Power distribution network fault monitoring system and monitoring method

Abstract

The invention relates to the technical field of power distribution network fault monitoring, in particular to a power distribution network fault monitoring system and method. Performing primary processing and feature extraction on multi-source data by edge calculation; the distributed computing node further fuses and analyzes the data by using a multi-source data fusion technology to generate a fault feature vector; summarizing analysis results of the distributed computing nodes, performing comprehensive analysis on the fault feature vector, and determining a fault position; comparing and verifying the fault position result with historical data, and if uncertainty exists, starting a secondary diagnosis process; and according to a fault position result, fault early warning information is automatically sent to operation and maintenance personnel. By collecting multi-source data such as traveling wave signals, current and voltage and combining the data fusion technology of edge calculation and distributed calculation nodes, single data sources are effectively reduced, and the accuracy of fault positioning is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network fault monitoring, and in particular to a distribution network fault monitoring system and a monitoring method. Background Art

[0002] The distribution network plays a crucial role in the power system. Once a fault occurs in the distribution network, huge losses will be caused if it is not repaired in time. In a multi-terminal distribution network, the time difference of the traveling wave arriving at different measurement terminals is commonly used for precise positioning, but this method is inefficient.

[0003] The existing publication number CN114113948A discloses a distribution network fault monitoring method, including: respectively using the head end, the tail end of the distribution network, and the tail ends of each branch as monitoring points to monitor the traveling wave signals; constructing a monitoring tree for each monitoring point; in response to the occurrence of a distribution network fault, obtaining the first distance from each monitoring point to the fault point; obtaining the first fault branch set of the fault point on the monitoring tree of each monitoring point; screening and reorganizing each first fault branch set to obtain a second fault branch set; adding the distances from the fault point to the first node among the elements with the same suspected branches in the second fault branch set to obtain multiple judgment distances; and selecting the maximum value of the judgment distances as the main judgment distance, and the remaining judgment distances as the secondary judgment distances; obtaining the fault distance according to the main judgment distance and the secondary judgment distances; and further obtaining the specific position of the fault point in the distribution network. The fault monitoring and positioning efficiency and accuracy are improved.

[0004] However, the above-mentioned distribution network fault monitoring method highly depends on the monitoring and analysis of traveling wave signals. If the traveling wave signals are interfered during transmission, such as electromagnetic interference, noise, etc., it may cause changes in the characteristics of the traveling wave signals, such as waveform, propagation time, etc., thus affecting the accuracy of fault location. Summary of the Invention

[0005] The purpose of the present invention is to provide a distribution network fault monitoring system and a monitoring method, which solve the problem that the existing distribution network fault monitoring method highly depends on the monitoring and analysis of traveling wave signals. If the traveling wave signals are interfered during transmission, such as electromagnetic interference, noise, etc., it may cause changes in the characteristics of the traveling wave signals, such as waveform, propagation time, etc., thus affecting the accuracy of fault location.

[0006] To achieve the above purpose, the present invention provides a distribution network fault monitoring method, including the following steps:

[0007] Install multi-source data monitoring devices at the head end, the tail end of the distribution network, and the tail ends of each branch to collect multi-source data in real time;

[0008] Preprocess and extract features from the collected multi-source data through edge computing, and upload the extracted feature data to the nearest distributed computing node;

[0009] The distributed computing node receives the feature data, further fuses and analyzes the data using multi-source data fusion technology to generate a fault feature vector;

[0010] Summarize the analysis results of each distributed computing node, comprehensively analyze the fault feature vector, realize precise positioning and diagnosis of the distribution network fault, and determine the fault location;

[0011] Compare and verify the fault location result with historical data. If there is uncertainty, start the secondary diagnosis process;

[0012] Automatically send a fault warning message to the operation and maintenance personnel according to the fault location result.

[0013] Among them, multi-source data monitoring devices are installed at the head end, tail end and the end of each branch of the distribution network to collect multi-source data in real time. The steps also include:

[0014] The monitoring device includes a traveling wave signal sensor, a current transformer and a voltage transformer, and the multi-source data includes traveling wave signals, current and voltage.

[0015] Among them, preprocess and extract features from the collected multi-source data through edge computing, and upload the extracted feature data to the nearest distributed computing node. The steps also include:

[0016] Perform preliminary processing on the collected multi-source data, including filtering, denoising and normalization operations;

[0017] Extract the characteristic parameters of the multi-source data after preliminary processing, including the amplitude, frequency, phase, mutation point of the traveling wave signal, and the effective value, peak value, harmonic content, power factor of voltage and current;

[0018] Use principal component analysis or autoencoder to perform dimensionality reduction processing on the extracted characteristic parameters;

[0019] Pack the dimensionality-reduced feature data into a standard format, and attach a timestamp and a monitoring point identifier;

[0020] Upload the feature data packet to the nearest distributed computing node.

[0021] Among them, the distributed computing node receives the feature data, further fuses and analyzes the data using multi-source data fusion technology to generate a fault feature vector. The steps also include:

[0022] The distributed computing node unpacks and verifies the feature data packet, extracts the feature data and verifies its integrity and consistency;

[0023] Align the feature data according to the timestamp and perform normalization processing;

[0024] Assign weights to the feature data from different sources, and the weights are dynamically adjusted according to the importance and reliability of the data;

[0025] Fuse the normalized feature data to generate a comprehensive feature vector;

[0026] Concatenate the fused feature data into a high-dimensional fault feature vector;

[0027] Optimize the high-dimensional fault feature vector, remove redundant information and then store it.

[0028] Among them, summarize the analysis results of each distributed computing node, comprehensively analyze the fault feature vector, realize the accurate positioning and diagnosis of the distribution network fault, determine the fault location, and the steps also include:

[0029] According to the feature parameters in the high-dimensional fault feature vector, use the predefined fault mode library for matching analysis to determine the specific type and severity of the fault;

[0030] Combine the topological structure of the distribution network and the traveling wave propagation model to locate the fault location. The location methods include the traveling wave method, the impedance method or the location algorithm based on machine learning;

[0031] For the fault location estimates provided by multiple distributed computing nodes, use the weighted average method, the voting method or the Bayesian fusion method for comprehensive judgment to determine the final fault location.

[0032] Among them, compare the fault location result with the historical data for verification. If there is uncertainty, start the secondary diagnosis process, and the steps also include:

[0033] Retrieve historical fault cases similar to the current fault feature from the historical fault database;

[0034] Compare the current fault location with the fault location in the historical fault case and calculate the location deviation;

[0035] If the location deviation exceeds the preset threshold or there are significant differences in the fault features, it is determined that there is uncertainty. At this time, re-collect multi-source data at the moment of fault occurrence and perform preliminary processing and feature extraction;

[0036] Use the updated fault feature vector to re-perform fault location and diagnostic analysis.

[0037] A distribution network fault monitoring system, including a multi-source data monitoring module, an edge computing module, a distributed computing node, a fault diagnosis module, a historical data comparison module, and an early warning module. The edge computing module is connected to the multi-source data monitoring module, the distributed computing node is connected to the edge computing module, the fault diagnosis module is connected to the distributed computing node, the historical data comparison module is respectively connected to the fault diagnosis module and the multi-source data monitoring module, and the early warning module is connected to the historical data comparison module;

[0038] The multi-source data monitoring module is used to install multi-source data monitoring devices at the head end, end end, and the end of each branch of the distribution network to collect multi-source data in real time;

[0039] The edge computing module is used to preprocess and extract features from the collected multi-source data through edge computing, and upload the extracted feature data to the nearest distributed computing node;

[0040] The distributed computing node is used to receive the feature data, further fuse and analyze the data using multi-source data fusion technology, and generate a fault feature vector;

[0041] The fault diagnosis module is used to summarize the analysis results of each distributed computing node, comprehensively analyze the fault feature vector, achieve precise positioning and diagnosis of the distribution network fault, and determine the fault location;

[0042] The historical data comparison module is used to compare and verify the fault location result with historical data. If there is uncertainty, a secondary diagnosis process is started;

[0043] The early warning module is used to automatically send fault warning information to the operation and maintenance personnel according to the fault location result.

[0044] A fault monitoring system and method for a distribution network according to the present invention installs multi-source data monitoring devices at the head end, end and end of each branch of the distribution network, including traveling wave signal sensors, current transformers and voltage transformers, and real-time collects multi-source data such as traveling wave signals, currents and voltages. Through edge computing, the collected multi-source data is filtered, denoised, normalized, etc., and feature parameters are extracted. Principal component analysis or autoencoders are used for dimensionality reduction processing, and then it is packaged and uploaded to the nearest distributed computing node. The distributed computing node receives the feature data, unpacks, verifies, time-aligns and normalizes it, dynamically assigns weights and then performs data fusion to generate a high-dimensional fault feature vector and removes redundant information. Summarize the analysis results of each distributed computing node, and combine the fault mode library, the distribution network topology structure and the traveling wave propagation model, and use the traveling wave method, impedance method or machine learning algorithm to accurately locate and diagnose faults. Compare and verify the fault location results with historical data. If there is uncertainty, re-collect data and start a secondary diagnosis process. According to the finally determined fault location, automatically send a fault warning message to the operation and maintenance personnel. By collecting multi-source data such as traveling wave signals, currents and voltages, and combining the data fusion technology of edge computing and distributed computing nodes, the single data source is effectively reduced, such as the error caused by electromagnetic interference or noise affecting the traveling wave signal, and the accuracy of fault location is improved. Using edge computing for preliminary data processing and feature extraction reduces the amount of data transmission. At the same time, efficient data fusion and analysis are performed through distributed computing nodes, significantly improving the efficiency of fault monitoring and location. By comparing and verifying with historical data and starting a secondary diagnosis process in case of uncertainty, the reliability and accuracy of fault diagnosis are further improved. Using the traveling wave method, impedance method and location algorithms based on machine learning can meet the fault location requirements in different scenarios, enhancing the applicability and flexibility of the system. Automatically sending a fault warning message to the operation and maintenance personnel shortens the fault response time and reduces the losses caused by faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art.

[0046] Figure 1 It is a flowchart of the steps of the distribution network fault monitoring method according to the first embodiment of the present invention.

[0047] Figure 2 It is a schematic block diagram of the distribution network fault monitoring system according to the second embodiment of the present invention.

[0048] In the figure: 201 - multi-source data monitoring module, 202 - edge computing module, 203 - distributed computing node, 204 - fault diagnosis module, 205 - historical data comparison module, 206 - warning module. Detailed implementation manners

[0049] The embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the present invention, and should not be construed as limiting the present invention.

[0050] The first embodiment of the present application is as follows:

[0051] Please refer to Figure 1 , where Figure 1 is the flowchart of the steps of the distribution network fault monitoring method according to the first embodiment of the present invention.

[0052] The present invention provides a distribution network fault monitoring method, including the following steps:

[0053] S101: Install multi-source data monitoring devices at the head end, the end end and the end of each branch of the distribution network, and collect multi-source data in real time;

[0054] Specifically, the monitoring device includes a traveling wave signal sensor, a current transformer, a voltage transformer, and an intelligent terminal device, etc. Among them, the traveling wave signal sensor is used to capture the transient voltage and current traveling wave signals generated when a fault occurs. The traveling wave signal sensor can record the waveform characteristics of the fault traveling wave at a high sampling rate, such as above 100 kHz, including amplitude, frequency, phase, and mutation points. The current transformer is used to monitor the current changes in the distribution network in real time. The current transformer can measure characteristics such as the effective value, peak value, harmonic content, and power factor of the current. The voltage transformer is used to monitor the voltage changes in the distribution network in real time. The voltage transformer can measure characteristics such as the effective value, peak value, harmonic content, and power factor of the voltage. The intelligent terminal device integrates data acquisition, preprocessing, storage, and communication functions. The intelligent terminal device supports edge computing and can perform preliminary processing and feature extraction on the collected data and upload the processed data to the distributed computing node 203. Installing the multi-source data monitoring device at the starting position of the distribution network can monitor the changes in electrical parameters on the power supply side. Installed at the end of the distribution network, it can capture the changes in electrical parameters on the load side. Installed at the ends of each branch of the distribution network, it can monitor the changes in electrical parameters of the branch line and detect faults in the branch line in a timely manner. During monitoring, the traveling wave signal sensor monitors the transient voltage and current changes at a high sampling rate, such as 100 kHz, in real time to ensure that high-frequency transient signals can be captured, timestamp the traveling wave signals, and record the accurate time information of the signals. The current transformer monitors the current changes in the distribution network in real time, including the effective value, peak value, harmonic content, and power factor. The sampling frequency is set to 1 kHz to 10 kHz to capture the dynamic changes of the current, timestamp the current data, and record the accurate time information of the data. The voltage transformer monitors the voltage changes in the distribution network in real time, including the effective value, peak value, harmonic content, and power factor. The sampling frequency is set to 1 kHz to 10 kHz to capture the dynamic changes of the voltage. Timestamp the voltage data and record the accurate time information of the data.

[0055] S102: Preprocess and extract features from the collected multi-source data through edge computing, and upload the extracted feature data to the nearest distributed computing node 203;

[0056] Specifically, a low-pass filter is used to remove high-frequency noise and smooth the signal, and wavelet transform or adaptive filtering algorithm is used to further remove noise and retain the useful components of the signal. Normalize the processed data to a unified range, such as 0 to 1, for subsequent processing. The normalization formula is: where X is the original data, X min and X maxThey are the minimum and maximum values of the data respectively. Extract the fault-related characteristic parameters from the preprocessed data, including the characteristic amplitude of the traveling wave signal: the maximum value of the traveling wave signal. Frequency: the main frequency component of the traveling wave signal. Phase: the phase information of the traveling wave signal. Mutation point: the mutation position of the traveling wave signal, used to capture the fault occurrence moment. Effective value (RMS) of the current characteristic: the effective value of the current. Peak value: the maximum value of the current. Harmonic content (THD): the total harmonic distortion rate of the current. Power factor: the power factor of the current. Effective value (RMS) of the voltage characteristic: the effective value of the voltage. Peak value: the maximum value of the voltage. Harmonic content (THD): the total harmonic distortion rate of the voltage. Power factor: the power factor of the voltage. Use principal component analysis (PCA) or autoencoders to perform dimensionality reduction on the extracted characteristic parameters, reduce the amount of data, and retain key information at the same time. The dimensionality-reduced characteristic data can be transmitted and processed more efficiently. Pack the dimensionality-reduced characteristic data into a standard format, and attach a timestamp and a monitoring point identifier. Upload the packed characteristic data to the nearest distributed computing node 203 through wireless communication, such as 4G / 5G, or fiber optic communication. Ensure the reliability and real-time performance of data transmission, and avoid data loss or delay. Edge computing is performed near the data source, reducing data transmission latency and improving the real-time performance of the system. Through feature extraction and dimensionality reduction, the amount of data uploaded to the distributed computing node 203 is reduced, and the communication cost is lowered. Even in the case of an unstable communication link, the edge computing module 202 can still complete preliminary processing to ensure data integrity. The data is preliminarily processed locally, reducing the exposure risk of the data during transmission and enhancing the security of the system.

[0057] S103: The distributed computing node 203 receives the characteristic data, and uses multi-source data fusion technology to further fuse and analyze the data to generate a fault feature vector;

[0058] Specifically, the distributed computing node 203 receives the characteristic data packet from edge computing. The data packet format should include a timestamp, a monitoring point identifier, and characteristic data. Perform integrity verification on the received data packet to ensure that the data has not been lost or damaged during transmission. Verify the timestamp and monitoring point identifier of the data to ensure the source and time consistency of the data. Align the characteristic data from different monitoring points according to the timestamp to ensure the consistency of the data in the time dimension. The alignment method uses interpolation or sliding window technology to handle time deviations. Perform normalization processing on the characteristic data to unify the data from different sources to the same range, such as 0 to 1. Assign weights to the characteristic data from different sources, and the weights are dynamically adjusted according to the importance and reliability of the data. For example, the weight of the traveling wave signal may be higher than that of the current or voltage data because the traveling wave signal is more sensitive to faults. Perform weighted averaging on the normalized characteristic data to generate a comprehensive feature vector. The calculation formula is: Where, Fi is the feature vector of the i-th monitoring point, and ω i is the weight. The fused feature data is concatenated into a high-dimensional fault feature vector. The fault feature vector should contain the characteristic parameters of multi-source data such as traveling wave signals, current, and voltage. Use principal component analysis (PCA) or autoencoders to reduce the dimension of the high-dimensional feature vector and remove redundant information. The optimized fault feature vector should retain the key fault features and improve the efficiency of subsequent fault diagnosis. Store the generated fault feature vector in the local memory of the distributed computing node 203 for subsequent analysis. The storage format should include the timestamp, the monitoring point identifier, and the fault feature vector. Upload the fault feature vector to the next step for global fault diagnosis and location. Through the distributed computing architecture, the data processing tasks are dispersed to multiple nodes, improving the overall efficiency of the system. Even if some nodes fail, the system can still operate normally, improving the fault tolerance of the system. The distributed computing node 203 can quickly process and fuse data to ensure the real-time generation of the fault feature vector. The system can increase or decrease the distributed computing node 203 according to needs, flexibly adapting to different scales of distribution networks.

[0059] S104: Summarize the analysis results of each distributed computing node 203, comprehensively analyze the fault feature vector, achieve precise location and diagnosis of the distribution network fault, and determine the fault location;

[0060] Specifically, construct a predefined fault mode library that contains the feature vector templates of different fault types, such as short circuit, open circuit, ground fault, etc. Each template contains the typical values and ranges of the fault feature vector. Perform matching analysis on the summarized fault feature vector and the templates in the fault mode library. Use similarity metrics, such as Euclidean distance and cosine similarity, to calculate the similarity between the feature vector and the template. Select the template with the highest similarity as the basis for judging the fault type. According to the fault feature vector and the topological structure of the distribution network, use the following methods to accurately locate the fault location: Traveling wave method: Utilize the propagation characteristics of traveling wave signals to calculate the distance from the fault point to each monitoring point. According to the time difference of arrival (TDOA) formula of traveling wave: where Δt is the time difference of arrival of the traveling wave, d is the distance from the fault point to the monitoring point, and v is the propagation speed of the traveling wave. Combine the time difference data of multiple monitoring points and determine the fault location through geometric location algorithms (such as trilateration). Impedance method: Utilize the current and voltage data to calculate the impedance of the fault point. According to the impedance formula: Among them, Z is the fault impedance, V is the fault voltage, and I is the fault current. Combining with the line parameters of the distribution network, the fault location is determined through impedance calculation. Location algorithm based on machine learning: Using machine learning models such as support vector machines and neural networks to train and classify the fault feature vectors. Input the feature vectors, and the model outputs the estimated value of the fault location. Optimize the model parameters through the training data set to improve the location accuracy. For the fault location estimates provided by multiple distributed computing nodes 203, the following methods are used for comprehensive judgment: Weighted average method: Different weights are assigned to the fault location estimates of each node, and the weights are dynamically adjusted according to the reliability of the nodes. Calculate the weighted average value as the final fault location: Among them, x i is the estimated fault location of the i-th node, and ω i is the weight. Voting method: Each node votes on the classification of the fault location, and selects the fault location with the most votes as the final result. Bayesian fusion method: Use Bayes' theorem to fuse the fault location estimates of multiple nodes, and calculate the posterior probability of each location. Select the location with the highest posterior probability as the final fault location. Based on the fault location estimates generated by the above methods, combined with the topological structure and operating status of the distribution network, the final fault location is determined. The fault location should include specific line numbers, branch locations, and distance information. Output the fault location information in a standardized format, including fault type, location coordinates, estimated error range, etc.

[0061] S105: Compare and verify the fault location result with the historical data. If there is uncertainty, start the secondary diagnosis process;

[0062] Specifically, retrieve historical fault cases similar to the current fault characteristics from the historical fault database. The retrieval conditions include fault type, similarity of feature vectors such as Euclidean distance and cosine similarity, and fault location information. Compare the current fault location with the fault locations in the historical fault cases, and calculate the location deviation. The location deviation formula is: Δd = ∣d current -d history ∣, where d current is the current fault location, and d historyThis is the historical fault location. Compare the differences between the current fault feature vector and the historical fault feature vector. Use similarity measures such as Euclidean distance and cosine similarity to evaluate the similarity of the feature vectors. Preset the location deviation threshold and the feature difference threshold (such as the similarity measure threshold). If the location deviation exceeds the preset threshold or the feature difference exceeds the preset threshold, it is determined that there is uncertainty. If the verification result shows uncertainty, start the secondary diagnosis process to further confirm the fault location and type. That is, re-collect data, perform data preprocessing and feature extraction, conduct data fusion and analysis, and compare and confirm the results. Compare the results of the secondary diagnosis with the results of the primary diagnosis and evaluate the differences. If the results of the secondary diagnosis are consistent with the results of the primary diagnosis, confirm the fault location and type. If there is still uncertainty in the results of the secondary diagnosis, make a final judgment based on the experience of on-site maintenance personnel. After confirming the fault location and type, generate a final fault diagnosis report. The report content includes the fault type, location, diagnosis method, verification process, and reliability assessment. Among them, conduct a reliability assessment of the final fault diagnosis result and calculate the confidence level. The confidence level calculation formula is: Among them, similarity is the similarity measure between the final fault feature vector and the historical data. Output the fault diagnosis result in a standardized format, including information such as the fault type, location, and confidence level.

[0063] By comparing historical data and re-collected data, reduce the possibility of misjudgment and missed judgment. Start the secondary diagnosis under uncertain conditions to ensure the reliability of the fault diagnosis result. Provide a detailed verification process and confidence level assessment to help maintenance personnel respond to faults quickly.

[0064] S106: Automatically send a fault warning message to the maintenance personnel according to the fault location result.

[0065] Specifically, pre-configure the contact information of the maintenance personnel, including mobile phone numbers, email addresses, or instant messaging tool accounts. According to the severity and location of the fault, send the warning message to the relevant maintenance team or individual. Store the fault warning message in the system database, recording the sending time, receiving personnel, fault details, etc. The system generates a detailed log file to record the generation, sending, and receiving process of the fault warning message. Provide a traceability function for the fault warning message, supporting queries of historical records according to conditions such as time, fault type, and location. Conduct statistical analysis on the fault warning message and generate a fault report to help maintenance personnel understand the distribution law and processing efficiency of faults.

[0066] By collecting multi-source data such as traveling wave signals, current, and voltage, and combining edge computing and data fusion technology of distributed computing node 203, the single data source is effectively reduced, such as the error caused by electromagnetic interference or noise affecting the traveling wave signal, improving the accuracy of fault location. Using edge computing for preliminary data processing and feature extraction reduces the amount of data transmission. At the same time, through the distributed computing node 203 for efficient data fusion and analysis, the efficiency of fault monitoring and location is significantly improved. By comparing and verifying with historical data and starting a secondary diagnosis process under uncertain conditions, the reliability and accuracy of fault diagnosis are further improved. Adopting traveling wave method, impedance method, and location algorithms based on machine learning can adapt to the fault location requirements in different scenarios, enhancing the applicability and flexibility of the system. Automatically sending fault warning information to operation and maintenance personnel shortens the fault response time and reduces the losses caused by faults.

[0067] The second embodiment of this application is as follows:

[0068] Based on the first embodiment, please refer to Figure 2 , where Figure 2 is the principle block diagram of the distribution network fault monitoring system according to the second embodiment of the present invention.

[0069] A distribution network fault monitoring system in this embodiment includes a multi-source data monitoring module 201, an edge computing module 202, a distributed computing node 203, a fault diagnosis module 204, a historical data comparison module 205, and a warning module 206.

[0070] For this specific embodiment, the edge computing module 202 is connected to the multi-source data monitoring module 201, the distributed computing node 203 is connected to the edge computing module 202, the fault diagnosis module is connected to the distributed computing node 203, the historical data comparison module 205 is respectively connected to the fault diagnosis module 204 and the multi-source data monitoring module 201, and the warning module 206 is connected to the historical data comparison module 205;

[0071] The multi-source data monitoring module 201 is used to install multi-source data monitoring devices at the head end, end end, and each branch end of the distribution network to collect multi-source data in real time;

[0072] The edge computing module 202 is used to preprocess and extract features from the collected multi-source data through edge computing, and upload the extracted feature data to the nearest distributed computing node 203;

[0073] The distributed computing node 203 is used to receive feature data, further fuse and analyze the data using multi-source data fusion technology, and generate a fault feature vector;

[0074] The fault diagnosis module 204 is used to summarize the analysis results of each distributed computing node 203, comprehensively analyze the fault feature vector, realize the accurate positioning and diagnosis of the distribution network fault, and determine the fault location.

[0075] The historical data comparison module 205 is used to compare and verify the fault location result with the historical data. If there is uncertainty, the secondary diagnosis process is started.

[0076] The early warning module 206 is used to automatically send fault warning information to the operation and maintenance personnel according to the fault location result.

[0077] Using a distribution network fault monitoring system according to this embodiment, the multi-source data monitoring module 201 collects multi-source data in real time and uploads it to the edge computing module 202. The edge computing module 202 preprocesses and extracts features from the collected multi-source data, and uploads the extracted feature data to the nearest distributed computing node 203. The distributed computing node 203 uses multi-source data fusion technology to further fuse and analyze the data, generates a fault feature vector, and uploads it to the fault diagnosis module 204. The fault diagnosis module 204 summarizes the analysis results of each distributed computing node 203, comprehensively analyzes the fault feature vector, realizes the accurate positioning and diagnosis of the distribution network fault, determines the fault location, and uploads it to the historical data comparison module 205. The historical data comparison module 205 compares and verifies the fault location result with the historical data. If there is uncertainty, it uploads a signal to the multi-source data monitoring module 201 to start the secondary diagnosis process. Finally, the early warning module 206 receives the fault location result and automatically sends fault warning information to the operation and maintenance personnel.

[0078] Combining multi-source data (traveling wave signals, current, voltage) and multi-source data fusion technology can reduce the error of a single data source and improve the accuracy of fault positioning and diagnosis. Using an edge computing and distributed computing architecture, the data processing tasks are dispersed to multiple nodes, reducing the computing pressure on the central system and improving the overall efficiency and real-time performance of the system. Through historical data comparison verification and the secondary diagnosis process, the reliability of the fault diagnosis result is ensured, and the possibility of misjudgment and missed judgment is reduced. Automatically sending fault warning information to the operation and maintenance personnel shortens the fault response time and reduces the power outage time and economic losses.

[0079] The above-disclosed are only one or more preferred embodiments of the present application, and the scope of rights of the present application cannot be limited thereby. Those of ordinary skill in the art can understand all or part of the processes of implementing the above embodiments, and the equivalent changes made according to the claims of the present application still fall within the scope covered by the present application.

Claims

1. A distribution network fault monitoring method, characterized in that: The following steps are involved: Install multi-source data monitoring devices at the head end, end end and end end of each branch of the distribution network to collect multi-source data in real time; Preprocess and extract features of the collected multi-source data through edge computing, and upload the extracted feature data to the nearest distributed computing node; The distributed computing nodes receive the feature data, and further fuse and analyze the data using multi-source data fusion technology to generate fault feature vectors; Summarize the analysis results of each distributed computing node, conduct a comprehensive analysis of the fault feature vector, accurately locate and diagnose the distribution network fault, and determine the fault location; Compare and verify the fault location results with historical data. If there is uncertainty, start the secondary diagnosis process; According to the fault location results, fault warning information is automatically sent to the operation and maintenance personnel.

2. The distribution network fault monitoring method according to claim 1, characterized in that: A multi-source data monitoring device is installed at the head end, the end end and the end end of each branch of the distribution network to collect multi-source data in real time. The steps also include: The monitoring device includes a traveling wave signal sensor, a current transformer and a voltage transformer, and the multi-source data includes a traveling wave signal, a current and a voltage.

3. The distribution network fault monitoring method according to claim 2, characterized in that: Preprocessing and feature extraction of the collected multi-source data are performed through edge computing, and the extracted feature data is uploaded to the nearest distributed computing node. The steps also include: The collected multi-source data is preliminarily processed, including filtering, denoising and normalization operations, where the normalization formula is: Among them, X is the original data, X min and X max are the minimum and maximum values ​​of the data respectively; Extract characteristic parameters of multi-source data after preliminary processing, including amplitude, frequency, phase, mutation point of traveling wave signal, effective value, peak value, harmonic content and power factor of voltage and current; Use principal component analysis or autoencoder to reduce the dimension of the extracted feature parameters; Package the reduced feature data into a standard format and attach timestamps and monitoring point identifiers; Upload the feature data package to the nearest distributed computing node.

4. The method for monitoring distribution network faults according to claim 3, characterized in that: The distributed computing node receives the feature data, further fuses and analyzes the data using a multi-source data fusion technology, and generates a fault feature vector. The steps also include: The distributed computing nodes unpack and verify the feature data packets, extract the feature data and verify its integrity and consistency; Time-align the feature data according to the timestamp and perform normalization; Assign weights to feature data from different sources, and the weights are dynamically adjusted based on the importance and reliability of the data; The normalized feature data is fused to generate a comprehensive feature vector. The calculation formula is: Among them, F i is the feature vector of the i-th monitoring point, ω i is the weight; The fused feature data are spliced ​​into a high-dimensional fault feature vector; The high-dimensional fault feature vector is optimized and stored after removing redundant information.

5. The distribution network fault monitoring method according to claim 4, characterized in that: Summarizing the analysis results of each distributed computing node, comprehensively analyzing the fault feature vector, accurately locating and diagnosing the distribution network fault, and determining the fault location, the steps also include: According to the characteristic parameters in the high-dimensional fault feature vector, a matching analysis is performed using a predefined fault mode library to determine the specific type and severity of the fault; Combine the topological structure of the distribution network and the traveling wave propagation model to locate the fault location. The positioning methods include the traveling wave method, the impedance method or the positioning algorithm based on machine learning. For the fault location estimates provided by multiple distributed computing nodes, a weighted average method, voting method or Bayesian fusion method is used for comprehensive judgment to determine the final fault location, where the weighted average value is calculated as the final fault location: Among them, x i is the fault location estimate of the i-th node, ω i is the weight.

6. The distribution network fault monitoring method according to claim 5, characterized in that: The fault location result is compared and verified with historical data. If there is uncertainty, a secondary diagnosis process is initiated, and the steps further include: Retrieve historical fault cases with similar fault characteristics to the current fault from the historical fault database; Compare the current fault location with the fault location in historical fault cases and calculate the location deviation; If the position deviation exceeds the preset threshold or there is a significant difference in the fault characteristics, it is determined that there is uncertainty. At this time, the multi-source data at the time of the fault occurrence is re-collected and preliminary processing and feature extraction are performed; The updated fault feature vector is used to re-perform fault location and diagnosis analysis.

7. A distribution network fault monitoring system, applicable to the distribution network fault monitoring method according to claim 1, characterized in that: It includes a multi-source data monitoring module, an edge computing module, a distributed computing node, a fault diagnosis module, a historical data comparison module and an early warning module, wherein the edge computing module is connected to the multi-source data monitoring module, the distributed computing node is connected to the edge computing module, the fault diagnosis module is connected to the distributed computing node, the historical data comparison module is connected to the fault diagnosis module and the multi-source data monitoring module respectively, and the early warning module is connected to the historical data comparison module; The multi-source data monitoring module is used to install multi-source data monitoring devices at the head end, the end and the end of each branch of the distribution network to collect multi-source data in real time; The edge computing module is used to preprocess and extract features of the collected multi-source data through edge computing, and upload the extracted feature data to the nearest distributed computing node; The distributed computing node is used to receive feature data, further fuse and analyze the data using multi-source data fusion technology, and generate a fault feature vector; The fault diagnosis module is used to summarize the analysis results of each distributed computing node, conduct a comprehensive analysis of the fault feature vector, achieve accurate positioning and diagnosis of the distribution network fault, and determine the fault location; The historical data comparison module is used to compare and verify the fault location result with the historical data, and if there is uncertainty, a secondary diagnosis process is initiated; The early warning module is used to automatically send fault early warning information to the operation and maintenance personnel according to the fault location result.

Images