Power distribution network intelligent diagnosis and self-adaptive protection method and device based on multi-source data fusion

Abstract

The invention provides a power distribution network intelligent diagnosis and self-adaptive protection method and device based on multi-source data fusion, and is applied to the technical field of data processing. The method comprises the following steps: processing power distribution network operation data information, fault recording data information, protection device parameter information, equipment state monitoring information and environment parameter information based on technical document reference information to generate intelligent diagnosis parameter information; processing the protection action information based on the historical fault information to generate fault processing strategy parameter information; processing the protection operation process information of the power distribution network to generate an operation optimization factor; processing the real-time operation information and the version number information corresponding to the real-time operation information based on the operation optimization factor, and generating operation state classification information and processing priority information of the real-time operation information; and processing the data based on the target adaptive protection model to generate adaptive protection optimization result information of the power distribution network.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a distribution network intelligent diagnosis and adaptive protection method and device based on multi-source data fusion. Background Art

[0002] With the adjustment of energy structures and the growth of electricity demand, distribution networks are rapidly developing towards intelligent and efficient development. Distributed generation is increasingly widely used in distribution networks, but its power generation characteristics differ from those of traditional power plants. When integrated into the previously radial distribution network, this complicates the topology of the distribution network, resulting in variable power flow directions and difficulty in determining the direction of short-circuit currents. This makes traditional, simple relay protection solutions incapable of meeting protection requirements.

[0003] At the same time, the development of smart distribution networks has generated massive amounts of data. Problems such as data expansion and poor data quality pose challenges to fault diagnosis. While topology-based methods exist for fault location and isolation, accurate topology modeling is crucial. Existing technologies for distribution network protection and fault diagnosis have numerous shortcomings, making it difficult to ensure safe, reliable, and efficient operation under complex operating conditions. A technical solution is urgently needed that can integrate multi-source data, improve diagnostic accuracy, and enhance protection adaptability to meet the demands of intelligent distribution network development.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore includes information that does not constitute prior art known to ordinary technicians in this field. Summary of the Invention

[0005] The purpose of this application is to provide a method and device for intelligent diagnosis and adaptive protection of distribution networks based on multi-source data fusion, which, at least to a certain extent, overcomes the problems existing in the prior art. By acquiring multi-source information such as operation and fault recording, intelligent diagnosis parameters and fault handling strategy parameter information are generated based on technical document processing. At the same time, the protection operation process information is processed in combination with technical documents and version numbers to generate operation optimization factors, and real-time operation information is analyzed to obtain operation status classification and processing priority information. The relevant information is input into the model, an evaluation value is generated, and the model decision vector is optimized to obtain optimization results, thereby achieving precise protection and optimization and improving the intelligence level and operational stability of the distribution network.

[0006] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0007] According to one aspect of the present application, a distribution network intelligent diagnosis and adaptive protection method based on multi-source data fusion is provided, including: obtaining distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, environmental parameter information, historical fault information, protection action information and technical document reference information; processing the distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information and environmental parameter information based on the technical document reference information to generate intelligent diagnosis parameter information; processing the protection action information based on the historical fault information to generate fault handling strategy parameter information; processing the distribution network protection operation process information based on the technical document reference information and the version number information corresponding to the technical document reference information to generate an operation optimization factor; processing the real-time operation information and the version number information corresponding to the real-time operation information based on the operation optimization factor to generate operation status classification information and processing priority information of the real-time operation information; inputting the intelligent diagnosis index, the operation status classification information and processing priority information of the real-time operation information and the fault handling strategy parameter information into the target adaptive protection model for processing to generate distribution network adaptive protection optimization result information.

[0008] Another aspect of the present application is a distribution network intelligent diagnosis and adaptive protection device based on multi-source data fusion, characterized in that it includes: an acquisition module for acquiring distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, environmental parameter information, historical fault information, protection action information and technical document reference information; a processing module for processing the distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, and environmental parameter information based on the technical document reference information to generate intelligent diagnosis parameter information; processing the protection action information based on the historical fault information to generate fault handling strategy parameter information; processing the distribution network protection operation process information based on the technical document reference information and the version number information corresponding to the technical document reference information to generate an operation optimization factor; processing the real-time operation information and the version number information corresponding to the real-time operation information based on the operation optimization factor to generate operation status classification information and processing priority information of the real-time operation information; inputting the intelligent diagnosis index, the operation status classification information of the real-time operation information and the processing priority information and the fault handling strategy parameter information into the target adaptive protection model for processing to generate distribution network adaptive protection optimization result information.

[0009] According to another aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a second processor, the computer program implements the above-mentioned distribution network intelligent diagnosis and adaptive protection method based on multi-source data fusion.

[0010] The present application provides a method and device for intelligent diagnosis and adaptive protection of distribution networks based on multi-source data fusion. The server is designed to integrate multi-source data to improve the reliability and safety of distribution network operation. First, obtain multi-source information such as operation and fault recording, and generate intelligent diagnosis parameters and fault handling strategy parameter information based on technical document processing. At the same time, combine the technical documents and version numbers to process the protection operation process information, generate operation optimization factors, and analyze the real-time operation information to obtain the operation status classification and processing priority information. Then obtain the target adaptive protection model, and determine the final model through data sampling and model training. Finally, input the relevant information into the model, generate an evaluation value, optimize the model decision vector, and obtain the optimization result to achieve precise protection and optimization, and improve the intelligence level and operation stability of the distribution network.

[0011] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A flow chart of a method for intelligent diagnosis and adaptive protection of a distribution network based on multi-source data fusion provided by an embodiment of the present application is shown;

[0013] Figure 2 A schematic structural diagram of a distribution network intelligent diagnosis and adaptive protection device based on multi-source data fusion provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0014] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0015] The following combination Figure 1 To describe the distribution network intelligent diagnosis and adaptive protection method based on multi-source data fusion according to the exemplary embodiment of the present application. In one embodiment, the present application also proposes a distribution network intelligent diagnosis and adaptive protection method and device based on multi-source data fusion. Figure 1 As shown, the method is applied to the server and includes:

[0016] S101, obtaining distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, environmental parameter information, historical fault information, protection action information and technical document reference information.

[0017] In one implementation, distribution network operational data reflects key data on the real-time operating status of the distribution network, encompassing electrical parameters such as voltage, current, power, and frequency. Voltage information reflects the stability of the grid's power supply; abnormal voltage fluctuations can cause equipment damage or operational anomalies. Current data helps determine whether lines are overloaded or contain potential faults such as short circuits. Power and frequency parameters are equally important, impacting the grid's energy transmission efficiency and overall operational quality. Continuous monitoring and analysis of this operational data can promptly identify abnormal distribution network conditions, providing real-time evidence for intelligent diagnosis.

[0018] When a fault occurs in a distribution network, a fault recorder accurately records the changes in electrical quantities over a period of time before and after the fault, including voltage and current waveforms, as well as frequency variations. This detailed data, which records the entire fault process, is crucial for analyzing the fault type (such as short circuit, open circuit, or ground fault), determining the fault location, and determining the cause. For example, by analyzing the sudden change characteristics of the current waveform, the nature of the fault can be determined; combining the phase relationship between voltage and current allows for more accurate location of the fault point. Fault recorder data provides strong support for in-depth understanding of fault mechanisms and optimizing protection strategies. Protection devices are crucial for ensuring the safe operation of distribution networks. Their parameters include action thresholds, action times, and protection ranges. The action thresholds determine the electrical condition under which a protection device triggers its action, such as the current threshold for overcurrent protection and the voltage threshold for overvoltage protection. The action time affects the protection device's response speed, and fast and accurate action is crucial for minimizing fault losses. The protection range defines the area that the protection device can monitor and protect. Obtaining this parameter information helps evaluate the performance of protection devices, ensuring that they operate correctly and promptly when an anomaly occurs in the distribution network, effectively protecting the grid.

[0019] Equipment status monitoring collects operational data on various equipment in the distribution network, such as transformers, switchgear, and transmission lines. For transformers, monitoring of indicators such as oil temperature, winding temperature, and oil color spectrum can determine whether there are internal problems such as overheating or insulation aging. Switchgear monitoring includes switch position, mechanical condition, and partial discharge to assess operational reliability and insulation performance. For transmission lines, monitoring focuses on conductor temperature, sag, and icing, which reflect the physical condition and environmental impacts of the line. This equipment status monitoring information can be used to proactively identify potential equipment failures, enabling condition-based maintenance and improving the reliability and safety of the distribution network.

[0020] The operation of distribution networks is significantly affected by environmental factors. Environmental parameter information includes meteorological data such as temperature, humidity, wind speed, rainfall, and lightning activity, as well as geographic information and electromagnetic interference. High temperatures can hinder equipment heat dissipation, causing equipment temperatures to rise and impacting performance and lifespan. High humidity can easily cause moisture to build up on equipment insulation, reducing insulation performance. Lightning activity can cause overvoltages on lines, threatening equipment safety. Furthermore, geographic information (such as topography and altitude) can influence line layout and operation, and electromagnetic interference can disrupt the normal operation of protection devices and monitoring equipment. Obtaining environmental parameter information allows for comprehensive consideration of the impact of environmental factors on distribution network operation, improving the accuracy of fault diagnosis and the relevance of protection strategies.

[0021] Historical fault information contains detailed records of various past faults in the distribution network, including the time, location, type, cause, handling process, and impact. Through in-depth analysis of historical fault information, patterns and trends in fault occurrence can be identified. For example, certain areas or equipment are more prone to failures in specific seasons or operating conditions. These patterns help predict future faults and provide a reference for developing preventive measures. Furthermore, the lessons learned from historical fault handling can inform current fault handling, improving the efficiency and quality of fault handling.

[0022] Protection action information records the actions of protection devices during distribution network operation, including the time, type (e.g., tripping, alarming), and action targets. Protection action information reflects the protection device's response to distribution network faults or abnormal conditions. By analyzing protection action information, the correct and timely action of the protection device can be assessed. For example, if a protection device fails to operate promptly when a fault occurs, or if a misoperation causes an unnecessary power outage, adjustments to the protection device's parameters or logic may be necessary. Combining protection action information with historical fault information provides a more comprehensive understanding of the distribution network's fault protection status, providing a basis for optimizing protection strategies. Technical document reference information covers various technical documents, including distribution network design specifications, operating procedures, and protection device technical manuals. These documents serve as the standards and foundation for distribution network construction, operation, and maintenance, encompassing industry best practices and technical requirements. When processing and analyzing other types of information, technical document reference information provides an important reference standard, ensuring scientific and standardized processing. For example, when generating intelligent diagnostic parameter information, data features are extracted and quantitatively analyzed based on the standards and specifications in the technical documentation. When optimizing protection operation procedures, the operating guidelines and requirements in the technical documentation are referenced to ensure correct and safe operation. Technical documentation reference information is crucial for ensuring the effective operation of the entire distribution network intelligent diagnosis and adaptive protection system.

[0023] S102 , based on technical document reference information, distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, and environmental parameter information are processed to generate intelligent diagnosis parameter information.

[0024] In one embodiment, feature extraction is performed on distribution network operation data, fault recording data, protection device parameter information, equipment status monitoring information, and environmental parameter information to generate distribution network operation features, fault recording features, protection device parameter features, equipment status features, and environmental features. Features such as voltage amplitude, phase, current magnitude, and power factor are extracted. For example, the fluctuation range of voltage amplitude can reflect the stability of the power grid. If the voltage amplitude frequently exceeds the normal range during a period of time, it indicates a possible abnormality in the grid. Fault-induced voltage and current waveform features are extracted, such as the degree of waveform distortion and the point of sudden change. Frequency variation features are also extracted. Fault-induced frequency fluctuations, and the amplitude and duration of these fluctuations are important features. Features such as action thresholds (such as the current threshold for overcurrent protection and the voltage threshold for overvoltage protection), action time, and protection range are also extracted. The action threshold determines when the protection device activates, and different threshold settings can affect the protection effectiveness. For transformers, features such as oil temperature, winding temperature, and oil color spectrum are extracted; for switchgear, features such as switch position, mechanical state, and partial discharge are extracted; and for transmission lines, features such as conductor temperature, sag, and icing are extracted. For example, excessively high transformer oil temperature may indicate an internal overheating fault. Meteorological characteristics such as temperature, humidity, wind speed, rainfall, and lightning activity, as well as geographic information and electromagnetic interference, are extracted. High humidity can degrade equipment insulation performance and is a key environmental characteristic affecting distribution network operation.

[0025] Quantitative analysis is performed on distribution network operating characteristics and protection device parameter characteristics to generate a basic quantitative factor for the distribution network. Quantitative analysis is performed on distribution network operating characteristics such as voltage amplitude and current magnitude. For example, the average and standard deviation of voltage amplitude, and the peak and effective value of current over a period of time are calculated. For example, if the average voltage amplitude of a line is 95% of the rated value and the standard deviation is small, this indicates that the voltage is relatively stable. Quantitative processing is performed on protection device parameter characteristics such as the operating threshold and operating time. The operating threshold is compared with the rated value to calculate the degree of deviation; the operating time is accurately measured and recorded. If the overcurrent protection operating threshold is set at 1.5 times the rated current, but the actual measured operating current is 1.6 times the rated current, this indicates that the operating threshold setting is generally reasonable, but there is a certain margin. These quantified characteristics are then comprehensively calculated to generate a basic quantitative factor for the distribution network. Methods such as weighted averaging can be used to assign weights to different characteristics based on their importance. For example, voltage stability is given a higher weight. This calculation results in a quantitative factor value that represents the basic operating and protection performance of the distribution network.

[0026] Quantitative analysis and processing of fault waveform features, equipment status features, and environmental features is performed to generate diagnostic operation quantitative factors. The degree of distortion of the voltage and current waveforms during a fault is quantified, for example, by calculating the harmonic content of the waveforms. The point of a sudden change is located and quantified, recording the amplitude and timing of the sudden change. A significant increase in the harmonic content of the current waveform during a fault indicates a potentially serious fault. Quantification of equipment status features is used to characterize transformer oil temperature. A normal temperature range is set, the actual oil temperature is compared with the normal range, and the deviation is calculated. For switchgear partial discharge, a quantified classification is performed based on the discharge magnitude. If the transformer oil temperature exceeds the normal range by 20%, it indicates a significant equipment hazard. Quantification of environmental features is used to characterize meteorological parameters such as temperature and humidity by comparing them with the equipment's suitable operating range to calculate the degree of deviation. Electromagnetic interference is measured and classified. For example, if the ambient humidity exceeds the equipment's suitable humidity range by 30%, the risk of moisture damage to the equipment's insulation increases. These quantified features are combined to generate diagnostic operation quantitative factors. Similarly, weighted averaging and other methods can be used to assign weights based on the impact of different features on fault diagnosis, resulting in a quantitative factor value that reflects the relevant factors in fault diagnosis.

[0027] Based on the reference information in the technical documentation, the distribution network basic quantitative factors and diagnostic operation quantitative factors are processed to generate intelligent diagnostic assessment information and corresponding weight calculation results. The distribution network basic quantitative factors and diagnostic operation quantitative factors are compared with the standards and specifications in the technical documentation. For example, if the technical documentation specifies the current variation range for a certain type of fault, if the current variation in the diagnostic operation quantitative factor exceeds this range, it indicates a fault risk.

[0028] Based on the comparison results, intelligent diagnostic assessment information is generated, such as determining whether the distribution network is currently operating normally, experiencing a minor anomaly, or experiencing a serious fault. Simultaneously, methods such as the Analytic Hierarchy Process (AHP) and Principal Component Analysis (PCA) are used to calculate the weights of various quantitative factors in the diagnostic assessment. If a fault recording feature plays a key role in multiple fault diagnoses, its weight is assigned a higher value. This intelligent diagnostic assessment information and the corresponding weight calculation results provide a crucial basis for the subsequent generation of intelligent diagnostic parameter information.

[0029] The intelligent diagnostic assessment information and the corresponding weight calculation results are processed to generate intelligent diagnostic parameter information. Based on the operating status determined by the intelligent diagnostic assessment information and the weight calculation results, a weighted comprehensive calculation is performed on each quantitative factor. For example, if the assessment indicates a minor abnormality and the fault recording feature weight is high, the quantitative factors associated with the fault recording will have a greater impact on the final intelligent diagnostic parameters.

[0030] The calculation results are converted into specific intelligent diagnostic parameters, such as a parameter value that comprehensively reflects the fault type, severity, and impact range. This parameter value can intuitively display the operating status of the distribution network. For example, a value within a certain range indicates normal operation, while values ​​outside of a certain range indicate faults of varying degrees, providing clear decision-making basis for operation and maintenance personnel.

[0031] S103: Process the protection action information based on the historical fault information to generate fault processing strategy parameter information.

[0032] In one implementation, historical fault information and protection action information are extracted and classified to generate historical fault codes, historical fault meanings, historical fault handling methods, protection action triggering conditions, protection action execution results, and periods of high frequency of similar faults. This classification process organizes these various types of information, providing a clearer picture of past distribution network faults and protection actions, and providing fundamental data support for the development of fault handling strategies. Historical faults are coded, with different codes representing different fault types. For example, "01" represents a short circuit, "02" represents an overload, and so on. This coding method facilitates rapid identification and categorization of large amounts of fault information. A detailed explanation of the fault corresponding to each fault code is provided, explaining the mechanism of occurrence, possible causes, and impact on distribution network operations. For example, a short circuit may be caused by line insulation damage, resulting in current looping without passing through the load, which can cause line overheating, equipment damage, and even power outages. The handling measures taken for each historical fault are recorded, including the details and steps of equipment repair and replacement. For example, for a short circuit fault, the solution is to replace the insulated conductors of the faulty line section and perform insulation inspection and repair on related equipment.

[0033] Clarify under what circumstances the protection device will be activated. For example, the overcurrent protection device is triggered when it detects that the line current exceeds the set action threshold. Record the threshold and relevant conditions such as the operating conditions at the time. If the overcurrent protection action threshold of a certain line is 1.2 times the rated current, the protection device will be activated when the line current reaches 1.3 times the rated current. Record the actual effect after the protection device is activated, such as whether the fault is successfully cut off and whether any equipment is damaged. For example, after a certain protection device is activated, it successfully cuts off the fault line, preventing the fault from expanding, but causing a brief power outage in some areas. Through statistical analysis of historical fault information, find out the time periods when similar faults often occur. For example, after analysis, it was found that in a certain area during the high temperature period in summer, due to the large increase in power load, line overload faults frequently occurred.

[0034] Based on historical fault information, the fault-related content in the protection action information is processed to generate fault type determination information and fault severity assessment information. The fault type targeted by the current protection action is determined by combining historical fault code information and fault meaning information. If the historical fault code is "01" and the fault characteristics, such as a transient current increase and a sudden voltage drop, are consistent with a short-circuit fault, the current fault type is determined to be a short-circuit fault. The severity of the fault is assessed by comprehensively considering the fault's impact on distribution network equipment, power supply reliability, and user power consumption. For a short-circuit fault, if it damages critical equipment and causes a large-scale, prolonged power outage, the severity is assessed as severe. If it only causes a short, localized power outage with no significant equipment damage, the severity is assessed as less severe. Quantitative indicators, such as the number of users experiencing power outages, the duration of the outage, and the extent of equipment damage, can be used for reference in this assessment. Based on the fault type determination information and fault severity assessment information, the target data in the protection action information is marked and filtered to generate abnormal protection action data screening results. For example, if a fault is determined to be a severe short circuit, but the protective device does not operate within the specified time, the protective action information is marked as abnormal. All marked abnormal protective action data is then filtered to form the abnormal protective action data screening results. This data may include records of protective device failure, malfunction, or excessively long or short action times.

[0035] The results of abnormal protection action data screening are integrated and quantified to generate fault handling strategy parameter information. This fault handling strategy parameter information characterizes the abnormal protection action conditions and fault severity caused by faults during distribution network protection. The results of abnormal protection action data screening are organized, and abnormal protection action data for the same fault type is aggregated to analyze their commonalities and differences. For example, data on all cases of protection device failure to operate in a timely manner under short-circuit faults is combined to investigate possible common causes, such as protection device aging and communication failures. The integrated data is quantitatively analyzed to calculate key indicators. For example, the frequency of protection device misoperation under a certain fault type is counted, and the average power outage duration caused by abnormal protection action is calculated. Based on the integration and quantification results, fault handling strategy parameter information is generated. These parameters may include recommended protection device adjustment parameters (such as action thresholds and action times), equipment repair or replacement plans, and fault prevention measures for different fault types and abnormal protection action conditions. For example, for short-circuit faults where protection device failure to operate in a timely manner occurs, it is recommended to shorten the overcurrent protection device's action time by a certain amount and regularly inspect and maintain the protection device.

[0036] S104: Process the distribution network protection operation procedure information based on the technical document reference information and the version number information corresponding to the technical document reference information to generate an operation optimization factor.

[0037] In one embodiment, distribution network protection operation process information, technical document reference information, and corresponding version number information are extracted and classified to generate operation step difference information, protection parameter setting deviation information, version compatibility difference information, and technical document compliance difference information. The actual distribution network protection operation process is compared in detail with the operation steps specified in the technical documentation, and any inconsistencies are recorded. For example, the technical documentation may specify that before performing line maintenance, the upstream power supply must be disconnected and the power test and grounding wire connection must be performed. However, in actual operation, this power test is skipped, resulting in operation step difference information. The deviation between the actual protection device parameters (such as the overcurrent protection current threshold and the zero-sequence protection operation time) and the standard parameters specified in the technical documentation is checked. If the technical documentation specifies that the overcurrent protection current threshold for a certain line is 500A, but the actual setting is 600A, this 100A deviation is protection parameter setting deviation information. The compatibility of the currently used technical document version with distribution network protection equipment, system software, etc. is analyzed. For example, if a new technical document version specifies a new protection device communication protocol, but some existing protection devices do not support this protocol, this will generate version compatibility difference information. Comprehensively assess the degree to which the entire distribution network protection operation process adheres to the specifications and standards in the technical documentation, and record any areas that are not followed or not fully implemented. For example, if the technical documentation requires regular functional testing of protection devices, but the actual test cycle exceeds the specified time, this is considered a discrepancy in technical documentation compliance.

[0038] Based on technical document reference information, information on operational procedure differences, protection parameter setting deviations, version compatibility differences, and technical document compliance differences is processed to generate difference type determination information and difference severity assessment information. Based on the technical document reference information, the operational procedure differences, protection parameter setting deviations, version compatibility differences, and technical document compliance differences are analyzed and determined to determine the type of each difference. For example, operational procedure differences may be due to operational specification differences, protection parameter setting deviations may be due to parameter configuration differences, and version compatibility differences may be due to system adaptation differences. The impact of these differences on distribution network protection effectiveness, equipment safety, and power supply reliability is assessed. For protection parameter setting deviations, if they are minor, they may only slightly affect the operating accuracy of the protection device, resulting in a low level of harm. However, if they are excessive, they may cause the protection device to malfunction or fail to operate, seriously impacting the safe operation of the distribution network and resulting in a high level of harm. This assessment assigns a severity level to each difference, such as minor, moderate, or severe.

[0039] Based on the discrepancy type and severity assessment information, target data within the distribution network protection operation process information is marked and filtered to generate discrepancy data screening results. For example, data such as operational step discrepancies and protection parameter setting deviations identified as severe discrepancies with a severity assessment of "serious" is marked. All marked data is then filtered to generate discrepancy data screening results. These screening results may include a series of operational records and parameter settings that do not conform to technical documentation specifications and have a significant impact on distribution network protection.

[0040] Integrate and quantify the results of the difference data screening to generate an operation optimization factor. Organize the difference data screening results, merge and analyze the same type of difference data, and find the correlation and commonality between different differences. For example, put all the difference data related to the deviation of protection parameter settings together and analyze whether there is any systematic reason between them, such as whether the deviation of multiple parameters is caused by parameter calculation errors in a certain link. Quantitatively analyze the integrated data and measure the degree of difference by calculating some key indicators. For example, count the number of different types of differences, calculate the average degree of deviation of protection parameter settings, and evaluate the proportion of missing functions caused by version compatibility differences.

[0041] Based on the results of integration and quantification, an operational optimization factor is generated. This factor can be a numerical value that comprehensively considers various variance factors. Its magnitude reflects the degree of deviation between the distribution network protection operation process and the technical documentation standards, and the degree of optimization required. For example, various variance quantitative indicators are weighted to obtain an operational optimization factor value. A larger value indicates a greater deviation from the standard and a greater need for optimization adjustments. Conversely, a lower value indicates a relatively standardized operation process with less room for optimization.

[0042] S105 : Processing the real-time operation information and the version number information corresponding to the real-time operation information based on the operation optimization factor to generate operation status classification information and processing priority information of the real-time operation information.

[0043] In one embodiment, data extraction and processing are performed on the real-time operation information to generate voltage value information, current value information, power value information, and other operation-related attribute information. Key electrical parameters and other related attribute information are separated from the real-time operation information to lay the foundation for the subsequent analysis of the real-time operation status of the distribution network. This information can intuitively reflect the current operating status of the distribution network and is an important basis for judging whether it is operating normally and evaluating its operating efficiency. The voltage of each node in the distribution network is monitored in real time to obtain data such as voltage amplitude and phase. For example, the real-time voltage amplitude of a line is 220V and the phase is 0°. These data can directly reflect the power supply voltage level and phase condition of the line. If the voltage amplitude exceeds the normal range, it may affect the normal operation of the equipment.

[0044] Collect information such as the current size and current change rate in the line. For example, at a certain moment, the current of a certain line is 50A, and the current change rate is stable. If the current suddenly increases or changes abnormally, it may indicate that the line has potential faults such as overload or short circuit. Calculate and obtain values ​​such as active power and reactive power. Taking a certain distribution network area as an example, the real-time active power is 100kW and the reactive power is 50kvar. The power value reflects the energy transmission and consumption of the power grid and is crucial for evaluating the operating efficiency and stability of the power grid. Other operation-related attribute information includes frequency, harmonic content, power factor, etc. For example, the frequency is 50Hz and the power factor is 0.9. These attribute information can further supplement the description of the operating characteristics of the distribution network. Excessive harmonic content may cause damage to power grid equipment.

[0045] Based on the version number information corresponding to the real-time operation information, voltage, current, power, and other operation-related attribute information are standardized and integrated to generate a standardized operation information set. Since real-time operation information from different sources or different periods may have inconsistent formats, units, or data standards, standardized integration ensures consistency and comparability of all types of information, facilitating subsequent comprehensive analysis. Based on the version number information corresponding to the real-time operation information, the corresponding technical standards and specifications are searched. For example, if the standard corresponding to the version number specifies that voltage values ​​are measured in volts (V), current values ​​are measured in amperes (A), and power values ​​are measured in watts (W), the extracted voltage, current, and power information are uniformly converted to these standard units. Furthermore, the data format is standardized, such as maintaining the same number of decimal places for all data. The processed voltage, current, power, and other operation-related attribute information are integrated into a single set to form a standardized operation information set. All information is integrated under the same standard, facilitating subsequent unified analysis and processing based on this set.

[0046] Based on the operational optimization factor, the importance of the standardized operational information set is evaluated and filtered, generating weights for each piece of information in operational status classification and processing priority determination. If the operational optimization factor indicates that a current overload on a particular line significantly impacts the safe operation of the power grid, then the current value information of that line is relatively more important in operational status classification and processing priority determination. Using methods such as the analytic hierarchy process and fuzzy comprehensive evaluation, the importance of each piece of information is quantitatively assessed and weights are generated. Assuming that, after evaluation, the current value information of a critical line has an importance weight of 0.4, the voltage value information has an importance weight of 0.3, the power value information has an importance weight of 0.2, and the other operational attribute information has an importance weight of 0.1, these weights reflect the relative importance of each piece of information in the comprehensive judgment.

[0047] The importance weights of each piece of information in the operational status classification and processing priority determination are processed to generate operational status classification and processing priority information for real-time operational information. If a line's current value exceeds a set threshold and has a high importance weight, and the voltage and power values ​​also exhibit abnormalities, the line's operational status is comprehensively determined to be "severely abnormal." If only some minor information exhibits minor anomalies and has a low importance weight, the operational status is classified as "minorly abnormal." Based on the importance weights and operational status classification results, the processing priority is determined. Information in the "severely abnormal" state is assigned the highest processing priority due to its significant impact on power grid security, requiring immediate action. Information in the "minorly abnormal" state has a relatively low processing priority, but still requires close attention. Ultimately, the operational status classification and processing priority information for real-time operational information is generated and presented to operation and maintenance personnel in an intuitive manner. For example, a visual interface displays each line's operational status (normal, minorly abnormal, severely abnormal, etc.) and its corresponding processing priority (high, medium, or low), facilitating timely decision-making.

[0048] S106 , inputting the intelligent diagnosis index, the operation status classification information of the real-time operation information, the processing priority information, and the fault processing strategy parameter information into the target adaptive protection model for processing to generate distribution network adaptive protection optimization result information.

[0049] In one embodiment, a sample set of historical distribution network operating data and a preset adaptive protection model are obtained. Long-term operating data is collected from the distribution network's monitoring system, database, and other sources. This data covers electrical parameters such as voltage, current, power, and frequency, as well as equipment operating status (e.g., transformer oil temperature, switch position of switchgear), and environmental parameters (e.g., temperature, humidity, and lightning activity). The data is organized chronologically to form a sample set containing distribution network operating information under different operating conditions. A multilayer perceptron neural network is constructed as the preset model, consisting of an input layer, several hidden layers, and an output layer. The number of input layer nodes is determined by the number of input data features. For example, if the input data contains 10 different distribution network operating characteristics (e.g., 5 electrical parameters, 3 equipment status parameters, and 2 environmental parameters), the input layer has 10 nodes. There are two hidden layers, and the number of neurons in each layer can be determined empirically or through trial and error, e.g., 30 neurons in the first hidden layer and 20 neurons in the second hidden layer. The Reinforced Lu (ReLU) function is used as the activation function, which effectively addresses the vanishing gradient problem and accelerates model convergence. The number of nodes in the output layer is determined by the model output task. If the model is used to determine the protection status of the distribution network (such as normal, warning, and fault), the output layer is set to three nodes, corresponding to different protection states. The initial learning rate is set to 0.001.

[0050] The number of each type of data feature in the distribution network historical operation data sample set was counted, and a sampling ratio was generated based on the balance of feature distribution. The collected distribution network historical operation data sample set was classified and counted. Data features were categorized into electrical parameter features, device status features, and environmental features. The number of occurrences of each type of feature in the sample set was counted. For example, the electrical parameter feature appeared 5,000 times, the device status feature appeared 3,000 times, and the environmental feature appeared 2,000 times. Analysis revealed that the electrical parameter feature accounts for a large proportion and may dominate model training, while the device status and environmental features account for a relatively small proportion and are easily overlooked. To ensure that the model can fully learn from each type of feature, the sampling ratio was adjusted based on the principle of balance. The adjusted sampling ratio was calculated to achieve a sampling ratio of 40% for electrical parameter features, 35% for device status features, and 25% for environmental features. This ensures a more balanced selection of different data types in subsequent sampling.

[0051] Based on the sampling ratio, a multimodal stratified sampling process is performed on the distribution network's historical operating data sample set to generate a preset number of sampling feature combinations. Sampling is performed according to the different modes and layers of the data. In terms of temporal modality, stratification is performed by season (spring, summer, autumn, and winter) and load period (peak, off-peak, and off-peak). For example, data is selected according to the previously generated sampling ratio in different time layers, such as the summer peak load period and the winter off-peak load period. In terms of spatial modality, sampling is stratified by distribution network region (such as the power supply areas of different substations) and line level (high-voltage lines and low-voltage lines). In terms of data type modality, stratification is performed by electrical parameters, equipment status, and environmental parameters. According to the adjusted sampling ratio, data is extracted from each layer to form sampling feature combinations. A set of 1,000 sampling feature combinations is generated, each containing data from different modes and layers. For example, a sampling feature combination may include the electrical parameters of a high-voltage line during the summer peak period, the equipment status of the corresponding area, and the environmental parameters at that time.

[0052] Based on statistical tests performed on any key feature and each sampled feature combination, the data is divided into a good protection effectiveness group and a poor protection effectiveness group. This data classification facilitates targeted model training, enabling it to distinguish distribution network operating states under different protection effects. "Protection action time" is selected as the key feature. For each sampled feature combination, the difference between its corresponding protection action time and the ideal protection action time is analyzed. Statistical tests are performed, such as calculating the difference between the two, and a threshold is set based on the distribution of the difference. If the difference between the protection action time and the ideal time corresponding to a sampled feature combination is within a reasonable range, and other relevant indicators (such as the time it takes for equipment to resume normal operation after a fault and the scope of the power outage) also meet the requirements, the sampled feature combination is classified as the good protection effectiveness group. Conversely, if the protection action time is too long, resulting in a wider fault impact range and severe equipment damage, the protection effectiveness group is classified as the poor protection effectiveness group. In this way, the 1,000 sampled feature combinations are divided into two groups, providing data samples with different protection effects for subsequent model training.

[0053] The pre-set adaptive protection model is iteratively trained based on a good protection effectiveness group and a poor protection effectiveness group, generating a trained adaptive protection model and performance metrics. During the training process, the model's weights and bias parameters are adjusted using the backpropagation algorithm. During each training session, the model predicts the protection status of the distribution network based on a combination of input sampled features. The predicted results are then compared with the actual protection effectiveness, and a loss value (such as cross-entropy loss) is calculated. The loss value is propagated from the output layer to the input layer using the backpropagation algorithm, and the weights and biases of neurons in each layer are adjusted to gradually reduce the loss value. After multiple iterative training (for example, setting the number of iterations to 500), the model gradually converges. After training is complete, the model is evaluated using performance metrics such as precision, recall, and F1 value. Precision indicates the proportion of protection statuses correctly predicted by the model, recall measures the proportion of data that actually fall within a certain protection status category, and F1 value comprehensively considers both precision and recall. Assuming the model achieves 85% precision, 80% recall, and 0.82 on the test set, these metrics reflect the model's ability to distinguish between good and poor protection effectiveness.

[0054] If the model identifies historical fault information and protection action information in the training results as key features influencing the evaluation of distribution network protection effectiveness, the trained model is designated as the target adaptive protection model. The trained multilayer perceptron neural network model is analyzed to examine the importance of historical fault information and protection action information in the model. This is determined by calculating the correlation between each input feature and the model output, or by observing the response of hidden layer neurons to different input features. If historical fault information and protection action information have a high correlation with the model output, or cause significant changes in the hidden layer response, the model has identified them as key features influencing the evaluation of distribution network protection effectiveness. For example, analysis reveals that when the fault type and severity in the historical fault information change, the model outputs a significant change in the protection status, and protection action information (such as action time and action type) has a significant impact on the model's decision-making. In this case, the trained model is designated as the target adaptive protection model. If the model does not identify these as key features, the sampling ratio is adjusted, the number of hidden layer neurons is increased or decreased, the learning rate is adjusted, and training is repeated until the requirements are met.

[0055] In another embodiment, a target-adaptive protection model processes intelligent diagnostic indicators, real-time operating status classification information, processing priority information, and fault handling strategy parameter information to generate a distribution network protection effectiveness correlation evaluation value. By comprehensively considering the intelligent diagnostic indicators, real-time operating status classification and processing priority information, and fault handling strategy parameter information, the target-adaptive protection model quantitatively evaluates the current distribution network protection effectiveness, generating a numerical value that reflects the degree of correlation between each aspect of information and the protection effectiveness, providing a basis for subsequent optimization. The target-adaptive protection model inputs intelligent diagnostic indicators (generated by processing multi-source information such as distribution network operating data and fault recording data, reflecting the potential fault risk of the distribution network), real-time operating status classification information (such as normal, minor abnormality, and severe abnormality, determined based on real-time operating data and operation optimization factors), processing priority information (determined based on operating status classification and information importance weighting), and fault handling strategy parameter information (generated based on historical fault and protection action information, including parameters related to the fault handling solution) into the target-adaptive protection model. The model internally integrates and analyzes this information based on its own algorithms and trained parameters. For example, the model may determine whether there are potential fault hazards in the current distribution network based on intelligent diagnostic indicators, determine the severity of the fault based on the status classification of real-time operation information, and then evaluate the effectiveness of the current protection strategy with reference to the fault handling strategy parameter information. Finally, it outputs a distribution network protection effect-related evaluation value, which can be a number between 0 and 1. The closer the value is to 1, the better the protection effect, and the closer it is to 0, the worse the protection effect.

[0056] Based on the distribution network protection effectiveness correlation evaluation values ​​generated above, the protection strategy decision vector within the target adaptive protection model is optimized and adjusted to generate a protection resource allocation deviation vector. The protection strategy decision vector contains a series of key parameters that determine the distribution network protection strategy, such as the protection device's action threshold, action time, protection range settings, and the order of fault handling. By analyzing the distribution network protection effectiveness correlation evaluation values, if the evaluation value is low, it indicates that there may be problems with the current protection strategy. For example, if the evaluation finds that a certain area frequently experiences faults and the protection action is not timely, it may be that the action threshold of the protection device in this area is set improperly or the action time is too long. Based on this, the protection strategy decision vector is optimized and adjusted. The difference between the decision vectors before and after the adjustment is calculated, and this difference forms the protection resource allocation deviation vector. For example, the original overcurrent protection action threshold of a certain line is 500A, which is adjusted to 450A based on the evaluation results, and the action time is adjusted from the original 0.5 second to 0.3 second. Then the changes in these parameters will be reflected in the protection resource allocation deviation vector, such as the action threshold change of -50A and the action time change of -0.2 second. These changes reflect the changes that need to be made in the protection resource allocation to improve the protection effect.

[0057] The protection resource allocation deviation vector is parsed and converted to generate distribution network adaptive protection optimization results. The changes in the action threshold and action time in the protection resource allocation deviation vector are analyzed to determine the specific impact of these changes on distribution network protection resource allocation. Based on the parsed results, they are converted into practical optimization measures. A decrease in the action threshold may indicate a need to adjust the parameters of the corresponding protection device to enable more sensitive fault detection. A shortened action time may indicate the need to optimize the protection device's communication link or control logic to speed up its operation. These specific optimization measures are organized into distribution network adaptive protection optimization results information, which can include detailed equipment adjustment plans, protection strategy modification recommendations, and fault handling process optimization plans. For example, the generated optimization results may include: "Adjust the action threshold of the overcurrent protection device on a certain line to 450A; inspect and optimize the device's communication link to ensure the action time is reduced to less than 0.3 seconds; develop a fault emergency plan for this area, prioritizing the handling of potential severe faults." This information can directly guide operations and maintenance personnel in optimizing and improving the distribution network protection system.

[0058] This application focuses on intelligent diagnosis and adaptive protection of distribution networks. The core is to integrate multi-source data to improve the reliability and safety of distribution network operations. Multi-source information such as distribution network operation, fault recording, and protection device parameters is obtained. Based on the reference information of technical documents, this data is processed to generate intelligent diagnosis parameter information to provide a basis for fault diagnosis; historical fault and protection action information is used to generate fault handling strategy parameter information to facilitate the formulation of targeted measures. At the same time, the protection operation process information is processed in combination with the technical documents and their version numbers to obtain the operation optimization factor, which is used to analyze real-time operation information and generate operation status classification and processing priority information.

[0059] Next, the target adaptive protection model is obtained. By sampling and grouping the historical operating data of the distribution network, the preset model is iteratively trained. If the historical fault and protection action information in the training results are identified as key features, it is determined as the target model. Finally, the intelligent diagnosis indicators, operating status classification and processing priority information, and fault handling strategy parameter information are input into the target model to obtain the distribution network protection effect correlation evaluation value. Based on this, the internal protection strategy decision vector of the model is optimized, and the protection resource allocation deviation vector is generated. After analytical conversion, the distribution network adaptive protection optimization result information is obtained, realizing the precise protection and optimization of the distribution network. This technology integrates multi-source data, from diagnosis, strategy formulation to model optimization, to form a complete system, effectively improving the intelligence level and operational stability of the distribution network.

[0060] In one embodiment, Figure 2 As shown, the present application also provides a distribution network intelligent diagnosis and adaptive protection device based on multi-source data fusion, including:

[0061] Acquisition module 201, used to acquire distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, environmental parameter information, historical fault information, protection action information and technical document reference information;

[0062] The processing module 202 is used to process the distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, and environmental parameter information based on the technical document reference information to generate intelligent diagnosis parameter information; process the protection action information based on the historical fault information to generate fault handling strategy parameter information; process the distribution network protection operation process information based on the technical document reference information and the version number information corresponding to the technical document reference information to generate an operation optimization factor; process the real-time operation information and the version number information corresponding to the real-time operation information based on the operation optimization factor to generate operation status classification information and processing priority information of the real-time operation information; input the intelligent diagnosis index, the operation status classification information and processing priority information of the real-time operation information and the fault handling strategy parameter information into the target adaptive protection model for processing to generate distribution network adaptive protection optimization result information.

[0063] Each embodiment in this application is described in a related manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the evaluation of the distribution network intelligent diagnosis and adaptive protection method based on multi-source data fusion, electronic device, electronic device, and readable storage medium embodiment, since it is basically similar to the distribution network intelligent diagnosis and adaptive protection method embodiment based on multi-source data fusion described above, the description is relatively simple, and the relevant parts can be referred to the partial description of the distribution network intelligent diagnosis and adaptive protection method embodiment based on multi-source data fusion described above.

Claims

1. A distribution network intelligent diagnosis and adaptive protection method based on multi-source data fusion, characterized in that: include: Obtain distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, environmental parameter information, historical fault information, protection action information and technical document reference information; Based on technical document reference information, distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, and environmental parameter information are processed to generate intelligent diagnosis parameter information; Processing protection action information based on historical fault information to generate fault handling strategy parameter information; Processing the distribution network protection operation process information based on the technical document reference information and the version number information corresponding to the technical document reference information to generate an operation optimization factor; Processing the real-time operation information and the version number information corresponding to the real-time operation information based on the operation optimization factor to generate operation status classification information and processing priority information of the real-time operation information; The intelligent diagnosis indicators, operation status classification information of real-time operation information, processing priority information and fault handling strategy parameter information are input into the target adaptive protection model for processing to generate distribution network adaptive protection optimization result information.

2. The method according to claim 1, wherein Based on the technical document reference information, the distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, and environmental parameter information are processed to generate intelligent diagnosis parameter information, including: Perform feature extraction and processing on distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, and environmental parameter information to generate distribution network operation characteristics, fault recording characteristics, protection device parameter characteristics, equipment status characteristics, and environmental characteristics; Quantitatively analyze and process the distribution network operation characteristics and protection device parameter characteristics to generate the basic quantitative factors of the distribution network; Quantitatively analyze and process fault recording characteristics, equipment status characteristics, and environmental characteristics to generate diagnostic operation quantitative factors; Based on the technical document reference information, the distribution network basic quantitative factors and diagnostic operation quantitative factors are processed to generate intelligent diagnostic evaluation information and corresponding weight calculation result information; The intelligent diagnosis evaluation information and the corresponding weight calculation result information are processed to generate intelligent diagnosis parameter information.

3. The method according to claim 1, wherein Process protection action information based on historical fault information and generate fault handling strategy parameter information, including: Extract and classify historical fault information and protection action information to generate historical fault code information, historical fault meaning information, historical fault handling method information, protection action triggering condition information, protection action execution result information, and similar fault frequent occurrence time period information; Process the fault-related content in the protection action information based on historical fault information to generate fault type determination information and fault hazard degree assessment information; Mark and filter the target data in the protection action information based on the fault type determination information and the fault hazard level assessment information, and generate abnormal protection action data screening results; The abnormal protection action data screening results are integrated and quantified to generate fault handling strategy parameter information, where the fault handling strategy parameter information is used to characterize the abnormal protection action conditions and fault hazard severity caused by faults during the distribution network protection process.

4. The method according to claim 1, wherein The distribution network protection operation process information is processed based on the technical document reference information and the version number information corresponding to the technical document reference information to generate an operation optimization factor, including: Extract and classify distribution network protection operation process information, technical document reference information, and corresponding version number information to generate operation step difference information, protection parameter setting deviation information, version compatibility difference information, and technical document compliance difference information; Based on the technical document reference information, the operation step difference information, protection parameter setting deviation information, version compatibility difference information, and technical document compliance difference information are processed to generate difference type judgment information and difference severity assessment information; Mark and filter target data in the distribution network protection operation process information based on the difference type judgment information and the difference severity assessment information to generate a difference data screening result; The results of differential data screening are integrated and quantified to generate operation optimization factors.

5. The method according to claim 4, wherein The real-time operation information and the version number information corresponding to the real-time operation information are processed based on the operation optimization factor to generate operation status classification information and processing priority information of the real-time operation information, including: Perform data extraction and processing on real-time operation information to generate voltage value information, current value information, power value information, and other operation-related attribute information; Based on the version number information corresponding to the real-time operation information, the voltage value information, current value information, power value information, and other operation-related attribute information are standardized and integrated to generate a standardized operation information set; Based on the operation optimization factors, the importance of the standardized operation information set is evaluated and screened, and the importance weight of each information is generated in the operation status classification and processing priority judgment; The importance weight of each information in the operation status classification and processing priority judgment is processed to generate the operation status classification information and processing priority information of the real-time operation information.

6. The method according to claim 1, wherein Obtain the target adaptive protection model, including: Obtain distribution network historical operation data sample sets and preset adaptive protection models; Count the number of various types of data features in the distribution network historical operation data sample set, and generate a sampling ratio based on the balance of feature distribution; Perform multimodal stratified sampling processing on the distribution network historical operation data sample set based on the sampling ratio to generate a preset number of sampling feature combinations; Based on the statistical test of any key feature combined with each sampling feature, the data were divided into a good protection effect group and a poor protection effect group; Iteratively train the preset adaptive protection model based on the good protection effect group and the poor protection effect group to generate the trained adaptive protection model and performance indicators; If the historical fault information and protection action information in the training results are identified by the model as key features that affect the evaluation of distribution network protection effectiveness, the trained model will be used as the target adaptive protection model.

7. The method according to claim 1, wherein The intelligent diagnosis indicators, the operating status classification information of the real-time operating information, the processing priority information, and the fault handling strategy parameter information are input into the target adaptive protection model for processing to generate the distribution network adaptive protection optimization result information, including: Based on the target adaptive protection model, the intelligent diagnosis indicators, the operation status classification information of the real-time operation information, the processing priority information and the fault handling strategy parameter information are processed to generate the distribution network protection effect correlation evaluation value; Based on the distribution network protection effect correlation evaluation value generated above, the protection strategy decision vector within the target adaptive protection model is optimized and adjusted to generate a protection resource allocation deviation vector; The protection resource allocation deviation vector is parsed and converted to generate distribution network adaptive protection optimization result information.

8. A distribution network intelligent diagnosis and adaptive protection device based on multi-source data fusion, characterized in that: The device comprises: An acquisition module is used to obtain distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, environmental parameter information, historical fault information, protection action information and technical document reference information; The processing module is used to process distribution network operation data information, fault recording data information, protection device parameter information, equipment status monitoring information, and environmental parameter information based on technical document reference information to generate intelligent diagnosis parameter information; process protection action information based on historical fault information to generate fault handling strategy parameter information; process distribution network protection operation process information based on technical document reference information and version number information corresponding to the technical document reference information to generate operation optimization factors; process real-time operation information and version number information corresponding to the real-time operation information based on the operation optimization factors to generate operation status classification information and processing priority information of the real-time operation information; input the intelligent diagnosis indicators, operation status classification information and processing priority information of the real-time operation information, and fault handling strategy parameter information into the target adaptive protection model for processing to generate distribution network adaptive protection optimization result information.

9. An electronic device, characterized in that: include: a first processor; and a memory for storing executable instructions of the first processor; The first processor is configured to execute the distribution network intelligent diagnosis and adaptive protection method based on multi-source data fusion according to any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the second processor, the method for intelligent diagnosis and adaptive protection of a distribution network based on multi-source data fusion as described in any one of claims 1 to 7 is implemented.

Images