Fault positioning and recovery decision-making method and system for power distribution network
By performing fuzzy analysis and fault prediction on the abnormal simulation characteristics of the distribution network, and utilizing a multi-source sensor monitoring array and a digital twin topology network, the complexity and uncertainty of distribution network fault identification are solved, enabling efficient fault location and recovery decision-making, and improving the accuracy and efficiency of fault prediction.
Patent Information
- Application Number
- CN202410985734.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-23
- Publication Date
- 2024-11-26
AI Technical Summary
The existing distribution network fault identification is complex and uncertain, resulting in low fault prediction accuracy and efficiency, lack of flexibility and adaptability, and fault recovery strategies that rely on the experience of operation and maintenance personnel are difficult to guarantee optimal results.
By performing fuzzy analysis on the abnormal simulation characteristics, a fault prediction sub-channel is built. By utilizing a multi-source sensor monitoring array and a digital twin topology network, combined with fuzzy analysis and fault prediction models, fault location and recovery decisions are realized, generating accurate fault recovery solutions.
It improves the accuracy and efficiency of fault prediction, reduces the workload of maintenance personnel, shortens the fault finding time, improves the automation and intelligence of fault diagnosis and recovery, and enhances power supply reliability.
Smart Images

Figure CN119024094A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of electrical data processing, and specifically to a fault location and recovery decision method and system for a distribution network. Background Art
[0002] With the development of smart grids, the automatic fault location technology of distribution networks has become a basic work to ensure the safe and reliable operation of smart distribution networks. The fault location technology of distribution networks is based on fault indicator technology and GIS technology, and is mainly used for the detection and location of various fault points in the distribution system, including phase-to-phase short circuit and single-phase grounding fault. This technology mainly involves fault line selection, section location, fault distance measurement, etc., aiming to quickly and accurately locate the fault point for subsequent fault recovery work. The distribution network has a complex network structure, many line branches, and various neutral point grounding methods. Compared with the traditional transmission network fault location technology, the concept of automatic fault location technology of distribution networks is broader and more complex to implement. This makes the accuracy and reliability of fault location technology subject to certain challenges. At present, the fault recovery strategy of distribution networks is mostly based on preset recovery plans, which lack flexibility and adaptability for specific fault conditions. At the same time, the formulation of fault recovery strategies often depends on the experience and judgment of operation and maintenance personnel, and it is difficult to ensure the optimal recovery effect.
[0003] In summary, the existing technology has the technical problem that due to the influence of various distribution factors, the complexity and uncertainty of distribution fault identification are relatively large, resulting in low distribution network fault prediction accuracy and efficiency. Summary of the invention
[0004] Based on this, it is necessary to provide a method and system for fault location and recovery decision-making in distribution networks that can perform fuzzy analysis on abnormal simulation characteristics, accurately set prediction nodes, build fault prediction sub-channels to execute distribution fault prediction of prediction nodes, and improve prediction accuracy and efficiency in response to the above technical problems.
[0005] In a first aspect, a fault location and recovery decision method for a distribution network is provided, the method comprising: identifying a target distribution system and determining a target distribution topology network, wherein the target distribution topology network is embedded with a plurality of monitoring nodes, and each monitoring node has a mapping-associated coverage device or / and coverage line; performing fault feature analysis based on the coverage device or / and coverage line to determine a plurality of monitoring feature information, and building a multi-source sensor monitoring array based on the plurality of monitoring feature information and the plurality of monitoring nodes, wherein the multi-source sensor monitoring array is embedded with a feature monitoring threshold; utilizing the multi-source sensor monitoring array to perform real-time monitoring of the target distribution system, and extracting abnormal information of a plurality of abnormal monitoring nodes that do not meet the feature monitoring threshold; characteristic data set; configuring multiple simulation operation windows of the multiple abnormal monitoring nodes, performing power distribution simulation in combination with the target twin topology network and the abnormal characteristic data set, and obtaining a simulation characteristic data set; inputting the simulation characteristic data set into a predetermined fault prediction channel for fuzzy analysis, performing fault prediction based on the fuzzy analysis results, and outputting multiple fault prediction results, wherein the fault prediction results include fault location and fault type; analyzing the fault location and fault type through a predetermined recovery decision channel, determining a matching maintenance plan and an optimized maintenance route; generating a fault recovery plan based on the fault location, fault type, matching maintenance plan, and optimized maintenance route, and sending the fault recovery plan to the operation and maintenance personnel of the associated maintenance node.
[0006] In a second aspect, a fault location and recovery decision system for a distribution network is provided, the system comprising: a target distribution topology network determination module, the target distribution topology network determination module is used to identify the target distribution system and determine the target distribution topology network, wherein the target distribution topology network is embedded with multiple monitoring nodes, and each monitoring node has a mapping-associated coverage device or / and coverage line; a fault feature analysis module, the fault feature analysis module is used to perform fault feature analysis based on the coverage device or / and coverage line, determine multiple monitoring feature information, and build a multi-source sensor monitoring array based on the multiple monitoring feature information and the multiple monitoring nodes, wherein the multi-source sensor monitoring array is embedded with a feature monitoring threshold; a real-time monitoring module, the real-time monitoring module is used to use the multi-source sensor monitoring array to perform real-time monitoring of the target distribution system, and extract abnormal feature data sets of multiple abnormal monitoring nodes that do not meet the feature monitoring threshold; simulated operation window configuration A configuration module, wherein the simulation operation window configuration module is used to configure multiple simulation operation windows of the multiple abnormal monitoring nodes, perform power distribution simulation in combination with the target twin topology network and the abnormal feature data set, and obtain a simulation feature data set; a fuzzy analysis module, wherein the fuzzy analysis module is used to input the simulation feature data set into a predetermined fault prediction channel for fuzzy analysis, perform fault prediction based on the fuzzy analysis results, and output multiple fault prediction results, wherein the fault prediction results include fault location and fault type; a fault location and type analysis module, wherein the fault location and type analysis module is used to analyze the fault location and fault type through a predetermined recovery decision channel, determine a matching maintenance plan and an optimized maintenance route; a fault recovery plan generation module, wherein the fault recovery plan generation module is used to generate a fault recovery plan based on the fault location, fault type, matching maintenance plan and optimized maintenance route, and send the fault recovery plan to the operation and maintenance personnel of the associated maintenance node.
[0007] The above-mentioned fault location and recovery decision-making method and system for distribution networks solve the technical problem in the prior art that due to the influence of various distribution factors, the complexity and uncertainty of distribution fault identification are large, resulting in low distribution network fault prediction accuracy and efficiency. By performing fuzzy analysis on abnormal simulation characteristics, accurately setting prediction nodes, and building fault prediction sub-channels to execute distribution fault prediction of prediction nodes, the technical effect of improving prediction accuracy and efficiency is achieved.
[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] Figure 1 A schematic diagram of a flow chart of a fault location and restoration decision method for a distribution network in one embodiment;
[0010] Figure 2 A schematic diagram of a flow chart of adding a fault location and restoration decision method for a distribution network to the plurality of monitoring feature information in one embodiment;
[0011] Figure 3 The figure is a structural block diagram of a fault location and restoration decision system for a distribution network in one embodiment.
[0012] Explanation of the reference numerals: target power distribution topology network determination module 11, fault characteristic analysis module 12, real-time monitoring module 13, simulation operation window configuration module 14, fuzzy analysis module 15, fault location and type analysis module 16, fault recovery plan generation module 17. DETAILED DESCRIPTION
[0013] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0014] like Figure 1 As shown, the present application provides a fault location and restoration decision method for a distribution network, the method comprising:
[0015] Identify a target power distribution system and determine a target power distribution topology network, wherein the target power distribution topology network has multiple monitoring nodes embedded therein, and each monitoring node has a mapping-associated coverage device or / and coverage line;
[0016] A distribution network refers to a power network that receives electrical energy from a transmission network or a regional power plant, and distributes it locally or step by step according to voltage to various types of users through distribution facilities. Fault location is mainly used to detect and locate various fault points in the distribution system, including short circuits and single-phase grounding faults; after the fault is located, the recovery decision system can help operators to formulate operating methods in a timely and reasonable manner to reduce their workload, improve work efficiency, and ensure the correctness of the plan. The present application provides a fault location and recovery decision method for a distribution network, which effectively reduces the labor intensity of distribution line patrol personnel, shortens the time for finding faults, improves the automation and intelligence of fault diagnosis and recovery, quickly and accurately diagnoses faults, automatically matches fault recovery strategies, improves the accuracy, timeliness and efficiency of distribution fault inspection and repair, and greatly improves power supply reliability.
[0017] Identify the target distribution system, which refers to the distribution system selected from multiple distribution networks for research, clarify the scope and boundaries of the target distribution system, including all relevant equipment and lines, collect basic information about the target distribution system, such as system capacity, voltage level, equipment type, etc., and build a distribution topology network based on the equipment and line connection relationship of the target distribution system. In the topology network diagram, clarify the location, type and connection relationship between each device and line. The target distribution topology network is embedded with multiple monitoring nodes, which refer to the location of the monitoring targets such as the monitored equipment and lines. Determine the location and type of each monitoring node, as well as the equipment or / and lines they cover, create a mapping relationship table, and associate each monitoring node with the equipment or / and lines it covers. Include the monitoring nodes in the distribution topology network, and mark their mapping relationship with the covered equipment or / and lines. Through the above steps, the target distribution system can be effectively identified and its distribution topology network can be determined, and the mapping association between the monitoring nodes and the covered equipment or / and lines can be realized. This helps to improve the operating efficiency and reliability of the distribution system and provide strong support for its subsequent maintenance and management.
[0018] Perform fault feature analysis based on the covered device and / or covered line, determine multiple monitoring feature information, and build a multi-source sensor monitoring array according to the multiple monitoring feature information and the multiple monitoring nodes, wherein the multi-source sensor monitoring array is embedded with a feature monitoring threshold;
[0019] Conduct detailed fault feature analysis on the covered equipment and / or covered lines in the target power distribution system, including analyzing the electrical characteristics, mechanical characteristics, thermal characteristics, etc. of the equipment or lines during normal operation; for possible fault modes, such as short circuit, overload, ground fault, etc., study the corresponding fault feature performance, such as voltage fluctuation, current anomaly, temperature change, etc., and determine the feature information that needs to be monitored based on the results of the fault feature analysis. The feature information should be able to directly or indirectly reflect the health status and fault status of the equipment or line, including voltage, current, temperature, power factor, harmonic content, etc. Depending on the specific application scenario and device type, the feature information that needs to be monitored will be different. Select appropriate monitoring node locations in the power distribution system to ensure that the equipment or lines that need to be monitored can be fully covered. According to the requirements of the monitoring feature information, configure corresponding sensors for the multiple monitoring nodes to accurately measure and record the required feature information. Connect the multiple monitoring nodes by wire or wireless means to obtain a multi-source sensor monitoring array to ensure that the multi-source sensor monitoring array can collect the monitoring data of each node in real time and accurately. For each monitored characteristic information, a corresponding characteristic monitoring threshold is set. The characteristic monitoring threshold is determined based on the normal operating range of the equipment or line, safety standards, and the results of fault characteristic analysis, and is used to determine whether the equipment or line is in a normal state or is about to fail. When the monitoring data exceeds the characteristic monitoring threshold, the system will issue an alarm or take corresponding control measures. Through the above method, fault characteristic analysis is performed based on the covered equipment and / or covered lines, multiple monitoring characteristic information is determined, and a multi-source sensor monitoring array is built accordingly. It helps to improve the operating safety and reliability of the distribution system, and provide strong support for its fault warning and fault location.
[0020] like Figure 2 As shown, a first monitoring node is randomly selected, and a first coverage device and / or a first coverage line of the first monitoring node are obtained;
[0021] Taking the first covered device and / or the first covered line as constraints, querying historical power distribution records to obtain a first fault data set;
[0022] Performing fault feature analysis based on the first fault data set to determine a first monitoring feature set and a first fault triggering frequency that meet a predetermined frequency threshold;
[0023] First monitoring feature information is obtained according to the first monitoring feature set and the first fault triggering frequency, and is added to the plurality of monitoring feature information.
[0024] A monitoring node is randomly selected from the multi-source sensor monitoring array as the first monitoring node. This selection is random to ensure that different devices and lines in the system have the same possibility for fault feature analysis; according to the previously set mapping relationship, the device or / and line covered by the first monitoring node is queried and determined, and the device or line will be used as a constraint for the subsequent fault feature analysis. In the historical records of the power distribution system, historical data related to the first covered device or / and the first covered line are searched, including monitoring data such as voltage, current, temperature, power factor, and related operation records, fault records, etc. Fault data related to the first covered device or / and the first covered line are screened out from the historical distribution records to form a first fault data set, which will be used for subsequent fault feature analysis. The first fault data set is deeply analyzed to identify feature information related to the fault, including voltage changes, current anomalies, temperature changes, etc. when the fault occurs. In the fault feature analysis, focus on those feature information that appear frequently. A predetermined frequency threshold is set, and the feature information that appears more frequently than this threshold is included in the first monitoring feature set, that is, multiple monitoring features that appear more frequently when a power distribution fault occurs, such as: voltage anomaly, current anomaly, frequency anomaly, harmonics, power distribution equipment temperature anomaly, etc. At the same time, the fault trigger frequency corresponding to these feature information is recorded, that is, the number of times they cause the fault to occur. The first monitoring feature set and the first fault trigger frequency are combined to form the first monitoring feature information for subsequent fault warning and fault location. The obtained first monitoring feature information is added to the multiple monitoring feature information that have been determined previously, and the set of the multiple monitoring feature information is continuously updated and improved so that it can more comprehensively reflect the fault characteristics of the power distribution system. Through the above method, a complete monitoring feature information library is established to support fault warning, fault location and fault recovery decisions of the power distribution system.
[0025] Based on the plurality of monitoring feature sets of the plurality of monitoring feature information, configure sensor types to construct a plurality of sensor monitoring combinations;
[0026] Performing a monitoring sensitivity analysis based on multiple fault triggering frequencies of the multiple monitoring feature information to determine multiple monitoring sensitivities;
[0027] Acquire a rated characteristic threshold of a monitoring characteristic, optimize and correct the rated characteristic threshold according to the multiple monitoring sensitivities, and determine multiple characteristic monitoring thresholds;
[0028] A multi-source sensor monitoring array of the plurality of monitoring nodes is established based on the plurality of sensor monitoring combinations and the plurality of feature monitoring thresholds.
[0029] For each monitoring feature set, select the appropriate sensor type according to its characteristics. For example, for voltage and current monitoring, voltage sensors and current sensors are required; for temperature monitoring, temperature sensors may be required. Combine sensors according to monitoring requirements to form multiple sensor monitoring combinations. Each combination may include one or more sensors for monitoring a specific feature set. Based on multiple fault triggering frequencies of multiple monitoring feature information, analyze the sensitivity of each monitoring feature to the occurrence of faults. The higher the fault triggering frequency, the more sensitive the feature is to the fault, that is, the higher the importance of the feature to the fault, and determine multiple monitoring sensitivities. Obtain the rated feature threshold of the monitoring feature. Each monitoring feature type corresponds to a rated feature threshold, which is provided by the equipment manufacturer or set during system design, such as the rated voltage threshold; optimize and correct to improve monitoring accuracy. The higher the fault frequency, the smaller the threshold range; optimize and correct the rated feature threshold according to the monitoring sensitivity analysis results. For features with high sensitivity, it may be necessary to set a stricter threshold; for features with low sensitivity, the threshold can be relaxed. The purpose of optimization and correction is to enable the monitoring array to more accurately identify potential faults and reduce false alarms and missed alarms. Install the configured multiple sensor monitoring combinations on the corresponding monitoring nodes to ensure that each node can monitor the equipment or lines it covers in real time. Set the optimized and corrected characteristic monitoring thresholds on each monitoring node to ensure that when the monitoring data exceeds or falls below these thresholds, an alarm can be triggered or corresponding control measures can be taken. Connect all monitoring nodes through the network to form a multi-source sensor monitoring array, which can collect and process the monitoring data from each node in real time, and provide strong support for fault warning, fault location and fault recovery decisions of the distribution system. According to the above method, an efficient and accurate multi-source sensor monitoring array is constructed to improve the operating safety and reliability of the distribution system.
[0030] Using the multi-source sensor monitoring array to monitor the target power distribution system in real time, extracting abnormal feature data sets of multiple abnormal monitoring nodes that do not meet the feature monitoring threshold;
[0031] Activate and start the multi-source sensor monitoring array, ensure that all monitoring nodes and their sensors are in normal working state, configure the parameters of real-time monitoring, including data sampling frequency, data transmission protocol, data storage method, etc., ensure that the data from the monitoring nodes can be collected and processed in real time and accurately, start the real-time monitoring process, and the multi-source sensor monitoring array begins to continuously collect data from each monitoring node in the target power distribution system, including electrical parameters such as voltage, current, temperature, power factor, and the operating status and fault information of the equipment, and compares them with the characteristic monitoring threshold. For each monitoring node, check whether its monitoring data meets the requirements of the characteristic monitoring threshold. If the data of a monitoring node does not meet the characteristic monitoring threshold, it is marked as an abnormal monitoring node. For each abnormal monitoring node, extract its abnormal feature data set that does not meet the characteristic monitoring threshold, and the abnormal feature data set contains all relevant monitoring data of the node when the abnormality occurs, as well as the time and location of the abnormality. Through the above method, the target power distribution system is monitored in real time using the multi-source sensor monitoring array, and the abnormal feature data set of the abnormal monitoring node that does not meet the characteristic monitoring threshold is accurately extracted.
[0032] configuring multiple simulation operation windows of the multiple abnormal monitoring nodes, performing power distribution simulation in combination with the target twin topology network and the abnormal feature data set, and obtaining a simulation feature data set;
[0033] Configure corresponding simulation operation windows for the multiple abnormal monitoring nodes to display and monitor the data and status during the simulation operation process. The simulation operation process is based on the distance between the monitoring node and the maintenance node. There are multiple maintenance nodes in the distribution network. The nearest node is selected. The longer the distance, the larger the window, that is, the longer the maintenance time required; predict the monitoring feature trend changes within the simulation operation time window. Integrate the twin topology network of the target distribution system into the simulation environment. The twin topology network is a digital representation of the structure and layout of the actual distribution system, including key elements such as equipment, lines, and substations. A digital twin model of the target distribution topology network is built based on digital twin technology; through the twin model, simulate trend analysis is performed based on the current abnormal data to determine the abnormal simulation characteristics. Load the previously extracted abnormal feature data set into the simulation operation window. The abnormal feature data set contains the monitoring data of the abnormal monitoring node in a specific time period, such as abnormal values such as voltage, current, and temperature. According to the operating parameters and rules of the actual distribution system, configure the relevant parameters of the simulation operation, including voltage range, current limit, equipment failure rate, etc. In the configured simulation environment, start the distribution simulation process. The simulation will simulate the operation of the actual distribution system based on the target twin topology network and the loaded abnormal feature data set. Through the simulation operation window, the data and status of the simulation process are monitored in real time. Pay attention to the performance of abnormal monitoring nodes and their impact on the entire distribution system. After the simulation process is completed, collect and record the data generated during the simulation process to form a simulation feature data set, which contains key information such as the data of each monitoring node during the simulation, system status, and fault occurrence. Through the above method, the abnormal feature data set and the target twin topology network are used to perform distribution simulation, so as to obtain a more accurate simulation feature data set, which provides strong support for the optimization and improvement of the distribution system.
[0034] Inputting the simulated feature data set into a predetermined fault prediction channel for fuzzy analysis, performing fault prediction based on the fuzzy analysis result, and outputting a plurality of fault prediction results, wherein the fault prediction results include fault location and fault type;
[0035] A predetermined fault channel is constructed. The predetermined fault prediction channel includes a fuzzy analysis subchannel and a fault prediction subchannel. The predetermined fault channel is a software system that integrates a fuzzy analysis algorithm and a fault prediction model, and is used for fuzzy analysis and fault prediction. The simulated feature data set is input into the predetermined fault prediction channel for fuzzy analysis. Fuzzy analysis is a mathematical method for processing imprecise and fuzzy information, allowing reasonable decisions to be made in an uncertain environment. Fuzzy logic can process uncertainty and fuzzy information, and improve the robustness and accuracy of fault location. In the distribution network, due to the influence of various factors such as the environment and equipment, there may be a certain degree of uncertainty in the fault information. Fuzzy logic can handle this situation well. In fuzzy analysis, a series of fuzzy sets and fuzzy rules can be defined to describe the relationship between the feature data set and the fault. Through fuzzy reasoning, the possibility of each fault prediction result can be obtained, and the fuzzy analysis result can be obtained; based on the result of the fuzzy analysis, fault prediction is performed, and the possible fault type and the specific location of the fault are predicted according to the input simulated feature data set and the fuzzy analysis result. The predetermined fault prediction channel will output multiple fault prediction results, each of which contains fault type and fault location information. Through the above method, corresponding preventive measures or emergency plans are formulated according to the output fault prediction results to reduce the impact of faults on the distribution system. At the same time, the prediction results can also provide valuable reference information for the optimization and improvement of the distribution system.
[0036] The predetermined fault prediction channel includes a fuzzy analysis subchannel and a fault prediction subchannel;
[0037] Inputting the simulation feature data set into the fuzzy analysis subchannel for fuzzy reasoning, and outputting a fuzzy predicted fault information set, wherein the fuzzy predicted fault information includes fuzzy fault location, fuzzy fault type and prediction confidence;
[0038] Extracting the abnormal monitoring nodes whose prediction confidence is greater than a predetermined confidence and setting them as fuzzy abnormal nodes, and acquiring multiple simulated abnormal feature data of the multiple fuzzy abnormal nodes;
[0039] The plurality of simulated abnormality feature data are input into the fault prediction subchannel for fault prediction, and the plurality of fault prediction results are output.
[0040] The predetermined fault prediction channel includes a fuzzy analysis subchannel and a fault prediction subchannel, wherein the fuzzy analysis subchannel processes the simulated feature data set by using a predetermined fuzzy reasoning algorithm. During the fuzzy reasoning process, the fuzzy fault location, fuzzy fault type and prediction confidence are calculated according to the fuzzy rule set and the simulated feature data set. The fuzzy analysis subchannel outputs a fuzzy prediction fault information set, which contains fuzzy fault location, fuzzy fault type and corresponding prediction confidence, wherein fuzzy fault location refers to location information, fuzzy fault type refers to fault type information, and corresponding prediction confidence refers to the credibility of the fuzzy prediction fault; the predetermined confidence is used to improve the precision of the prediction node setting, save computing resources, and improve prediction efficiency and accuracy. Traverse the fuzzy prediction fault information set, extract abnormal monitoring nodes whose prediction confidence is greater than the predetermined confidence, set these nodes as fuzzy abnormal nodes, and obtain the corresponding simulated abnormal feature data for each fuzzy abnormal node. Input the simulated abnormal feature data of multiple fuzzy abnormal nodes into the fault prediction subchannel. In the fault prediction subchannel, the fault prediction subchannel includes a BP decision unit, a random forest decision unit, and a fusion connection unit, and uses a predetermined fault prediction model to process the simulated abnormal feature data. The fault prediction model outputs a specific fault prediction result based on the input data, including the fault type and fault location, wherein the fault type refers to the specific nature or cause of the problem in the device or system, and the fault location refers to the process of determining the specific location or component where the fault occurs in the device or system. Through the above method, the fuzzy analysis subchannel and the fault prediction subchannel in the predetermined fault prediction channel are used to perform fuzzy reasoning and fault prediction on the simulated feature data set, thereby obtaining multiple fault prediction results including fault type and fault location.
[0041] Defining a power distribution fuzzy data set and a fuzzy membership function, and building a fuzzy processing window based on the power distribution fuzzy data set and the fuzzy membership function;
[0042] Retrieving and acquiring a historical power distribution fault data set, and dividing the historical power distribution fault data set into a first data set and a second data set according to a preset data division ratio;
[0043] Performing feature analysis based on the first data set to determine fuzzy reasoning features and fuzzy reasoning rules, and establishing a fuzzy reasoning system according to the fuzzy reasoning features and fuzzy reasoning rules;
[0044] Performing supervised training on the fuzzy inference system through the second data set to obtain a convergent fuzzy inference system that meets preset training constraints, wherein the convergent fuzzy inference system has a convergence degree identifier;
[0045] The fuzzy analysis subchannel is generated based on the fuzzy processing window and the convergent fuzzy inference system.
[0046] The distribution fuzzy data set contains fuzzy descriptions of various operating states of the distribution system, such as the fluctuation range of voltage, the fluctuation range of current, the change trend of temperature, etc. The fuzzy data is usually expressed in the form of linguistic variables or fuzzy sets. The fuzzy membership function is a function that maps a specific value to the degree of a fuzzy set. For example, for temperature, a membership function is defined to indicate the degree to which a specific temperature value belongs to the fuzzy set of "high". This function is usually continuous, and its output value is between 0 (completely not belonging) and 1 (completely belonging). Using the defined distribution fuzzy data set and fuzzy membership function, a fuzzy processing window is built, which is responsible for converting the specific input data, such as specific values of voltage, current, temperature, etc., into fuzzy data, and applying fuzzy reasoning rules for fault prediction. The distribution fault data set is retrieved from the historical database, which contains the faults that occurred in the past and their related operating data. According to the preset data division ratio, such as 70% of the first data set and 30% of the second data set, the historical distribution fault data set is divided into a first data set for feature analysis and the construction of the fuzzy reasoning system and a second data set for supervised training of the fuzzy reasoning system. Perform feature analysis on the first data set to determine which features, such as voltage, current, temperature, etc., are closely related to the occurrence of the fault, and determine the fuzzy reasoning features and rules of the features. According to the fuzzy reasoning features and rules obtained by the analysis, a fuzzy reasoning system is established. This system usually includes fuzzification, that is, converting specific data into fuzzy data; fuzzy reasoning, that is, applying fuzzy rules for reasoning and defuzzification, and converting fuzzy results into specific predictions. Input the second data set into the fuzzy reasoning system, and supervise the fuzzy reasoning system by comparing the prediction results of the fuzzy reasoning system with the actual fault situation. During the training process, the parameters of the fuzzy reasoning system, such as the shape of the membership function and the weight of the fuzzy rules, are continuously adjusted to make the prediction results of the fuzzy reasoning system more accurate. When the prediction performance of the fuzzy reasoning system reaches the preset training constraints, the training is stopped to obtain a converged fuzzy reasoning system, which has a convergence mark to indicate the degree of completion of its training and the accuracy of the prediction. Based on the constructed fuzzy processing window and the converged fuzzy reasoning system, a fuzzy analysis subchannel is generated. This sub-channel will be responsible for receiving the input distribution operation data, performing fuzzy processing and analysis, and outputting fault prediction results. Through the above method, a distribution fault prediction sub-channel based on fuzzy analysis is constructed to provide strong support for fault prediction of distribution systems.
[0047] Acquire a sample power distribution fault data set, and equally divide the sample power distribution fault data set into a first sample data set and a second sample data set;
[0048] Using the first sample data set and the second sample data set, respectively, the BP decision unit and the random forest decision unit are trained to a convergence state, and the convergence degree is marked, wherein the BP decision unit is constructed based on a BP neural network, and the random forest decision unit is constructed based on a random forest decision principle;
[0049] Configure a predetermined fusion strategy based on the convergence indicator, and embed the predetermined fusion strategy in the fusion connection unit;
[0050] The fault prediction subchannel is constructed based on the BP decision unit, the random forest decision unit and the fusion connection unit.
[0051] Obtain a sample power distribution fault data set from an actual power distribution system or a historical database, including characteristic data of the power distribution system under different operating conditions, such as voltage, current, temperature, etc., and corresponding fault labels, such as fault type, fault location, etc. The obtained sample power distribution fault data set is equally divided, that is, ensuring that the first sample data set and the second sample data set contain the same number of samples, and the data distribution is as similar as possible, through random sampling or other sampling methods. Divide the data set into two subsets, the first sample data set is used to train the BP decision unit, and the second sample data set is used to train the random forest decision unit. Construct a BP decision unit based on a BP neural network, use the first sample data set to train the BP decision unit, adjust the weight and bias of the BP decision unit, so that the BP decision unit can accurately predict faults, and stop training when the performance of the network on the validation set reaches the preset convergence standard, and mark the convergence degree. A random forest decision unit is constructed based on the random forest decision principle, and the random forest decision unit is trained using the second sample data set. The performance of the random forest decision unit is optimized by adjusting the number of trees, feature selection and other parameters. When the performance of the random forest decision unit on the validation set reaches the preset convergence standard, the training is stopped and the convergence is marked. According to the convergence marks of the BP decision unit and the random forest decision unit, a predetermined fusion strategy is configured based on a variety of methods, such as weighted average, voting method, stack method, etc. For example, different weights can be assigned according to the performance of the two decision units on the validation set, and then their prediction results are weighted and fused. In this application, the weight is configured based on the convergence degree. The greater the convergence degree, the greater the weight. Then, a weighted calculation is performed to obtain a predetermined fusion strategy. The predetermined fusion strategy is embedded in the fusion connection unit for use in the fault prediction subchannel. Based on the trained BP decision unit, the random forest decision unit and the fusion connection unit, a fault prediction subchannel is built. The fault prediction subchannel receives the input distribution system feature data, predicts through the BP decision unit and the random forest decision unit respectively, and then passes the prediction results to the fusion connection unit for fusion processing. The fusion connection unit fuses the prediction results of the two decision units according to the predetermined fusion strategy, and outputs the final fault prediction results, including information such as fault type and fault location, to provide valuable reference for fault prevention and maintenance of the distribution system. Through the above method, a fusion fault prediction subchannel based on BP decision unit and random forest decision unit is constructed to improve the accuracy and reliability of fault prediction.
[0052] Analyze the fault location and fault type through a predetermined recovery decision channel to determine a matching maintenance plan and optimize the maintenance route;
[0053] The fault location and fault type information output by the fault prediction subchannel is connected to the predetermined recovery decision channel, which is used to obtain the maintenance plan for the target power grid. The received fault location information is analyzed in detail to clarify the specific location of the fault, the scope of impact and the possible consequences. The fault type is analyzed in depth to understand the nature, cause and possible development trend of the fault. According to the fault analysis and historical experience, the maintenance plan matching the current fault is searched in the preset maintenance plan library. If there is no ready-made matching plan, a decision support system or expert system is used to generate a customized maintenance plan based on the fault location, fault type, equipment status and other information. Through the above method, it is ensured that the matching maintenance plan is determined quickly and accurately for specific fault location and fault type, and the maintenance efficiency is improved by optimizing the maintenance route, so as to effectively restore the normal operation of the equipment.
[0054] Retrieving and acquiring a plurality of sample fault repair data, and performing repair quality evaluation on the plurality of sample fault repair data based on a predetermined repair evaluation strategy to obtain a plurality of repair evaluation coefficients;
[0055] Extracting sample fault maintenance data corresponding to maintenance evaluation coefficients greater than the evaluation coefficient threshold, and obtaining a plurality of standard fault maintenance data;
[0056] Performing cluster analysis on the plurality of standard fault repair data, and constructing a repair scheme matching unit according to the plurality of clustering results based on the decision tree principle;
[0057] Acquire multiple maintenance nodes of the target power distribution system, and build a maintenance line analysis unit based on the multiple maintenance nodes and the GIS positioning system;
[0058] The predetermined restoration decision channel is established according to the maintenance scheme matching unit and the maintenance line analysis unit.
[0059] Retrieve multiple sample fault maintenance data from the historical database, including complete fault description, maintenance process, maintenance results and other information, and perform maintenance quality evaluation on the sample fault maintenance data based on a predetermined maintenance evaluation strategy, wherein the predetermined maintenance evaluation strategy includes maintenance time, resource consumption, re-failure rate, etc.; calculate the maintenance evaluation coefficient of the multiple sample fault maintenance data, wherein the maintenance evaluation coefficient can quantitatively represent the quality of the maintenance work, set an evaluation coefficient threshold, extract the sample fault maintenance data corresponding to the maintenance evaluation coefficient greater than the threshold, and the corresponding sample fault maintenance data can be regarded as high-quality maintenance cases, i.e., multiple standard fault maintenance data. Perform cluster analysis on the multiple standard fault maintenance data, group the multiple standard fault maintenance data according to the similarity of the data, each cluster represents a similar fault and maintenance mode, and construct a maintenance plan matching unit based on the decision tree principle using the clustering results. This unit can quickly match the corresponding maintenance plan according to the input fault information. The information of multiple maintenance nodes of the target distribution system is obtained, including the geographical location, equipment type, maintenance history, etc. of the nodes. The maintenance nodes are visualized on the map using GIS, i.e., geographic information system positioning technology, and a maintenance line analysis unit is constructed. This unit can analyze the optimal maintenance route based on the GIS system according to the maintenance requirements, including the shortest path, shortest time, etc. Combined with the maintenance plan matching unit and the maintenance line analysis unit, a scheduled recovery decision channel is constructed. When a fault occurs in the target distribution system, the decision channel can automatically receive the fault information, quickly find a matching maintenance plan through the maintenance plan matching unit, and plan the optimal maintenance route using the maintenance line analysis unit. According to the above method, by constructing a scheduled recovery decision channel, the maintenance quality, decision efficiency, resource allocation, maintenance cost, system reliability, and intelligent operation and maintenance can be effectively improved, and strong guarantees can be provided for the safe and stable operation of the distribution system.
[0060] A fault recovery plan is generated based on the fault location, fault type, matching maintenance plan and optimized maintenance route, and the fault recovery plan is sent to the operation and maintenance personnel of the associated maintenance node.
[0061] Once the fault is located and classified, the corresponding maintenance plan can be quickly matched, and a fault recovery plan can be generated based on the optimized maintenance route. The time for fault response and recovery is shortened, and the availability of the system is improved. The fault recovery plan is generated based on accurate fault location and type analysis, as well as verified maintenance plans and optimized routes. Therefore, the plan has high accuracy and reliability, and can ensure that the fault is handled correctly and effectively. The operation and maintenance personnel can directly receive the fault recovery plan related to the associated maintenance node without additional query or analysis. This greatly reduces the workload and time cost of the operation and maintenance personnel and improves the operation and maintenance efficiency. Through the above method, a fault recovery plan is generated based on fault location, fault type, matching maintenance plan, and optimized maintenance route, and sent to the operation and maintenance personnel of the associated maintenance node, which can significantly improve the efficiency and quality of fault handling, reduce operation and maintenance costs, and enhance the stability and reliability of the system.
[0062] like Figure 3 As shown, the embodiment of the present application includes a fault location and restoration decision system for a distribution network, the system comprising:
[0063] A target power distribution topology network determination module 11, the target power distribution topology network determination module 11 is used to identify the target power distribution system and determine the target power distribution topology network, wherein the target power distribution topology network is embedded with multiple monitoring nodes, and each monitoring node has a mapping associated coverage device and / or coverage line;
[0064] A fault feature analysis module 12, the fault feature analysis module 12 is used to perform fault feature analysis based on the coverage device and / or coverage line, determine multiple monitoring feature information, and build a multi-source sensor monitoring array according to the multiple monitoring feature information and the multiple monitoring nodes, wherein the multi-source sensor monitoring array is embedded with a feature monitoring threshold;
[0065] A real-time monitoring module 13, wherein the real-time monitoring module 13 is used to use the multi-source sensor monitoring array to perform real-time monitoring on the target power distribution system, and extract abnormal feature data sets of multiple abnormal monitoring nodes that do not meet the feature monitoring threshold;
[0066] A simulation operation window configuration module 14, wherein the simulation operation window configuration module 14 is used to configure multiple simulation operation windows of the multiple abnormal monitoring nodes, perform power distribution simulation in combination with the target twin topology network and the abnormal feature data set, and obtain a simulation feature data set;
[0067] A fuzzy analysis module 15, wherein the fuzzy analysis module 15 is used to input the simulation feature data set into a predetermined fault prediction channel for fuzzy analysis, perform fault prediction based on the fuzzy analysis result, and output a plurality of fault prediction results, wherein the fault prediction results include fault location and fault type;
[0068] A fault location and type analysis module 16, which is used to analyze the fault location and fault type through a predetermined recovery decision channel, determine a matching maintenance plan and optimize the maintenance route;
[0069] A fault recovery plan generation module 17 is used to generate a fault recovery plan based on the fault location, fault type, matching maintenance plan and optimized maintenance route, and send the fault recovery plan to the operation and maintenance personnel of the associated maintenance node.
[0070] Furthermore, the embodiment of the present application also includes:
[0071] A monitoring node random selection module, wherein the monitoring node random selection module is used to randomly select a first monitoring node and obtain a first coverage device and / or a first coverage line of the first monitoring node;
[0072] A historical power distribution record query module, the historical power distribution record query module is used to query the historical power distribution record and obtain a first fault data set based on the first coverage device and / or the first coverage line as constraints;
[0073] A monitoring feature set determination module, the monitoring feature set determination module is used to perform fault feature analysis based on the first fault data set to determine a first monitoring feature set and a first fault triggering frequency that meet a predetermined frequency threshold;
[0074] A monitoring feature information adding module is used to obtain first monitoring feature information according to the first monitoring feature set and the first fault triggering frequency, and add the first monitoring feature information to the multiple monitoring feature information.
[0075] Furthermore, the embodiment of the present application also includes:
[0076] A sensor monitoring combination building module, the sensor monitoring combination building module is used to configure sensor types based on multiple monitoring feature sets of the multiple monitoring feature information to build multiple sensor monitoring combinations;
[0077] A monitoring sensitivity determination module, the monitoring sensitivity determination module is used to perform monitoring sensitivity analysis based on multiple fault triggering frequencies of the multiple monitoring feature information to determine multiple monitoring sensitivities;
[0078] An optimization and correction module, the optimization and correction module is used to obtain a rated feature threshold of a monitoring feature, optimize and correct the rated feature threshold according to the multiple monitoring sensitivities, and determine multiple feature monitoring thresholds;
[0079] A multi-source sensor monitoring array building module is used to build a multi-source sensor monitoring array of the multiple monitoring nodes based on the multiple sensor monitoring combinations and the multiple feature monitoring thresholds.
[0080] Furthermore, the embodiment of the present application also includes:
[0081] The predetermined fault prediction channel includes a module, and the predetermined fault prediction channel includes a module for the predetermined fault prediction channel to include a fuzzy analysis subchannel and a fault prediction subchannel;
[0082] A fuzzy prediction fault information set output module, the fuzzy prediction fault information set output module is used to input the simulation feature data set into the fuzzy analysis subchannel for fuzzy reasoning, and output a fuzzy prediction fault information set, wherein the fuzzy prediction fault information includes fuzzy fault location, fuzzy fault type and prediction confidence;
[0083] A simulated abnormal feature data module, wherein the simulated abnormal feature data module is used to extract the abnormal monitoring nodes whose prediction confidence is greater than a predetermined confidence and set them as fuzzy abnormal nodes, and obtain multiple simulated abnormal feature data of multiple fuzzy abnormal nodes;
[0084] A fault result output module is used to input the multiple simulated abnormality feature data into the fault prediction sub-channel to perform fault prediction and output the multiple fault prediction results.
[0085] Furthermore, the embodiment of the present application also includes:
[0086] A fuzzy processing window building module, wherein the fuzzy processing window building module is used to define a power distribution fuzzy data set and a fuzzy membership function, and to build a fuzzy processing window based on the power distribution fuzzy data set and the fuzzy membership function;
[0087] A historical power distribution fault data set acquisition module, the historical power distribution fault data set acquisition module is used to retrieve and acquire a historical power distribution fault data set, and divide the historical power distribution fault data set into a first data set and a second data set according to a preset data division ratio;
[0088] A feature analysis module, the feature analysis module is used to perform feature analysis based on the first data set, determine fuzzy reasoning features and fuzzy reasoning rules, and establish a fuzzy reasoning system according to the fuzzy reasoning features and fuzzy reasoning rules;
[0089] A supervised training module, wherein the supervised training module is used to perform supervised training on the fuzzy inference system through the second data set to obtain a convergent fuzzy inference system that meets preset training constraints, wherein the convergent fuzzy inference system has a convergence degree identifier;
[0090] A fuzzy analysis sub-channel generation module is used to generate the fuzzy analysis sub-channel based on the fuzzy processing window and the convergent fuzzy inference system.
[0091] Furthermore, the embodiment of the present application also includes:
[0092] A sample power distribution fault data set acquisition module, the sample power distribution fault data set acquisition module is used to acquire a sample power distribution fault data set, and divide the sample power distribution fault data set into a first sample data set and a second sample data set by equal value;
[0093] A convergence identification module, the convergence identification module is used to train the BP decision unit and the random forest decision unit to a convergence state respectively through the first sample data set and the second sample data set, and perform convergence identification, wherein the BP decision unit is constructed based on the BP neural network, and the random forest decision unit is constructed based on the random forest decision principle;
[0094] A fusion connection unit module, the fusion connection unit module is used to configure a predetermined fusion strategy based on the convergence degree indicator, and embed the predetermined fusion strategy into the fusion connection unit;
[0095] A fault prediction sub-channel building module is used to build the fault prediction sub-channel based on the BP decision unit, the random forest decision unit and the fusion connection unit.
[0096] Furthermore, the embodiment of the present application also includes:
[0097] A maintenance evaluation coefficient acquisition module, the maintenance evaluation coefficient acquisition module is used to retrieve and acquire multiple sample fault maintenance data, perform maintenance quality evaluation on the multiple sample fault maintenance data based on a predetermined maintenance evaluation strategy, and obtain multiple maintenance evaluation coefficients;
[0098] A standard fault repair data acquisition module, the standard fault repair data acquisition module is used to extract sample fault repair data corresponding to a repair evaluation coefficient greater than an evaluation coefficient threshold, and obtain a plurality of standard fault repair data;
[0099] A maintenance scheme matching unit module, wherein the maintenance scheme matching unit module is used to perform cluster analysis on the plurality of standard fault maintenance data, and construct a maintenance scheme matching unit according to the plurality of clustering results based on the decision tree principle;
[0100] A maintenance node acquisition module, the maintenance node acquisition module is used to acquire multiple maintenance nodes of the target power distribution system, and build a maintenance line analysis unit based on the multiple maintenance nodes and the GIS positioning system;
[0101] A predetermined restoration decision channel module is used to build the predetermined restoration decision channel according to the maintenance scheme matching unit and the maintenance line analysis unit.
[0102] For the specific embodiment of the fault location and restoration decision system for the distribution network, please refer to the embodiment of the fault location and restoration decision method for the distribution network described above, which will not be repeated here. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
[0103] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0104] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application.
Claims
1. Fault location and restoration decision method for distribution network, It is characterized in that The method comprises: Identify a target power distribution system and determine a target power distribution topology network, wherein the target power distribution topology network has multiple monitoring nodes embedded therein, and each monitoring node has a mapping-associated coverage device or / and coverage line; Perform fault feature analysis based on the covered device and / or covered line, determine multiple monitoring feature information, and build a multi-source sensor monitoring array according to the multiple monitoring feature information and the multiple monitoring nodes, wherein the multi-source sensor monitoring array is embedded with a feature monitoring threshold; Using the multi-source sensor monitoring array to monitor the target power distribution system in real time, extracting abnormal feature data sets of multiple abnormal monitoring nodes that do not meet the feature monitoring threshold; configuring multiple simulation operation windows of the multiple abnormal monitoring nodes, performing power distribution simulation in combination with the target twin topology network and the abnormal feature data set, and obtaining a simulation feature data set; Inputting the simulated feature data set into a predetermined fault prediction channel for fuzzy analysis, performing fault prediction based on the fuzzy analysis result, and outputting a plurality of fault prediction results, wherein the fault prediction results include fault location and fault type; Analyze the fault location and fault type through a predetermined recovery decision channel to determine a matching maintenance plan and optimize the maintenance route; A fault recovery plan is generated based on the fault location, fault type, matching maintenance plan and optimized maintenance route, and the fault recovery plan is sent to the operation and maintenance personnel of the associated maintenance node.
2. The method according to claim 1, It is characterized in that Based on the covered device and / or the covered line, a fault feature analysis is performed to determine a plurality of monitoring feature information, including: Randomly select a first monitoring node, and obtain a first coverage device and / or a first coverage line of the first monitoring node; Taking the first covered device and / or the first covered line as constraints, querying historical power distribution records to obtain a first fault data set; Performing fault feature analysis based on the first fault data set to determine a first monitoring feature set and a first fault triggering frequency that meet a predetermined frequency threshold; First monitoring feature information is obtained according to the first monitoring feature set and the first fault triggering frequency, and is added to the plurality of monitoring feature information.
3. The method according to claim 2, It is characterized in that Building a multi-source sensor monitoring array according to the plurality of monitoring feature information and the plurality of monitoring nodes includes: Based on the plurality of monitoring feature sets of the plurality of monitoring feature information, configure sensor types to construct a plurality of sensor monitoring combinations; Performing a monitoring sensitivity analysis based on multiple fault triggering frequencies of the multiple monitoring feature information to determine multiple monitoring sensitivities; Acquire a rated characteristic threshold of a monitoring characteristic, optimize and correct the rated characteristic threshold according to the multiple monitoring sensitivities, and determine multiple characteristic monitoring thresholds; A multi-source sensor monitoring array of the plurality of monitoring nodes is established based on the plurality of sensor monitoring combinations and the plurality of feature monitoring thresholds.
4. The method according to claim 1, It is characterized in that The simulated feature data set is input into a predetermined fault prediction channel for fuzzy analysis, fault prediction is performed based on the fuzzy analysis result, and multiple fault prediction results are output, including: The predetermined fault prediction channel includes a fuzzy analysis subchannel and a fault prediction subchannel; Inputting the simulation feature data set into the fuzzy analysis subchannel for fuzzy reasoning, and outputting a fuzzy predicted fault information set, wherein the fuzzy predicted fault information includes fuzzy fault location, fuzzy fault type and prediction confidence; Extracting the abnormal monitoring nodes whose prediction confidence is greater than a predetermined confidence and setting them as fuzzy abnormal nodes, and obtaining a plurality of simulated abnormal feature data of the plurality of fuzzy abnormal nodes; The multiple simulated abnormality feature data are input into the fault prediction sub-channel to perform fault prediction, and the multiple fault prediction results are output.
5. The method according to claim 4, It is characterized in that The method further comprises: Defining a power distribution fuzzy data set and a fuzzy membership function, and building a fuzzy processing window based on the power distribution fuzzy data set and the fuzzy membership function; Retrieving and acquiring a historical power distribution fault data set, and dividing the historical power distribution fault data set into a first data set and a second data set according to a preset data division ratio; Performing feature analysis based on the first data set to determine fuzzy reasoning features and fuzzy reasoning rules, and establishing a fuzzy reasoning system according to the fuzzy reasoning features and fuzzy reasoning rules; Performing supervised training on the fuzzy inference system through the second data set to obtain a convergent fuzzy inference system that meets preset training constraints, wherein the convergent fuzzy inference system has a convergence degree identifier; The fuzzy analysis subchannel is generated based on the fuzzy processing window and the convergent fuzzy inference system.
6. The method according to claim 4, It is characterized in that Inputting the plurality of simulated abnormal feature data into the fault prediction subchannel to perform fault prediction includes: Acquire a sample power distribution fault data set, and equally divide the sample power distribution fault data set into a first sample data set and a second sample data set; Using the first sample data set and the second sample data set, respectively, the BP decision unit and the random forest decision unit are trained to a convergence state, and the convergence degree is marked, wherein the BP decision unit is constructed based on a BP neural network, and the random forest decision unit is constructed based on a random forest decision principle; Configure a predetermined fusion strategy based on the convergence indicator, and embed the predetermined fusion strategy in the fusion connection unit; The fault prediction subchannel is constructed based on the BP decision unit, the random forest decision unit and the fusion connection unit.
7. The method according to claim 1, It is characterized in that The fault location and fault type are analyzed through a predetermined recovery decision channel, including: Retrieving and acquiring a plurality of sample fault repair data, and performing repair quality evaluation on the plurality of sample fault repair data based on a predetermined repair evaluation strategy to obtain a plurality of repair evaluation coefficients; Extracting sample fault maintenance data corresponding to maintenance evaluation coefficients greater than the evaluation coefficient threshold, and obtaining a plurality of standard fault maintenance data; Performing cluster analysis on the plurality of standard fault repair data, and constructing a repair scheme matching unit according to the plurality of clustering results based on the decision tree principle; Acquire multiple maintenance nodes of the target power distribution system, and build a maintenance line analysis unit based on the multiple maintenance nodes and the GIS positioning system; The predetermined restoration decision channel is established according to the maintenance scheme matching unit and the maintenance line analysis unit.
8. Fault location and restoration decision system for distribution network, It is characterized in that The system comprises: A target power distribution topology network determination module, the target power distribution topology network determination module is used to identify the target power distribution system and determine the target power distribution topology network, wherein the target power distribution topology network is embedded with multiple monitoring nodes, and each monitoring node has a mapping-associated coverage device and / or coverage line; A fault feature analysis module, the fault feature analysis module is used to perform fault feature analysis based on the coverage device and / or coverage line, determine multiple monitoring feature information, and build a multi-source sensor monitoring array according to the multiple monitoring feature information and the multiple monitoring nodes, wherein the multi-source sensor monitoring array is embedded with a feature monitoring threshold; A real-time monitoring module, the real-time monitoring module is used to use the multi-source sensor monitoring array to perform real-time monitoring on the target power distribution system, and extract abnormal feature data sets of multiple abnormal monitoring nodes that do not meet the feature monitoring threshold; A simulation operation window configuration module, wherein the simulation operation window configuration module is used to configure multiple simulation operation windows of the multiple abnormal monitoring nodes, perform power distribution simulation in combination with the target twin topology network and the abnormal feature data set, and obtain a simulation feature data set; A fuzzy analysis module, the fuzzy analysis module is used to input the simulation feature data set into a predetermined fault prediction channel for fuzzy analysis, perform fault prediction based on the fuzzy analysis result, and output multiple fault prediction results, wherein the fault prediction results include fault location and fault type; A fault location and type analysis module, which is used to analyze the fault location and fault type through a predetermined recovery decision channel, determine a matching maintenance plan and optimize the maintenance route; A fault recovery plan generation module is used to generate a fault recovery plan based on the fault location, fault type, matching maintenance plan and optimized maintenance route, and send the fault recovery plan to the operation and maintenance personnel of the associated maintenance node.
Citation Information
Cited By
Automatic fault detection method and system for power distribution network frame
CN119247045A
Intelligent monitoring system for operation state of distribution transformer
CN119341207A
Single Beidou RTK high-precision positioning method and system for monitoring working condition of power transmission line
CN119689177A
Single Beidou RTK high-precision positioning method and system for transmission line condition monitoring
CN119689177B
Fault identification and prediction method based on low-voltage distribution network topology modeling and intelligent algorithm
CN120539535A
