one or one-to-many mapping relationship, and build a distributed edge computing network.

[0026] Specifically, it is necessary to determine the available edge computing node resources and allocate a corresponding edge computing node to each line node in the initial set of key nodes. The allocation method can be one-to-one, that is, one initial key node corresponds to one edge computing node; or it can be one-to-many, that is, one edge computing node is responsible for the data processing tasks of multiple initial key nodes. After the allocation is completed, the mapping relationship between the initial key nodes and edge computing nodes is established, and they are connected through network configuration and other means to form a distributed edge computing network.

[0027] In the specific implementation process, step S2 is as follows: S201. Establish an environmental perception mechanism, the steps of which include: collecting historical sensing data and operating parameters collected by the intelligent sensing system, and arranging the historical sensing data and operating parameter data in chronological order to form time series data, wherein the intelligent sensing system includes multiple types of sensors; based on the time series data, analyzing the changing trend of the data using the moving average method in existing technology to obtain time series analysis results; according to the time series analysis results, setting environmental data deviation thresholds corresponding to multiple time periods within the same period, and when the operating environmental data in the current time period exceeds its environmental data deviation threshold, it is included in the set of initial key nodes.

[0028] Specifically, historical sensing data and operating parameters collected from various types of sensors in the intelligent sensing system are acquired. These sensors include temperature sensors, humidity sensors, wind speed sensors, voltage or current sensors, etc., used to monitor various factors in the power distribution network's operating environment. The collected data is arranged chronologically to form time-series data. For example, hourly temperature and humidity data are arranged sequentially to construct a data sequence that changes over time. The moving average method is used to analyze the time-series data. The moving average method smooths data fluctuations by calculating the average value of data within a certain time window, thus providing a clearer view of data trends. For example, a 5-hour moving average window is used to calculate the average value of data every 5 hours, resulting in a new sequence that reflects the long-term trend of the data. Based on the time-series analysis results, corresponding environmental data deviation thresholds are set for multiple time periods within the same period. For example, the environmental data changes differently during the day and night; based on the analyzed trends of daytime and nighttime data, environmental data deviation thresholds are set for different time periods during the day and night. When the monitored operating environment data exceeds its corresponding environmental data deviation threshold during the current time period, it indicates that the operating status of the node in the current environment may have changed significantly and requires close attention. At this time, the node is included in the initial set of critical nodes.

[0029] S202. Monitor the distribution network operation environment data in real time, and adjust the initial set of key nodes based on the established environment perception mechanism, and correct it to the updated set of key nodes.

[0030] Specifically, an intelligent sensing system continuously monitors the power distribution network's operating environment data in real time, including various environmental parameters such as temperature, humidity, and wind speed. Based on an established environmental perception mechanism, the monitored environmental data is analyzed and judged. If the operating environment data of a node exceeds its corresponding environmental data deviation threshold, that node is added to the initial set of critical nodes; conversely, if the operating environment data of an initial critical node remains within the normal range for a period of time, and its importance is determined to decrease according to the environmental perception mechanism, it can be removed from the initial set of critical nodes. Through dynamic adjustment, the initial set of critical nodes is revised to an updated set.

[0031] In this embodiment, a temperature sensor records the ambient temperature hourly within a power distribution network area. Temperature data from the past month is collected and compiled into a time series. Analysis using the moving average method reveals that daytime (8 AM - 6 PM) temperatures fluctuate between 25-30°C, while nighttime (6 PM - 8 AM the following day) temperatures fluctuate between 15-20°C. Based on this analysis, a daytime temperature deviation threshold of ±5°C and a nighttime temperature deviation threshold of ±3°C are set. If, at 10 AM on a certain day, the temperature suddenly rises to 36°C, exceeding the daytime temperature deviation threshold, then the power distribution network node where the temperature sensor is located is included in the initial set of critical nodes.

[0032] S203. Collect the historical normal fluctuation range of each key node in the updated key node set, and set a corresponding operation anomaly judgment threshold for each key node in the updated key node set, wherein the operation anomaly judgment threshold is within the historical normal fluctuation range.

[0033] Specifically, historical operational data for each critical node in the updated critical node set is collected, and the fluctuation range of this data under normal conditions is analyzed. For example, for a critical node, data on parameters such as voltage, current, and temperature over a period of time are collected, and statistical analysis methods such as calculating standard deviation and range are used to determine the normal fluctuation range of these parameters. Based on the collected historical normal fluctuation range, a corresponding operational anomaly judgment threshold is set for each critical node. The operational anomaly judgment threshold should be outside the historical normal fluctuation range, but it should not be too lenient or too strict. For example, if the historical normal fluctuation range of the voltage of a critical node is 220V±5V, then the operational anomaly judgment threshold can be set to 220V±8V. When the voltage exceeds this range, the voltage of that node is considered abnormal.

[0034] In the specific implementation process, step S3 is as follows: S301. Set a fixed time interval as the monitoring cycle, and within the current monitoring cycle, collect the operating data of the distribution network lines through the intelligent sensing system, including voltage, current, temperature and environmental parameters.

[0035] S302. The collected voltage, current, temperature and environmental parameters are processed to form the current cycle line information.

[0036] In the specific implementation process, step S4 is as follows: S401. Transmit the current cycle line information to the corresponding edge computing node. In the edge computing node, call the generated anomaly judgment threshold of the corresponding updated key node, and compare the voltage, current and temperature in the current cycle line information with the anomaly judgment threshold one by one to obtain the comparison analysis results.

[0037] Specifically, after organizing the current cycle line information, this information is accurately transmitted to the corresponding edge computing nodes. Upon receiving the current cycle line information, the edge computing nodes, based on the updated key node identifiers, retrieve the pre-generated anomaly detection thresholds for the corresponding key nodes from the stored database. The voltage, current, and temperature data in the current cycle line information are then compared one by one with the retrieved anomaly detection thresholds. Specifically, for voltage data, it is checked whether it is within the set normal voltage range (between the upper and lower thresholds); for current data, it is similarly determined whether it is within the normal current range; temperature data is compared in the same way. Through this step-by-step comparison, it is possible to accurately determine whether each parameter is abnormal.

[0038] S402. Based on the comparison and analysis results, output the node abnormal status identifier for each updated key node in the current cycle. If any one of the voltage, current, or temperature indicators exceeds the abnormal judgment threshold, it is marked as an abnormal state.

[0039] Specifically, based on the comparison and analysis results in step S401, a comprehensive judgment is made on the abnormality of the voltage, current, and temperature parameters of each updated key node. If any one of these three parameters exceeds the abnormality judgment threshold, the key node is determined to be in an abnormal state within the current cycle; only when all three parameters are within the normal range is the key node determined to be in a normal state. A clear node abnormality status identifier is output for each updated key node, represented by a simple string, such as "normal" indicating normal node operation, and "abnormal" indicating an abnormality. This identifier will be used in subsequent steps for further analysis and processing.

[0040] In the specific implementation process, step S5 is as follows: S501, aggregate the abnormal status identifiers of all nodes in the current period under different characteristic frequencies, i.e. different sampling frequencies.

[0041] Specifically, abnormal node status identifiers generated at different sampling frequencies are collected from various edge computing nodes. Since the timestamps of data from different sampling frequencies may differ, for the accuracy of subsequent analysis, all collected abnormal node status identifiers need to be unified to the time base of the current monitoring period. For example, if the current monitoring period is 1 minute, high-frequency sampled data needs to be aggregated into 1-minute intervals, retaining only the key abnormal status information within that 1-minute period; low-frequency sampled data, if its sampling time falls within the current monitoring period, is used directly. The abnormal node status identifiers at different characteristic frequencies, after being processed with a unified time base, are then aggregated into a centralized data storage area, such as a database table or data file.

[0042] In this embodiment, in a power distribution network monitoring system, the voltage of key node A is sampled at high frequency (10 times per second), the current at medium frequency (once every 10 seconds), and the temperature at low frequency (once per minute). The current monitoring cycle is 1 minute. When collecting data, the abnormal status identifiers (0 for normal and 1 for abnormal) corresponding to 600 voltage sampling data points within that minute are retrieved from the edge computing node cache storing voltage data. After aggregation, the overall abnormal status identifiers for the voltage within that minute are obtained (if the percentage of abnormal occurrences exceeds a certain threshold, it is marked as abnormal). The abnormal status identifiers corresponding to 6 current sampling data points within that minute are retrieved from the database storing current data, and are similarly aggregated. Since the temperature sampling frequency is consistent with the monitoring cycle, the abnormal temperature status identifiers within that minute are directly obtained. Finally, these node abnormal status identifiers processed under different characteristic frequencies are aggregated into a single database table.

[0043] S502. Based on the preset fluctuation deviation threshold and the abnormal status identifiers of all nodes in the current cycle, analyze the correlation of abnormal status of adjacent line nodes in the same time period, as well as the abnormal status change trend of the same line node in different time periods, and finally compile and generate the identification and analysis results.

[0044] Specifically, for adjacent line nodes in a distribution network, their operating states exhibit a certain correlation within the same time period. For example, a fault in one line may cause changes in parameters such as voltage and current of adjacent lines, thus affecting their abnormal states. A preset fluctuation deviation threshold is used to determine whether the abnormal states of adjacent nodes within the same time period exceed the normal correlation range. Specifically, the difference or ratio of certain key parameters (such as voltage and current) of adjacent nodes is calculated and compared with the fluctuation deviation threshold. If the difference or ratio exceeds the threshold, an abnormal correlation is considered to exist between the two adjacent nodes, suggesting a possible common fault cause or mutual influence. Secondly, for the same line node, changes in its abnormal state over different time periods can reflect the stability of its operating state and the trend of fault development. By comparing the abnormal state indicators of the same node in different monitoring cycles, its changing trends are analyzed. For example, if a node frequently exhibits abnormal states with similar abnormal types in several consecutive monitoring cycles, it can be inferred that the node may have potential fault hazards. The results of the above-mentioned correlation analysis of abnormal states of adjacent nodes and the analysis of the changing trends of abnormal states of the same node over different time periods are compiled and summarized to form the final identification and analysis results.

[0045] Example 2

[0046] like Figure 2 As shown, a power distribution network line abnormality identification system includes: The node allocation module is used to identify and determine the initial set of key nodes based on the topology analysis of the power distribution network, allocate edge computing nodes to the initial key nodes, and build a distributed edge computing network. The dynamic update module is used to establish an environmental perception mechanism based on the intelligent sensing system with multiple sensors and historical data, monitor the operating environment data of the power distribution network in real time, and dynamically correct the initial set of key nodes to form updated key nodes while setting their corresponding operating anomaly judgment thresholds. The data acquisition module is used to collect line operation data through the intelligent sensing system during the current monitoring period to form the line information for the current period, which includes voltage, current, temperature and environmental parameters. The comparison and analysis module is used to input the current cycle line information to the corresponding edge computing node, and use the generated anomaly judgment threshold to compare and analyze the running data of each updated key node, and output the node anomaly status identifier of each updated key node in the current cycle. The results generation module is used to collect the node anomaly status identifiers of each updated key node under different characteristic frequencies within the current period. Combined with the preset fluctuation deviation threshold, it performs spatiotemporal aggregation analysis on discrete node anomaly information to generate identification and analysis results.

[0047] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0048] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0049] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A method for identifying abnormal states of power distribution lines, characterized in that, Includes the following steps: S1. Based on the analysis of the power distribution network topology, identify and determine the initial set of key nodes, allocate edge computing nodes to the initial key nodes, and build a distributed edge computing network. S2. Based on a multi-sensor intelligent sensing system and historical data, an environmental perception mechanism is established to monitor the operating environment data of the power distribution network in real time, and the initial set of key nodes is dynamically corrected to form an updated set of key nodes while setting their corresponding operating anomaly judgment thresholds. S3. During the current monitoring cycle, the line operation data is collected through the intelligent sensing system to form the line information for the current cycle, which includes voltage, current, temperature and environmental parameters. S4. Input the current cycle line information to the corresponding edge computing node, use the generated anomaly judgment threshold to compare and analyze the running data of each updated key node, and output the node anomaly status identifier of each updated key node in the current cycle. S5. Collect the node anomaly status identifiers of each updated key node under different characteristic frequencies within the current period, and combine them with the preset fluctuation deviation threshold to perform spatiotemporal aggregation analysis on the discrete node anomaly information to generate identification and analysis results.

2. The method for identifying abnormal states of distribution network lines according to claim 1, characterized in that, Step S1 is as follows: S101. Read the topology data of the distribution network and organize the connection relationships of line nodes and load distribution characteristics; S102. Based on the electrical distance centerness and load importance, select the matching line nodes from the line node connection relationships and organize them into an initial set of key nodes. S103. Assign a corresponding edge computing node to each line node in the initial set of key nodes, establish a one-to-one or one-to-many mapping relationship, and build a distributed edge computing network.

3. The method for identifying abnormal states of distribution network lines according to claim 1, characterized in that, Step S2 is as follows: S201. Establish an environmental perception mechanism, the steps of which include: collecting historical sensing data and operating parameters collected by the intelligent sensing system, and arranging the historical sensing data and operating parameter data in chronological order to form time series data, wherein the intelligent sensing system includes multiple types of sensors; based on the time series data, analyzing the changing trend of the data using the moving average method to obtain time series analysis results; according to the time series analysis results, setting environmental data deviation thresholds for multiple time periods within the same period, and when the operating environmental data in the current time period exceeds its environmental data deviation threshold, it is included in the initial set of key nodes; S202. Monitor the distribution network operation environment data in real time, and adjust the initial set of key nodes based on the established environment perception mechanism, and correct it to the updated set of key nodes. S203. Collect the historical normal fluctuation range of each key node in the updated key node set, and set a corresponding operation anomaly judgment threshold for each key node in the updated key node set, wherein the operation anomaly judgment threshold is within the historical normal fluctuation range.

4. The method for identifying abnormal states of distribution network lines according to claim 1, characterized in that, Step S3 is as follows: S301. Set a fixed time interval as the monitoring cycle, and within the current monitoring cycle, collect the operating data of the distribution network lines through the intelligent sensing system, including voltage, current, temperature and environmental parameters. S302. The collected voltage, current, temperature and environmental parameters are processed to form the current cycle line information.

5. The method for identifying abnormal states of distribution network lines according to claim 1, characterized in that, Step S4 is as follows: S401. Transmit the current cycle line information to the corresponding edge computing node. In the edge computing node, call the generated anomaly judgment threshold of the corresponding updated key node, and compare the voltage, current and temperature in the current cycle line information with the anomaly judgment threshold one by one to obtain the comparison analysis results. S402. Based on the comparison and analysis results, output the node abnormal status identifier for each updated key node in the current cycle. If any one of the voltage, current, or temperature indicators exceeds the abnormal judgment threshold, it is marked as an abnormal state.

6. The method for identifying abnormal states of distribution network lines according to claim 1, characterized in that, Step S5 is as follows: S501, Aggregate the node abnormal status identifiers of all nodes in the current period under different characteristic frequencies, i.e., different sampling frequencies in the current period. S502. Based on the preset fluctuation deviation threshold and the abnormal status identifiers of all nodes in the current cycle, analyze the correlation of abnormal status of adjacent line nodes in the same time period, as well as the abnormal status change trend of the same line node in different time periods, and finally compile and generate the identification and analysis results.

7. A distribution network line abnormal state identification system, employing a distribution network line abnormal state identification method according to any one of claims 1-6, characterized in that, The identification system includes: The node allocation module is used to identify and determine the initial set of key nodes based on the topology analysis of the power distribution network, allocate edge computing nodes to the initial key nodes, and build a distributed edge computing network. The dynamic update module is used to establish an environmental perception mechanism based on the intelligent sensing system with multiple sensors and historical data, monitor the operating environment data of the power distribution network in real time, and dynamically correct the initial set of key nodes to form updated key nodes while setting their corresponding operating anomaly judgment thresholds. The data acquisition module is used to collect line operation data through the intelligent sensing system during the current monitoring period to form the line information for the current period, which includes voltage, current, temperature and environmental parameters. The comparison and analysis module is used to input the current cycle line information to the corresponding edge computing node, and use the generated anomaly judgment threshold to compare and analyze the running data of each updated key node, and output the node anomaly status identifier of each updated key node in the current cycle. The results generation module is used to collect the node anomaly status identifiers of each updated key node under different characteristic frequencies within the current period. Combined with the preset fluctuation deviation threshold, it performs spatiotemporal aggregation analysis on discrete node anomaly information to generate identification and analysis results.

Images