Method and system for diagnosing high-impedance fault of power distribution network

Abstract

A method and system for diagnosing a high-impedance fault of a power distribution network. The method comprises: collecting voltages and currents of feeder lines and branch nodes in a power distribution network, converting the voltages and currents into three-phase and zero-sequence V-I curve images by means of an artificial intelligence algorithm model, using the three-phase and zero-sequence V-I curve images as dual features to perform iterative computation, identifying fault occurrence, fault type, and fault cause, converting the voltages and currents into a fault phase V-I curve image by means of the artificial intelligence algorithm model and then verticalizing the fault phase V-I curve image into a verticalized V-I curve image, using the fault phase V-I curve and the verticalized V-I curve image as dual features to perform iterative computation, locating a fault section and a fault distance, and generating a corresponding fault handling scheme on the basis of a computation result. The method and system can effectively identify a high-impedance fault, and provide accurate fault location information, and stability and reliability in actual operation.

Description

Methods and systems for high-resistance fault diagnosis in distribution networks Technical Field

[0001] This disclosure relates to the field of power grid operation and maintenance technology, and specifically to a method and system for high-resistance fault diagnosis in distribution networks based on deep learning algorithm models, capable of identifying the causes of high-resistance faults. Background Technology

[0002] The power distribution network is a system composed of branches that transmits electricity to the final consumption units, and it is widespread in rural areas and cities. As the final end of the power system, the power distribution network is directly connected to users. Therefore, rapid fault diagnosis and minimizing repair time are prerequisites for providing users with a high-quality power supply, which is crucial to the production and daily life of thousands of households.

[0003] High-resistance faults are a common type of fault in distribution networks. They refer to faults caused by factors such as insulation aging, animal intrusion, sand and vegetation disturbance, and natural disasters, resulting in an increase in the line's potential to ground. Because high-resistance faults are usually subtle and involve small fault currents, they are unlikely to trigger line protection devices to trip, making them difficult to detect and prevent. However, high-resistance faults can generate electric arcs, which are extremely dangerous and can easily lead to electrical fires and high-voltage electric shocks.

[0004] Existing power grid fault diagnosis and identification methods based on algorithm models, such as extracting features from current waveforms and then using set thresholds to determine whether a high-resistance fault has occurred, mainly focus on building models to determine whether a high-resistance fault has occurred. There is no technology to determine the cause of high-resistance faults, which cannot effectively alert distribution network operation and maintenance personnel to the on-site situation and is not conducive to maintenance personnel arranging emergency repair work.

[0005] Existing fault location methods for power distribution networks mainly use steady-state methods for fault segment location and fault ranging, which have low accuracy and poor disturbance rejection capabilities. Furthermore, existing fault ranging methods relying on artificial intelligence employ ANN models, which have few layers and shallow networks. They have not been tested on large datasets, fail to leverage the advantages of deep learning, and cannot accurately locate fault positions, hindering maintenance personnel from carrying out emergency repairs. Summary of the Invention

[0006] The purpose of this disclosure is to provide a method and system for high-resistance fault diagnosis in distribution networks, which can effectively reflect fault characteristics, establish a mapping relationship with fault characteristics, and fully develop and utilize artificial intelligence methods in fault location research to solve the problems existing in the prior art.

[0007] To solve the above-mentioned technical problems, the embodiments of this disclosure adopt the following technical solutions:

[0008] This disclosure provides a method for diagnosing high-resistance faults in a distribution network, including the following steps:

[0009] Collect raw characteristic data from the power distribution network.

[0010] The original feature data is processed into first feature data and second feature data, and the first feature data and second feature data are input into a first model to be transformed into a first image and a second image. The first model is used to calculate the first image and the second image to identify the occurrence of the fault, the type of the fault, and the cause of the fault.

[0011] The original feature data is converted into a third image, and then the third image is verticalized into a fourth image. The third image and the fourth image are input into a second model for calculation to locate the fault section. The third image and the fourth image are input into a third model for calculation to locate the fault distance.

[0012] The nature of the fault is determined based on the fault type and the fault cause; the precise location of the fault is determined based on the fault section and the fault distance; and a corresponding fault handling plan is generated based on the fault nature and the precise location of the fault.

[0013] In some embodiments, a corresponding fault handling plan is generated based on the fault nature and the precise location of the fault, including:

[0014] Based on the nature of the fault and the precise location of the fault, a corresponding reclosing command is generated to block and repair the fault point;

[0015] After the power distribution network successfully recloses the circuit according to the reclosing command, a corresponding notification message is sent to the operation and maintenance personnel to notify them to verify the information.

[0016] In some embodiments, the causes of failure include permanent failures and transient failures.

[0017] Based on the nature of the fault and its precise location, a corresponding reclosing command is generated to block and repair the fault point, including:

[0018] If the cause of the fault is a permanent fault, the relevant circuit breaker in the protection device at the precise location of the fault is controlled to trip faster by the reclosing command.

[0019] If the fault is transient, the circuit breaker is controlled to reclose after tripping by the reclosing command.

[0020] In some embodiments, the original feature data includes: denoised and preprocessed voltage waveform data of the feeder, current waveform data of the feeder, voltage waveform data of the branch node, and current waveform data of the branch node.

[0021] And / or, the fault types include single-phase short-circuit ground fault, single-phase open-circuit ground fault, two-phase short-circuit ground fault, two-phase phase-to-phase fault, and three-phase short-circuit fault.

[0022] In some embodiments, the first feature data is obtained from the original feature data through a three-phase transformation, and the first image is a three-phase VI curve image; the second feature data is obtained from the original feature data through a zero-sequence transformation, and the second image is a zero-sequence VI curve image.

[0023] The first model is used to calculate the first image and the second image to identify the occurrence of the fault, the type of the fault, and the cause of the fault, including: normalizing the three-phase VI curve image and the zero-sequence VI curve image to obtain the first dual feature.

[0024] Based on the first dual features, iterative calculations are performed to extract features used to identify whether a fault has occurred, determine the fault type, and identify the cause of the fault.

[0025] In some embodiments, the third image is a fault phase VI curve image obtained from the original feature data, and the fourth image is a verticalized VI curve generated by performing a verticalization operation on the fault phase VI curve image.

[0026] The third and fourth images are input into the second model for calculation to locate the faulty section, including:

[0027] The fault phase VI curve image and the verticalized VI curve are normalized to obtain the second dual feature;

[0028] Based on the second dual feature, iterative calculations are performed to extract features used to identify faulty sections;

[0029] The third image and the fourth image are input into the third model for calculation to locate the fault distance, including:

[0030] The fault phase VI curve image and the verticalized VI curve are normalized to obtain the third dual feature;

[0031] Based on the third dual feature, iterative calculations are performed to extract features used to identify the fault distance.

[0032] In some embodiments, the construction of the first model includes: obtaining a fault sample set of three-phase and zero-sequence VI curves at the feeder end; inputting the fault sample set of three-phase and zero-sequence VI curves at the feeder end into an initial first model for training to obtain the first model.

[0033] In some embodiments, the construction of the second model includes: obtaining a fault sample set of the VI curve of the faulty phase at the feeder end; inputting the fault sample set of the VI curve of the faulty phase at the feeder end into the initial second model for training to obtain the second model.

[0034] The construction of the third model includes: a fault sample set of VI curves of branch node fault phases; inputting the fault sample set of VI curves of branch node fault phases into the initial third model for training to obtain the third model.

[0035] In some embodiments, the method further includes:

[0036] Store the original data and the calculation results of the first model, the second model, and the third model; add the original data and the calculation results of the first model, the second model, and the third model to the sample set used to train the first model, the second model, and the third model.

[0037] Another aspect of the embodiments of this disclosure provides a system for diagnosing high-resistance faults in a distribution network, comprising:

[0038] Data acquisition module: Used to collect raw characteristic data in the power distribution network.

[0039] Fault identification module: used to process the original feature data into first feature data and second feature data, input the first model to convert it into first image and second image for calculation, and identify the occurrence of fault, fault type and fault cause.

[0040] Fault location module: It is used to convert the original feature data into a third image, then verticalize the third image into a fourth image, input the third image and the fourth image into a second model for calculation to locate the fault section, and then input the third image and the fourth image into a third model for calculation to locate the fault distance.

[0041] Fault Repair Module: Used to determine the nature of the fault based on the fault type and the fault cause, determine the precise location of the fault based on the fault section and the fault distance, send a reclosing command, and notify the operation and maintenance personnel to verify after successful reclosing.

[0042] Data storage module: Stores the original data and the calculation results of the first model, the second model, and the third model, and adds them to the sample set used to train the first model, the second model, and the third model.

[0043] The system for high-resistance fault diagnosis in distribution networks can implement the method for high-resistance fault diagnosis in distribution networks according to any one of claims 2 to 8.

[0044] This disclosure provides a method and system for diagnosing high-resistance faults in distribution networks. It enables fault occurrence judgment, fault cause identification, fault type identification, fault section location, fault distance measurement, and fault recovery. This helps power system operation and maintenance personnel to more comprehensively understand and handle fault situations, improving the efficiency and accuracy of fault handling. The system integrates a fault identification model, a fault section identification model, and a fault distance measurement model. Utilizing VI curves obtained through different processing methods, it can fully and accurately reflect the fault cause, type, section, and distance. Compared with existing technologies, this method and system have more systematic functionality.

[0045] For high-resistance faults, this model can effectively identify the cause, type, section, and distance of the fault, preventing protection from failing to trip due to weak high-resistance fault characteristics, thus avoiding disasters such as electrical fires. Simultaneously, the model can help to promptly locate high-resistance fault points and provide accurate fault information, thereby avoiding potential catastrophic consequences. For non-high-resistance faults, the model has an even more accurate identification effect. This model system can effectively cope with interference factors such as load imbalance, line parameter errors, and transformer errors, ensuring the stability and reliability of the model system in actual operation and improving the accuracy of fault diagnosis and identification. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this disclosure 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 recorded in this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 is a schematic diagram of the structure of the power distribution network fault diagnosis system according to a disclosed embodiment;

[0048] Figure 2 is a schematic diagram of the steps of the power distribution network fault diagnosis method according to the disclosed embodiment;

[0049] Figure 3 is a schematic diagram of the structure of the first model in the power distribution network fault diagnosis method of the disclosed embodiment;

[0050] Figure 4 is a schematic diagram of the structure of the second model in the power distribution network fault diagnosis method of the disclosed embodiment;

[0051] Figure 5 is a schematic diagram of the structure of the third model in the power distribution network fault diagnosis method of the disclosed embodiment;

[0052] Figure 6 is a vertical VI curve diagram of the disclosed embodiment;

[0053] Figure 7 is a single-outgoing-line multi-branch distribution network model of the disclosed embodiment;

[0054] Figure 8 is a multi-outgoing-line, multi-branch distribution network model of the disclosed embodiment. Detailed Implementation

[0055] Various embodiments and features of this disclosure are described herein with reference to the accompanying drawings.

[0056] It should be understood that various modifications can be made to the embodiments described herein. Therefore, the above description should not be considered as limiting, but merely as an example of embodiments. Other modifications within the scope and spirit of this disclosure will be apparent to those skilled in the art.

[0057] The accompanying drawings, which are included in and form part of this specification, illustrate embodiments of the present disclosure and, together with the general description of the disclosure given above and the detailed description of the embodiments given below, serve to explain the principles of the disclosure.

[0058] These and other features of this disclosure will become apparent from the following description of preferred forms of embodiments given as non-limiting examples, with reference to the accompanying drawings.

[0059] It should also be understood that although this disclosure has been described with reference to some specific examples, those skilled in the art can certainly implement many other equivalent forms of this disclosure, which have the features described in the claims and are therefore all within the scope of protection defined herein.

[0060] The above and other aspects, features and advantages of this disclosure will become more apparent when taken in conjunction with the accompanying drawings and in view of the following detailed description.

[0061] Specific embodiments of this disclosure are described thereafter with reference to the accompanying drawings; however, it should be understood that the claimed embodiments are merely examples of this disclosure, which may be implemented in various ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that could obscure this disclosure. Therefore, the specific structural and functional details claimed herein are not intended to be limiting, but merely to serve as the basis and representative basis for the claims to teach those skilled in the art to use this disclosure in a variety of substantially any suitable detailed structures.

[0062] This specification may use the phrases “in one embodiment,” “in another embodiment,” “in yet another embodiment,” or “in still another embodiment,” all of which may refer to one or more of the same or different embodiments according to this disclosure.

[0063] The first embodiment of this disclosure provides a power distribution network fault diagnosis method, applied to the power distribution network fault diagnosis system shown in Figure 1. Referring to Figure 2, the power distribution network fault diagnosis method includes the following steps:

[0064] S1: Data acquisition module 1 collects raw characteristic data from the power distribution network.

[0065] In this step, the raw characteristic data collected by the data acquisition module 1 in the distribution network are the voltage and current of the feeders and the voltage and current waveform data of the branch nodes. That is, it includes: voltage waveform data of the feeders, current waveform data of the feeders, voltage waveform data of the branch nodes, and current waveform data of the branch nodes. All of the above data need to be denoised and preprocessed. The subsequent steps will use the raw characteristic data for calculation.

[0066] Preferably, in this step, the voltage recording data of the feeder, the current recording data of the feeder, the voltage recording data of the branch node, and the current recording data of the branch node are extracted in real time in a data window with a length of two cycles, and then denoised and preprocessed. More preferably, the two-cycle signal after the fault time is extracted, the signal is processed using a low-pass filter, the high-frequency cutoff frequency is 150Hz, and the secondary signal is converted into a primary signal through the transformer ratio.

[0067] S2: The fault identification module 2 processes the original feature data into first feature data and second feature data, and inputs the first feature data and second feature data into the first model to convert them into a first image and a second image. The first model calculates the first image and the second image to identify the fault occurrence, fault type and fault cause.

[0068] Specifically, in this step, the first feature data is obtained from the original feature data through three-phase transformation, and the first image is a three-phase VI curve image. That is, the original feature data is transformed into the form of three-phase current and three-phase voltage, and then curve images are generated on the vertical coordinate axis with the three-phase voltage and three-phase current as coordinate points respectively.

[0069] The second feature data is obtained by zero-sequence transformation of the original feature data. The second image is a zero-sequence VI curve image, that is, the original feature data is transformed into the form of zero-sequence current and zero-sequence voltage, and then curve images are generated on the vertical coordinate axis with zero-sequence voltage and zero-sequence current as coordinate points respectively.

[0070] The first model is used to identify whether a fault has occurred, and if so, to determine the type and cause of the fault.

[0071] Specifically, as shown in Figure 3, the first model includes an input layer, an intermediate layer, and an output layer.

[0072] The input layer converts the collected feeder terminal and node voltage and current data into three-phase VI curve images and zero-sequence VI curve images and normalizes them. The normalized three-phase VI curve images and zero-sequence VI curve images are then used as dual image features input to the intermediate layer.

[0073] Preferably, the three-phase VI curve image is normalized using the following method: limiting the x-range of the coordinates to [-X, X] and the y-range to [-Y, Y], with each phase curve distinguished by a different color. Preferably, X = 400A and Y = 10000V. The zero-sequence VI curve image is normalized using the following method: using the maximum absolute values ​​of voltage and current as reference values ​​to obtain a normalized curve, with both x and y ranging from [-1, 1]. The RGB values ​​of the curve color are [r, r, r], where the value of r is measured by the magnitude of the zero-sequence current. The zero-sequence current magnitude range [-Imax, Imax] corresponds to RGB values ​​[0, 0, 0] to [255, 255, 255]. Preferably, Imax = 100A.

[0074] The intermediate layer consists of a convolutional layer and a channel attention mechanism. Specifically, each of the two residual blocks forms two pathways, which extract two image features respectively, convert them into feature maps, and merge them into a comprehensive feature map set according to the channel direction. Then, the comprehensive feature map set features are extracted by the three residual block structures.

[0075] The output layer consists of an average pooling layer, a fully connected layer, a softmax layer, and a classification layer. It can determine whether a fault has occurred based on the calculation results of the intermediate layers, and if a fault is determined to have occurred, it can determine the fault type and the cause of the fault.

[0076] Specifically, the fault types include: single-phase short-circuit ground fault, single-phase open-circuit ground fault, two-phase short-circuit ground fault, two-phase phase-to-phase fault, and three-phase short-circuit fault.

[0077] The causes of the faults include wet sand and gravel faults, dry sand and gravel faults, tree-related faults, animal-related faults, lightning strike faults, and wildfire faults. Among these, wet sand and gravel faults, dry sand and gravel faults, tree-related faults, animal-related faults, and volcanic faults are permanent faults, while lightning strike faults are considered transient faults. It is worth noting that, in practical terms, wet sand and gravel faults, dry sand and gravel faults, tree-related faults, and animal-related faults fall under the category of high-resistance grounding faults.

[0078] S3: The fault location module 3 converts the original feature data into a third image, then verticalizes the third image into a fourth image, and inputs the third image and the fourth image into the second model for calculation to locate the fault section; the third image and the fourth image are input into the third model for calculation to locate the fault distance.

[0079] Specifically, in this step, the original feature data is transformed into a third image, and then the third image is verticalized into a fourth image. The third image refers to the fault phase VI curve image generated by using the voltage and current of the original feature data as coordinate points, and the fourth image refers to the verticalized VI curve generated by performing a verticalization operation on the fault phase VI curve image.

[0080] In a specific embodiment, referring to Figure 6, the fault phase VI curve is generated by fitting an ellipse to the fault phase VI curve. The verticalization method is to rotate the fault phase VI curve counterclockwise by 90°-θ angle, and divide the curve horizontally into four parts with the same height and width. These parts are then merged from left to right and from top to bottom in a 2×2 matrix to form a verticalized VI curve image.

[0081] The advantage of verticalizing the curve is that it can better highlight the curve's changing characteristics in the coordinate graph, and can avoid the curve's changing characteristics being obscured by the reduction in image dimensions due to small changes in features.

[0082] The second model is used for fault segment identification, that is, to determine the location of the fault segment.

[0083] Furthermore, when it is determined that the fault occurred in a certain section of the outgoing line, the fault phase VI curve and the verticalized VI curve at the upstream branch point of that section are taken as dual features and input into the fault ranging model to measure the fault distance; when it is determined that the fault occurred at a branch point of the outgoing line, the fault phase VI curve and the verticalized VI curve at that branch point are taken as dual features and input into the fault ranging model to measure the fault distance.

[0084] Furthermore, referring to Figure 4, the second model includes an input layer, an intermediate layer, and an output layer.

[0085] The input layer converts the collected feeder-end fault phase voltage and current data into fault phase VI curve images and normalizes them. The normalization method can refer to the normalization method of the three-phase VI curve image and the zero-sequence VI curve image in step 2, which will not be repeated here.

[0086] The intermediate layer consists of convolutional layers, preferably composed of 5 residual structure blocks, to extract image features.

[0087] The output layer consists of an average pooling layer, a fully connected layer, a softmax layer, and a classification layer to identify faulty sections or branch points.

[0088] The third model is used to determine the distance to the fault.

[0089] In a specific embodiment, referring to Figure 5, the third model includes an input layer, an intermediate layer, and an output layer.

[0090] The input layer converts the branch point voltage and current data into a fault phase VI curve image and normalizes it as the first feature. The normalized fault phase VI curve is then verticalized and divided into four parts along the y-axis and spliced ​​together to form a vertical VI curve image as the second feature. The normalization method can refer to the normalization method of the three-phase VI curve image and the zero-sequence VI curve image in step 2, which will not be repeated here.

[0091] The intermediate layer consists of a convolutional layer and a channel attention mechanism. Preferably, each channel consists of 8 residual structural blocks forming two pathways, which extract two image features respectively, convert them into feature map sets, and merge them into a comprehensive feature map set according to the channel direction.

[0092] The output layer consists of an average pooling layer, a fully connected layer, and a regression layer, and outputs the distance to the fault branch point.

[0093] By combining the second model to locate the faulty section and the third model to locate the distance to the faulty branch point, the fault location can be accurately determined.

[0094] S4: The fault recovery module 4 determines the nature of the fault based on the fault type and the fault cause, determines the precise location of the fault based on the fault section and the fault distance, and generates a corresponding fault handling plan based on the fault nature and the precise location of the fault.

[0095] Specifically, the fault recovery module determines the nature of the fault based on the fault type and cause output by the fault identification module 2 in step S2, and outputs a fault reclosing command based on the fault nature and fault section. The corresponding protection device executes protection actions according to the reclosing command, and the fault type, cause, nature, section, and distance are displayed visually for maintenance personnel to verify. This allows for the rapid interruption of high-resistance faults in the distribution network, preventing them from generating arcs that could cause fires or electric shocks.

[0096] Specifically, generating a corresponding fault handling plan based on the nature of the fault and the precise location of the fault also includes:

[0097] For the faulty section, a reclosing decision instruction is given. After the line circuit breaker trips the corresponding phase, the upstream circuit breaker of the faulty section is accelerated for reclosing. After the reclosing time T1, preferably 0.8s, the reclosing faulty section is close to the power supply side. If it is identified as a permanent fault, the three-phase circuit breaker is accelerated to trip. If it is identified as a transient fault, the circuit breaker reclosing is successful.

[0098] For branch lines, a reclosing decision instruction is given; after the corresponding phase of the line circuit breaker trips, the non-faulty line section and the non-faulty branch line circuit breaker are reclosed at an accelerated speed. After a reclosing time T1, preferably 0.8s, the branch line circuit breaker is reclosed. If a permanent fault is identified, the three-phase circuit breaker trips at an accelerated speed; if a transient fault is identified, the circuit breaker reclosing is successful.

[0099] Preferably, the data collected by the data acquisition module and the results calculated by the fault identification module and the fault location module are used as historical samples by the data storage module 5 to expand the dataset, and the dataset is used to train the first model, the second model and the third model.

[0100] Specifically, data storage module 5 stores the information and reclosing instructions extracted by the data acquisition module after the fault is recovered, as historical reclosing samples to expand the sample set. When a sample set of a certain size is accumulated, the parameters of fault identification module 2 and fault location module 3 are updated.

[0101] A second embodiment of this disclosure provides an initial construction method for a first model, a second model, and a third model, including:

[0102] Three sample sets of feeder and branch node faults are obtained. Specifically, the sample sets include: a sample set of three-phase and zero-sequence VI curve faults at the feeder end for the first training and testing; a sample set of phase VI curve faults at the feeder end for the second model training and testing; and a sample set of phase VI curve faults at the branch node for the third training and testing.

[0103] The three fault sample sets obtained are respectively input into the first model, the second model, and the third model for training until convergence, to obtain the trained first model, the second model, and the third model.

[0104] The three fault sample sets are generated through a simulated power distribution network model, as shown in Figures 7 and 8. The generation steps include:

[0105] Fault sample data were obtained for use as samples. The causes of the faults included wet sand and gravel high-resistance grounding faults, dry sand and gravel high-resistance grounding faults, tree-touching high-resistance grounding faults, animal-touching high-resistance grounding faults, wildfire faults, and lightning strike faults. The fault distance was randomly set on the line section or branch line.

[0106] Based on the acquired fault sample data, construct the corresponding three-phase VI curve, zero-sequence VI curve, fault phase VI curve, and verticalized VI curve, and perform normalization processing.

[0107] The three-phase VI curves and the zero-sequence VI curves are used to form the first sample set for the first model.

[0108] The fault phase VI curves and verticalized VI curves of the feeder are used to form a second sample set for the second model.

[0109] The fault phase VI curves and verticalized VI curves of the branch nodes are used to form the third sample set for the third model.

[0110] In summary, the embodiments of this disclosure provide a method and system for high-resistance fault diagnosis in distribution networks. This system enables fault occurrence judgment, fault cause identification, fault type identification, fault section location, fault distance measurement, and fault recovery. It helps power system operation and maintenance personnel to more comprehensively understand and handle fault situations, improving the efficiency and accuracy of fault handling. The system integrates a fault identification model, a fault section identification model, and a fault distance measurement model. Utilizing VI curves obtained through different processing methods, it can fully and accurately reflect the fault cause, type, section, and distance. Compared with existing technologies, this method and system have more systematic functionality.

[0111] For high-resistance faults, this model can effectively identify the cause, type, section, and distance of the fault, preventing protection from failing to trip due to weak high-resistance fault characteristics, thus avoiding disasters such as electrical fires. Simultaneously, the model can help to promptly locate high-resistance fault points and provide accurate fault information, thereby avoiding potential catastrophic consequences. For non-high-resistance faults, the model has an even more accurate identification effect. This model system can effectively cope with interference factors such as load imbalance, line parameter errors, and transformer errors, ensuring the stability and reliability of the model system in actual operation and improving the accuracy of fault diagnosis and identification.

[0112] The aforementioned storage medium may be included in the aforementioned electronic device; or it may exist independently and not be assembled into the electronic device.

[0113] The aforementioned storage medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: acquire at least two Internet Protocol (IP) addresses; send a node evaluation request, including at least two IP addresses, to a node evaluation device, wherein the node evaluation device selects an IP address from the at least two IP addresses and returns it; and receive the IP address returned by the node evaluation device; wherein the acquired IP address indicates an edge node in the content delivery network.

[0114] Alternatively, the storage medium may carry one or more programs that, when executed by the electronic device, cause the electronic device to: receive a node evaluation request including at least two Internet Protocol (IP) addresses; select an IP address from the at least two IP addresses; and return the selected IP address; wherein the received IP address indicates an edge node in the content delivery network.

[0115] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including but not limited to object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the passenger's computer, partially on the passenger's computer, as a standalone software package, partially on the passenger's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer can be connected to the passenger's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0116] It should be noted that the storage medium described in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any storage medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the storage medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0117] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0118] The units described in the embodiments of this disclosure can be implemented in software or hardware. The names of the units are not, in some cases, intended to limit the specific unit.

[0119] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0120] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0121] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

[0122] Furthermore, while the operations are described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in a sequential order. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0123] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

[0124] The foregoing has provided a detailed description of several embodiments of this disclosure. However, this disclosure is not limited to these specific embodiments. Those skilled in the art can make various variations and modifications based on the concept of this disclosure, and all such variations and modifications should fall within the scope of protection claimed by this disclosure.

Claims

1. A method for diagnosing high-resistance faults in power distribution networks, characterized in that, Includes the following steps: Collect raw characteristic data from the power distribution network; The original feature data is processed into first feature data and second feature data, and the first feature data and second feature data are input into a first model to be transformed into a first image and a second image. The first model is used to calculate the first image and the second image to identify the occurrence of the fault, the type of the fault, and the cause of the fault. The original feature data is converted into a third image, and then the third image is verticalized into a fourth image. The third image and the fourth image are input into a second model for calculation to locate the fault section. The third image and the fourth image are input into a third model for calculation to locate the fault distance. The nature of the fault is determined based on the fault type and the fault cause; the precise location of the fault is determined based on the fault section and the fault distance; and a corresponding fault handling plan is generated based on the fault nature and the precise location of the fault.

2. The method for high-resistance fault diagnosis in distribution networks according to claim 1, characterized in that, Based on the nature of the fault and its precise location, a corresponding fault handling plan is generated, including: Based on the nature of the fault and the precise location of the fault, a corresponding reclosing command is generated to block and repair the fault point; After the power distribution network successfully recloses the circuit according to the reclosing command, a corresponding notification message is sent to the operation and maintenance personnel to notify them to verify the information.

3. The method for high-resistance fault diagnosis in distribution networks according to claim 2, characterized in that, The causes of the failures include permanent failures and transient failures. Based on the nature of the fault and its precise location, a corresponding reclosing command is generated to block and repair the fault point, including: If the cause of the fault is a permanent fault, the relevant circuit breaker in the protection device at the precise location of the fault is controlled to trip faster by the reclosing command. If the fault is transient, the circuit breaker is controlled to reclose after tripping by the reclosing command.

4. The method for high-resistance fault diagnosis in distribution networks according to claim 1, characterized in that, The original feature data includes: voltage waveform data of the feeder after noise reduction and preprocessing, current waveform data of the feeder, voltage waveform data of the branch node, and current waveform data of the branch node. and / or The fault types include single-phase short-circuit ground fault, single-phase open-circuit ground fault, two-phase short-circuit ground fault, two-phase phase-to-phase fault, and three-phase short-circuit fault.

5. The method for high-resistance fault diagnosis in distribution networks according to claim 1, characterized in that, The first feature data is obtained from the original feature data through a three-phase transformation, and the first image is a three-phase VI curve image. The second feature data is obtained from the original feature data through a zero-order transformation, and the second image is a zero-order VI curve image. The first model is used to calculate the first image and the second image to identify the occurrence of the fault, the type of the fault, and the cause of the fault, including: normalizing the three-phase VI curve image and the zero-sequence VI curve image to obtain the first dual feature; Based on the first dual features, iterative calculations are performed to extract features used to identify whether a fault has occurred, determine the fault type, and identify the cause of the fault.

6. The method for high-resistance fault diagnosis in distribution networks according to claim 1, characterized in that, The third image is the fault phase VI curve image obtained from the original feature data, and the fourth image is the vertical VI curve generated by performing a verticalization operation on the fault phase VI curve image. The third and fourth images are input into the second model for calculation to locate the faulty section, including: The fault phase VI curve image and the verticalized VI curve are normalized to obtain the second dual feature; Based on the second dual feature, iterative calculations are performed to extract features used to identify faulty sections; The third image and the fourth image are input into the third model for calculation to locate the fault distance, including: The fault phase VI curve image and the verticalized VI curve are normalized to obtain the third dual feature; Based on the third dual feature, iterative calculations are performed to extract features used to identify the fault distance.

7. The method for high-resistance fault diagnosis in distribution networks according to claim 5, characterized in that, The construction of the first model includes: Obtain a fault sample set of three-phase and zero-sequence VI curves at the feeder end; The fault sample set of the three phases and zero sequence VI curves at the feeder end is input into the initial first model for training to obtain the first model.

8. The method for high-resistance fault diagnosis in distribution networks according to claim 6, characterized in that, The construction of the second model includes: Obtain the fault sample set of the VI curve of the faulty phase at the feeder end; The fault sample set of the VI curve of the faulty phase at the feeder end is input into the initial second model for training to obtain the second model; The construction of the third model includes: Fault sample set of VI curves for branch node failures; The fault sample set of the VI curve of the branch node fault phase is input into the initial third model for training to obtain the third model.

9. The method for diagnosing high-resistance faults in a distribution network according to claim 1, characterized in that, The method further includes: Store the original data and the calculation results of the first model, the second model, and the third model; The original data, along with the calculation results of the first model, the second model, and the third model, are added to the sample set used to train the first model, the second model, and the third model.

10. A system for diagnosing high-resistance faults in power distribution networks, characterized in that, include: Data acquisition module: Used to collect raw characteristic data from the power distribution network. Fault identification module: This module processes the original feature data into first feature data and second feature data, inputs them into a first model to convert them into a first image and a second image for calculation, and identifies the occurrence, type, and cause of the fault. Fault location module: This module converts the original feature data into a third image, then vertically transforms the third image into a fourth image. The third and fourth images are then input into a second model for calculation to locate the fault segment. Finally, the third and fourth images are input into a third model for calculation to determine the fault distance. Fault Repair Module: Used to determine the nature of the fault based on the fault type and cause, pinpoint the exact location of the fault based on the fault section and distance, send a reclosing command, and notify maintenance personnel for verification upon successful reclosing. Data storage module: Stores the original data and the calculation results of the first model, the second model, and the third model, and adds them to the sample set used for training the first model, the second model, and the third model. The system for high-resistance fault diagnosis in distribution networks can implement the method for high-resistance fault diagnosis in distribution networks according to any one of claims 2 to 8.

Images