Power distribution network partial discharge fault positioning device and method

Abstract

The invention discloses a power distribution network partial discharge fault positioning device and method, and relates to the technical field of power distribution network fault positioning, and the device comprises a data collection module, a multi-dimensional analysis module, an intelligent traceability module, a directional verification module, a report generation module and a local database. All the modules collaboratively integrate discharge pulses and load data of corresponding lines, through feature extraction, correlation analysis and historical data verification, a plurality of discharge sources are accurately distinguished, a fault branch is locked, then through space-time signal propagation path analysis and multi-algorithm collaborative optimization, a fault is accurately positioned to a specific equipment node in the power supply branch, and the fault is accurately positioned. And a structured report is output, so that the positioning accuracy and efficiency are greatly improved, and reliable support is provided for operation and maintenance of the power distribution network.

Description

Technical Field

[0001] This invention relates to the field of distribution network fault location technology, specifically to a distribution network partial discharge fault location device and method. Background Technology

[0002] Partial discharge fault location in power distribution networks is now one of the key technologies to ensure the stable operation of the power grid. In recent years, the development of signal processing, data analysis and machine learning technologies has promoted the evolution of the location approach based on multi-dimensional data fusion, and gradually realized the upgrade from coarse branch location to specific equipment node location. At the same time, through the optimization of feature extraction and correlation analysis technologies, the reliability and efficiency of fault location have been further improved, laying a technical foundation for the rapid investigation of faults in power distribution networks.

[0003] Existing technologies, such as the invention patent application with announcement number CN119881524A, disclose a method for locating partial discharge faults in power distribution networks based on directional calibration. The method includes: installing Rogowski coil directional calibration sensors at power distribution network nodes to identify partial discharges on the CC line, comparing the synchronization time difference of the end sensor signals, and locating the fault by combining multi-sensor data, thereby reducing costs and increasing efficiency. Existing technologies, such as the invention patent application with announcement number CN115754625B, disclose a fault monitoring device for power distribution equipment in a power distribution network. The device includes: a spliced ​​ring body, a moving and moving monitoring mechanism, which can move along the cable and drive the monitoring components to reciprocate in a spiral motion, thereby achieving all-round monitoring, positioning, image acquisition and marking.

[0004] As can be seen from the above solutions, the existing solutions each have their limitations. The Rogowski coil-based location method only locates faults by comparing the time difference of sensor signals, without combining it with line load change data for correlation analysis, making it difficult to accurately trace the cause of the fault. While monitoring devices relying on mechanical moving mechanisms can achieve omnidirectional cable scanning, they lack the ability to effectively distinguish multiple discharge sources, and the location accuracy is mostly limited to the general area level, making it difficult to pinpoint specific equipment nodes. At the same time, existing technologies generally do not fully integrate historical fault data and load trend verification. When facing complex scenarios with multiple causes and multiple discharge sources in the distribution network, the reliability and specificity of the location are insufficient, making it difficult to meet the actual needs of refined fault diagnosis. Summary of the Invention

[0005] To address the aforementioned technical shortcomings, the present invention aims to provide a device and method for locating partial discharge faults in power distribution networks.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The first aspect of the present invention provides a partial discharge fault location device for a distribution network, including a data acquisition module: used to synchronously and in real time acquire discharge pulse signals and load change data of the corresponding lines from partial discharge sensors deployed at key nodes of the distribution network and monitoring and data acquisition systems.

[0007] Multi-dimensional analysis module: used to perform fusion analysis on the received discharge pulse signal and the corresponding line load change data.

[0008] Intelligent tracing module: It is used to determine the correlation strength between discharge activities and load events according to preset load correlation judgment rules, and generate initial tracing assumptions for specific power supply branches.

[0009] Directional verification module: Used to analyze the propagation path of the power supply branch pointed to by the initial tracing hypothesis by calling the spatiotemporal signal data of all monitoring points of the branch, so as to verify the hypothesis and locate the fault to the specific equipment node in the power supply branch.

[0010] Report generation module: Used to generate and output structured reports containing specific faulty device nodes and correlation criteria.

[0011] The second aspect of the present invention provides a location method for a partial discharge fault location device in a power distribution network, comprising step 1. Data acquisition: synchronously and in real time acquiring discharge pulse signals and load change data of the corresponding lines from partial discharge sensors deployed at key nodes of the power distribution network and from a monitoring and data acquisition system.

[0012] Step 2. Multidimensional analysis: Perform fusion analysis on the received discharge pulse signals and load change data.

[0013] Step 3. Intelligent source tracing: Based on the preset load association judgment rules, determine the correlation strength between discharge activities and load events, and generate initial source tracing hypotheses for specific power supply branches.

[0014] Step 4. Targeted Verification: For the power supply branch pointed to by the initial tracing hypothesis, the spatiotemporal signal data of all monitoring points in the branch are called to perform propagation path analysis in order to verify the hypothesis and locate the fault to a specific equipment node in the power supply branch.

[0015] Step 5. Report Generation: Generate and output a structured report containing the specific faulty device node and correlation criteria.

[0016] The beneficial effects of the present invention are as follows: (1) The first part of the present invention: synchronously captures the discharge pulse signal and the load change data of the corresponding line, accurately locks the correspondence between the two in the time dimension, avoids the deviation of the correlation analysis caused by the asynchronous data, and the real-time acquisition of dual-source data not only enriches the analysis dimension, but also improves the data credibility through complementary verification, effectively filters the environmental interference when a single signal is acquired, and provides high-fidelity and strong correlation basic data support for subsequent in-depth fusion analysis.

[0017] (2) The second part of the present invention: By deeply integrating discharge signals and load data, the intrinsic relationship between the two is accurately explored. At the same time, the correlation strength is scientifically divided based on the load correlation judgment rules, multiple discharge sources are efficiently distinguished, and the initial source tracing hypothesis is generated by combining historical data and load trends to avoid blind positioning and significantly improve the targeting and reliability of source tracing, laying a solid foundation for accurate fault location.

[0018] (3) The third part of the invention: Directional verification uses spatiotemporal signal analysis and multi-algorithm optimization to accurately locate faults from branches to specific equipment nodes, breaking through the limitations of traditional coarse location. The structured report integrates core criteria and fault information, presenting key content intuitively, which facilitates maintenance personnel to quickly obtain effective information and carry out maintenance work, greatly shortening the troubleshooting time and reducing maintenance costs. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the system modules of the present invention.

[0021] Figure 2 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Reference Figure 1As shown, the present invention provides a partial discharge fault location device for power distribution networks, including a data acquisition module, a multi-dimensional analysis module, an intelligent tracing module, a directional verification module, a report generation module, and a local database.

[0024] It should be noted that the data acquisition module is connected to the multi-dimensional analysis module, the multi-dimensional analysis module is connected to the intelligent traceability module, the intelligent traceability module is connected to the targeted verification module, and the targeted verification module is connected to the report generation module. The local database is connected to the data acquisition module, the multi-dimensional analysis module, the intelligent traceability module, the targeted verification module, and the report generation module.

[0025] The data acquisition module is used to synchronously and in real time acquire discharge pulse signals and corresponding line load change data from partial discharge sensors deployed at key nodes of the power distribution network and from the monitoring and data acquisition system.

[0026] The multi-dimensional analysis module is used to perform fusion analysis on the received discharge pulse signal and the corresponding line load change data.

[0027] In a specific embodiment of the present invention, the fusion analysis of the received discharge pulse signal and the load change data of the corresponding line includes: a discharge feature extraction unit: used to identify and extract the timing occurrence pattern and intensity change trend of the discharge pulse from the received discharge pulse signal to form feature information characterizing the discharge activity.

[0028] Correlation analysis unit: Used to compare discharge activities with specific load events identified from the load change data of the corresponding line to form correlation information between discharge activities and load events.

[0029] In a specific embodiment of the present invention, the method for forming characteristic information representing discharge activity is as follows: the received discharge pulse signal is filtered, and an adaptive dynamic threshold algorithm based on signal statistical features is used to identify and extract the effective discharge pulse signal. The effective discharge pulse signal is clustered according to the amplitude, polarity and phase characteristics of the discharge pulse signal, and the effective discharge pulse signal representing the unified discharge power source is divided into a cluster. By extracting the statistical features and temporal features of each cluster of effective discharge pulse signals per unit time, characteristic information representing discharge activity is formed.

[0030] The filtering process can be implemented using mature digital filtering techniques in the field, such as bandpass filters, to suppress noise and periodic interference outside a specific frequency band, thereby improving the signal-to-noise ratio. This technique is common knowledge in the field of signal processing, and those skilled in the art can select and implement it as needed.

[0031] For example, statistical characteristics include pulse repetition rate and intensity value, while temporal characteristics include, but are not limited to, the time regularity of pulse occurrence and intensity variation trend.

[0032] It should be noted that clustering analysis of effective discharge pulse signals based on the amplitude, polarity, and phase characteristics of the discharge pulse signals refers to the automatic classification of identified effective discharge pulses using clustering algorithms in unsupervised machine learning. The technical basis lies in the inherent stability and consistency of the amplitude distribution, polarity characteristics, and phase points occurring within the power frequency cycle of discharge pulses generated by the same discharge source. This forms a dense cluster in a multi-dimensional feature space, while the pulse characteristics of different discharge sources are distributed in different spatial regions. Through clustering analysis, mixed pulse signals can be automatically separated and grouped into several independent clusters, each corresponding to an independent discharge source. This process lays the foundation for distinguishing and accurately locating multiple concurrent discharge faults. Those skilled in the art can flexibly configure this based on conventional experience and specific scenarios without affecting the implementation and reproduction of the technical solution of this invention.

[0033] It should be noted that the adaptive dynamic threshold algorithm based on signal statistical characteristics refers to the algorithm that dynamically sets a threshold associated with the statistical distribution characteristics of background noise. When the amplitude of the discharge pulse signal is greater than the threshold, it is determined to be a valid discharge pulse signal. This method can effectively adapt to the fluctuation of the signal strength on site and is a commonly used technique in this field for extracting transient pulses from noise. Those skilled in the art can flexibly configure it based on conventional experience and specific scenarios without affecting the implementation and reproduction of the technical solution of this invention.

[0034] In a specific embodiment of the present invention, the correlation information between the discharge activity and the load event is specifically obtained by the following method: if the load change amplitude in the received load data is greater than the preset warning load change amplitude threshold, it is determined as a load event, and its load change rate, load change amplitude and duration are recorded. Each load event is statistically obtained, and each identified load event is aligned with each cluster of discharge activities obtained through cluster analysis on the time axis.

[0035] For each load event, the feature information of each cluster discharge activity within a preset time correlation window is extracted. The Pearson correlation coefficient algorithm is used to calculate the correlation coefficient between the feature data of each load event and the feature information of each cluster discharge activity. If the correlation coefficient between the feature data of a load event and the feature information of a discharge activity is greater than the preset correlation threshold, it is determined that there is a correlation between the load event and the discharge activity. The feature data of the load event and the feature information of the corresponding discharge activity are integrated to form the correlation information between the discharge activity and the load event.

[0036] The characteristic data of the load event refers to its load change rate, load change magnitude, and duration.

[0037] It should be noted that the Pearson correlation coefficient algorithm is a well-known and mature method in the field of data analysis for measuring the degree of linear correlation between two variables. Its specific calculation process is well known to those skilled in the art. That is, the correlation coefficient is obtained by calculating the ratio of the product of the covariance of the two variables to their respective standard deviations. This invention directly applies this standard algorithm and performs calculations based on the extracted load event feature data and discharge activity feature information. Those skilled in the art can flexibly configure it according to conventional experience and specific scenarios without affecting the implementation and reproduction of the technical solution of this invention.

[0038] The intelligent tracing module is used to determine the correlation strength between discharge activities and load events according to preset load correlation judgment rules, and to generate initial tracing hypotheses for specific power supply branches.

[0039] In a specific embodiment of the present invention, the preset load association determination rule includes a correlation strength rule and a timing matching rule, wherein the correlation strength rule means that the correlation coefficient between the characteristic data of the load event and the characteristic information of the discharge activity must be greater than a preset association threshold.

[0040] The timing matching rule means that the absolute value of the time difference between the occurrence time of discharge activity characteristic information and the occurrence time of load event must be less than the preset timing tolerance threshold.

[0041] In a specific embodiment of the present invention, the method for determining the correlation strength between a discharge activity and a load event is as follows: when both a discharge activity and a load event satisfy the correlation strength rule and the timing matching rule, the discharge activity and the load event are determined to be strongly correlated.

[0042] If a discharge activity and a load event satisfy only one of the correlation strength rule and the timing matching rule, then the discharge activity and the load event are determined to be moderately correlated.

[0043] If neither the correlation strength rule nor the timing matching rule is satisfied between a certain discharge activity and a certain load event, then the discharge activity and the load event are determined to be weakly correlated.

[0044] In a specific embodiment of the present invention, the method for generating the initial tracing hypothesis for a specific power supply branch is as follows: when it is determined that there is a strong correlation between the discharge activity and the load event, the power supply branch where the discharge activity occurred is directly taken as the initial tracing hypothesis.

[0045] When the discharge activity and load event are determined to be moderately correlated, the matching degree analysis is performed on the characteristic information of the current discharge activity and the typical fault development mode in the historical discharge characteristic database of the branch. At the same time, the consistency verification of the type attribute of the current load event and the change trend of the historical load event is combined. If the matching degree between the characteristics of the current discharge activity and the historical fault development mode is greater than the preset high matching threshold and the type attribute of the current load event is consistent with the change trend of the historical load event, then the power supply branch corresponding to the discharge activity is taken as the initial source hypothesis. Otherwise, no initial source hypothesis for the specific power supply branch is generated.

[0046] When the discharge activity is determined to be weakly correlated with the load event, no initial source tracing assumptions for the specific power supply branch are generated.

[0047] In one specific embodiment, a matching degree analysis is performed on the characteristic information of the current discharge activity and the typical fault development patterns in the historical discharge characteristic database of the branch. The specific method is as follows: extract the characteristic information of each cluster of discharge activities obtained after cluster analysis from the current discharge pulse signal, and normalize it to construct a multi-dimensional current feature vector. Call the historical feature vector corresponding to the typical fault development pattern most similar to the current operating condition from the historical discharge characteristic database of the power supply branch. Use the cosine similarity algorithm to calculate the matching degree between the two feature vectors. Treat the two feature vectors as two line segments in multi-dimensional space and calculate the cosine value of their included angle, that is, the matching degree S = (current feature vector · historical feature vector) / (||current feature vector|| × ||historical feature vector||). The matching degree S value ranges between [-1, 1]. The closer it is to 1, the higher the matching degree. The calculated matching degree S is compared with the preset high matching threshold for judgment.

[0048] In one specific embodiment, the consistency verification of the type attributes of the current load event with the changing trends of historical load events in the same period is carried out by: parsing the type attributes of the current load event and quantifying its load change rate, load change amplitude, and duration; retrieving the sequence data of the same type of load event that occurred in the same period in history from the database; using a dynamic time warping algorithm to align the length difference between the current load event and the historical load event sequence on the time axis; extracting key trend features, such as the peak distribution and fluctuation period of the load change rate; and applying the Pearson correlation coefficient algorithm to calculate the correlation coefficient between the characteristic data of the current load event and the trend features of the historical load event sequence to determine the consistency of the changing trends.

[0049] The directional verification module is used to analyze the propagation path of the power supply branch pointed to by the initial tracing hypothesis by calling the spatiotemporal signal data of all monitoring points of the branch, so as to verify the hypothesis and locate the fault to a specific equipment node in the power supply branch.

[0050] In a specific embodiment of the present invention, the method for verifying the hypothesis and locating the fault to a specific device node in the power supply branch is as follows: based on the specific power supply branch used as the initial tracing hypothesis, the spatiotemporal signal data of all monitoring points deployed along the power supply branch are obtained, and the timestamps corresponding to the same valid discharge pulse signal recorded by each monitoring point are compared using the signal arrival time difference positioning method to obtain the timestamp difference between each pair of monitoring points.

[0051] The propagation speed of electromagnetic waves in the cable is extracted from the local database. The difference in timestamps between each pair of monitoring points is multiplied by the propagation speed to obtain the corresponding relative distance difference. Based on this, a hyperbola positioning algorithm is used to construct a hyperbola with each pair of monitoring points as the focus. Several hyperbolas are obtained by statistical analysis. The least squares method is used to optimize the intersection area of ​​several hyperbolas and locate the fault point to the specific equipment node in the power supply branch.

[0052] The spatiotemporal signal data refers to a data set containing discharge pulse signals with timestamp information and corresponding spatial location information of monitoring points. The timestamp information is used to record the absolute time when the signal arrives at each monitoring point, and the spatial location information is used to identify the specific coordinates or number of each monitoring point in the power distribution network topology.

[0053] It should also be noted that the signal arrival time difference positioning method, hyperbolic positioning algorithm, and least squares method involved in this invention are all well-known classic algorithms in the fields of signal processing and mathematical positioning. Among them, the signal arrival time difference positioning method is a standard method for coarse positioning by calculating the time delay difference of the signal arriving at different monitoring points. The hyperbolic positioning algorithm is a core algorithm based on geometric principles that uses the time delay difference to construct the intersection of hyperbolas to solve for the target position. The least squares method is a classic mathematical optimization tool used to optimize positioning results and reduce measurement errors. The basic principles, calculation processes, and implementation methods of these algorithms are all common knowledge in the prior art. Those skilled in the art can standardize and adjust the conventional parameters in the algorithms according to the actual application scenario, such as the time synchronization error tolerance in the signal arrival time difference positioning method, the convergence accuracy of the hyperbolic positioning algorithm, and the number of iterations in the least squares method. Those skilled in the art can flexibly configure them based on conventional experience and specific scenarios without affecting the implementation and reproduction of the technical solution of this invention.

[0054] For example, the fault point is located to a specific equipment node in the power supply branch. For instance, suppose there are three monitoring points on a power supply branch of the distribution network, located at the midpoint and end of the line respectively. When the power supply branch is identified as the initial source of the fault, the system calls the spatiotemporal signal data collected synchronously by the three monitoring points. By comparison, the timestamps of the same valid discharge pulse signal arriving at monitoring points A, B and C are identified as T1, T2 and T3 respectively. The time differences ΔT_AB, ΔT_BC and ΔT_AC are calculated. Based on the known propagation speed V of electromagnetic waves in the cable, the time differences are multiplied by the propagation speed to obtain the relative distance differences ΔD_AB and ΔD_BC and ΔD_AC. Based on this, three hyperbolas are constructed with the coordinates of monitoring points A and B, B and C and A and C as the focal points respectively. These three hyperbolas will form a smaller intersection area. Finally, the least squares method is used to optimize the calculation of this area to accurately locate the fault point to the specific equipment node of the branch, thereby using all available information to achieve the highest accuracy of the location.

[0055] The report generation module is used to generate and output a structured report containing the specific faulty device node and correlation criteria.

[0056] In a specific embodiment of the present invention, the method for generating and outputting a structured report containing specific faulty device nodes and correlation criteria is as follows: integrating the determined specific faulty device node information and the generated correlation criteria, wherein the specific faulty device node information includes the device number, location coordinates and the power supply branch to which it belongs, and the generated correlation criteria include the correlation information between discharge activity and load event, the correlation strength, the characteristic information of discharge activity and the corresponding load event characteristic data, and using a built-in report template engine to structurally encapsulate the determined specific faulty device node information and the generated correlation criteria, automatically generating a standardized report content, and outputting the complete structured report to the power distribution network monitoring system.

[0057] Reference Figure 2 As shown, the present invention provides a location method for a partial discharge fault location device in a power distribution network, including step 1. Data acquisition: synchronously and in real time acquiring discharge pulse signals and load change data of the corresponding lines from partial discharge sensors deployed at key nodes of the power distribution network and the monitoring and data acquisition system.

[0058] Step 2. Multidimensional analysis: Perform fusion analysis on the received discharge pulse signals and load change data.

[0059] Step 3. Intelligent source tracing: Based on the preset load association judgment rules, determine the correlation strength between discharge activities and load events, and generate initial source tracing hypotheses for specific power supply branches.

[0060] Step 4. Targeted Verification: For the power supply branch pointed to by the initial tracing hypothesis, the spatiotemporal signal data of all monitoring points in the branch are called to perform propagation path analysis in order to verify the hypothesis and locate the fault to a specific equipment node in the power supply branch.

[0061] Step 5. Report Generation: Generate and output a structured report containing the specific faulty device node and correlation criteria.

[0062] The examples described in this invention are not limited to the specific embodiments listed above. The examples are merely illustrative to facilitate understanding of the invention and do not constitute a limitation on the scope of protection of this invention. Any modifications, equivalent substitutions, etc., made within the spirit and principles of this invention should be included within the scope of protection.

[0063] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A partial discharge fault location device for a power distribution network, characterized in that, Includes the following modules: Data acquisition module: used to synchronously and in real time acquire discharge pulse signals and corresponding line load change data from partial discharge sensors and monitoring and data acquisition systems deployed at key nodes of the distribution network; Multi-dimensional analysis module: used to perform fusion analysis on the received discharge pulse signal and the corresponding line load change data; Intelligent tracing module: used to determine the correlation strength between discharge activities and load events according to preset load correlation judgment rules, and generate initial tracing assumptions for specific power supply branches; Targeted verification module: Used to analyze the propagation path of the power supply branch pointed to by the initial tracing hypothesis by calling the spatiotemporal signal data of all monitoring points of the branch, so as to verify the hypothesis and locate the fault to the specific equipment node in the power supply branch; Report generation module: Used to generate and output structured reports containing specific faulty device nodes and correlation criteria.

2. The partial discharge fault location device for a power distribution network according to claim 1, characterized in that, The fusion analysis of the received discharge pulse signal and the corresponding line load change data includes: Discharge feature extraction unit: used to identify and extract the timing pattern and intensity variation trend of the discharge pulse from the received discharge pulse signal, forming feature information characterizing the discharge activity; Correlation analysis unit: Used to compare discharge activities with specific load events identified from the load change data of the corresponding line to form correlation information between discharge activities and load events.

3. The partial discharge fault location device for a power distribution network according to claim 2, characterized in that, The specific method for forming the characteristic information representing the discharge activity is as follows: The received discharge pulse signal is filtered, and an adaptive dynamic threshold algorithm based on signal statistical characteristics is used to identify and extract the effective discharge pulse signal. The effective discharge pulse signal is clustered according to the amplitude, polarity and phase characteristics of the discharge pulse signal, and the effective discharge pulse signal representing the unified discharge power source is divided into a cluster. By extracting the statistical characteristics and time sequence characteristics of each cluster of effective discharge pulse signals per unit time, characteristic information representing the discharge activity is formed.

4. The partial discharge fault location device for a power distribution network according to claim 2, characterized in that, The specific method for obtaining the correlation information between the discharge activity and the load event is as follows: If the load change amplitude in the received load data is greater than the preset warning load change amplitude threshold, it is determined as a load event, and its load change rate, load change amplitude and duration are recorded. Each load event is statistically obtained, and each identified load event is aligned with each cluster discharge activity obtained through cluster analysis on the time axis. For each load event, the feature information of each cluster discharge activity within a preset time correlation window is extracted. The Pearson correlation coefficient algorithm is used to calculate the correlation coefficient between the feature data of each load event and the feature information of each cluster discharge activity. If the correlation coefficient between the feature data of a load event and the feature information of a discharge activity is greater than the preset correlation threshold, it is determined that there is a correlation between the load event and the discharge activity. The feature data of the load event and the feature information of the corresponding discharge activity are integrated to form the correlation information between the discharge activity and the load event.

5. A partial discharge fault location device for a power distribution network according to claim 1, characterized in that, The preset load association determination rules include correlation strength rules and time series matching rules, wherein the correlation strength rule means that the correlation coefficient between the characteristic data of the load event and the characteristic information of the discharge activity must be greater than the preset association threshold. The timing matching rule means that the absolute value of the time difference between the occurrence time of discharge activity characteristic information and the occurrence time of load event must be less than the preset timing tolerance threshold.

6. A partial discharge fault location device for a power distribution network according to claim 5, characterized in that, The specific method for determining the correlation strength between discharge activity and load events is as follows: When a discharge activity and a load event both satisfy the correlation strength rule and the time sequence matching rule, then the discharge activity and the load event are determined to be strongly correlated. If a discharge activity and a load event satisfy only one of the correlation strength rule and the timing matching rule, then the discharge activity and the load event are determined to be moderately correlated. If neither the correlation strength rule nor the timing matching rule is satisfied between a certain discharge activity and a certain load event, then the discharge activity and the load event are determined to be weakly correlated.

7. A partial discharge fault location device for a power distribution network according to claim 6, characterized in that, The specific method for generating the initial source tracing hypothesis for a specific power supply branch is as follows: When it is determined that there is a strong correlation between the discharge activity and the load event, the power supply branch where the discharge activity occurred is directly taken as the initial source tracing assumption. When it is determined that there is a moderate correlation between the discharge activity and the load event, the matching degree analysis is performed on the characteristic information of the current discharge activity and the typical fault development mode in the historical discharge characteristic database of the branch. At the same time, the consistency verification of the type attribute of the current load event and the change trend of the historical load event is combined. If the matching degree between the current discharge activity characteristics and the historical fault development mode is greater than the preset high matching threshold and the type attribute of the current load event is consistent with the change trend of the historical load event, then the power supply branch corresponding to the discharge activity is taken as the initial source hypothesis; otherwise, no initial source hypothesis for the specific power supply branch is generated. When the discharge activity is determined to be weakly correlated with the load event, no initial source tracing assumptions for the specific power supply branch are generated.

8. A partial discharge fault location device for a power distribution network according to claim 7, characterized in that, The specific method for verifying this hypothesis and locating the fault to a specific device node in the power supply branch is as follows: Based on the specific power supply branch used as the initial source tracing hypothesis, the spatiotemporal signal data of the power supply branch is obtained by all monitoring points deployed along the power supply branch. The timestamps corresponding to the same effective discharge pulse signal recorded by each monitoring point are compared by the signal arrival time difference positioning method to obtain the timestamp difference between each pair of monitoring points. The propagation speed of electromagnetic waves in the cable is extracted from the local database. The difference in timestamps between each pair of monitoring points is multiplied by the propagation speed to obtain the corresponding relative distance difference. Based on this, a hyperbola positioning algorithm is used to construct a hyperbola with each pair of monitoring points as the focus. Several hyperbolas are obtained by statistical analysis. The least squares method is used to optimize the intersection area of ​​several hyperbolas and locate the fault point to the specific equipment node in the power supply branch.

9. A partial discharge fault location device for a power distribution network according to claim 8, characterized in that, The specific method for generating and outputting a structured report containing the specific faulty device node and correlation criteria is as follows: The system integrates the identified fault-specific device node information and the generated correlation criteria. The fault-specific device node information includes the device number, location coordinates, and the power supply branch to which it belongs. The generated correlation criteria include the correlation information between discharge activities and load events, the correlation strength, the characteristic information of discharge activities, and the corresponding load event characteristic data. The system then uses a built-in report template engine to structurally encapsulate the identified fault-specific device node information and the generated correlation criteria, automatically generates a standardized report, and outputs the complete structured report to the power distribution network monitoring system.

10. A method for locating a partial discharge fault in a power distribution network according to any one of claims 1-9, characterized in that, include: Step 1. Data Acquisition: Simultaneously and in real time acquire discharge pulse signals and corresponding line load change data from partial discharge sensors deployed at key nodes of the distribution network and the monitoring and data acquisition system; Step 2. Multi-dimensional analysis: Perform fusion analysis on the received discharge pulse signals and load change data; Step 3. Intelligent source tracing: Based on the preset load association judgment rules, determine the correlation strength between discharge activities and load events, and generate initial source tracing hypotheses for specific power supply branches; Step 4. Targeted verification: For the power supply branch pointed to by the initial tracing hypothesis, the spatiotemporal signal data of all monitoring points of the branch are called to perform propagation path analysis in order to verify the hypothesis and locate the fault to the specific equipment node in the power supply branch; Step 5. Report Generation: Generate and output a structured report containing the specific faulty device node and correlation criteria.

Images