A method, device, equipment and storage medium for locating fault wave head of a transmission line

Abstract

The present invention discloses a method, device, equipment, and storage medium for locating fault waveheads on power transmission lines, which are used to address the technical problems of existing traveling wave wavehead locating algorithms, such as a high false alarm rate for fault waveforms and low wavehead locating accuracy. The present invention comprises the following steps: collecting first fault waveform data; performing data normalization processing on the first fault waveform data to obtain input data; extracting first latent space variables from the input data; classifying the first latent space variables to obtain second fault waveform data; extracting suspected fault wavehead data from the second fault waveform data; extracting second latent space variables from the suspected fault wavehead data; determining whether the second latent space variables belong to a fault waveform; and if so, identifying the fault wavehead position in the fault waveform by presetting a first recurrent neural network model.

Description

Technical Field

[0001] The present invention relates to the technical field of fault location, and in particular to a method, device, equipment and storage medium for locating a fault wave head of a transmission line. Background Art

[0002] With the development of my country's social economy, the industry has placed higher demands on the reliability and stability of power transmission lines. When a fault occurs, the longer it takes to resolve it, the greater the economic impact. Therefore, quickly locating the fault location is crucial to maximize response time for resolution. Traditional fault location methods rely heavily on manual line inspections, which require significant manpower and material resources, are time-consuming, and inefficient.

[0003] Traveling wave acquisition typically operates at a high frequency, making it susceptible to interference from various noise sources and the generation of clutter. These issues make it extremely difficult to locate the wave head within a traveling wave waveform. Currently, the accuracy of traveling wave fault alarms in the industry is low, often requiring manual intervention to screen suspected fault waveforms. Furthermore, manual experience is required to diagnose the wave head location and manually calibrate it within the system. This method is time-consuming and requires a high level of experience from the operations and maintenance personnel.

[0004] Using machine learning methods for fault location currently presents the following challenges: Using a power frequency waveform as input offers good regularity, enabling high accuracy in fault identification and classification. However, due to the low sampling frequency of the power frequency, errors in wave head positioning can significantly impact the effectiveness of fault location. Therefore, the power frequency is generally only suitable for fault location determination. Using a traveling wave waveform as input, however, often lacks regularity and contains significant noise, making it difficult to achieve high accuracy even with manual judgment. Summary of the Invention

[0005] The present invention provides a method, device, equipment and storage medium for locating a fault wave head of a transmission line, which are used to solve the technical problems of a high false alarm rate for fault waveforms and low wave head positioning accuracy in existing traveling wave wave head positioning algorithms.

[0006] The present invention provides a method for locating a fault wave head of a transmission line, comprising:

[0007] collecting first fault waveform data;

[0008] performing data normalization processing on the first fault waveform data to obtain input data;

[0009] Extracting a first latent space variable of the input data;

[0010] classifying the first latent space variables to obtain second fault waveform data;

[0011] extracting suspected fault wave head data from the second fault waveform data;

[0012] extracting a second latent space variable of the suspected fault wave head data;

[0013] determining whether the second latent space variable belongs to a fault waveform;

[0014] If so, a first recurrent neural network model is preset to identify the fault wave head position in the fault waveform.

[0015] Optionally, the step of collecting first fault waveform data includes:

[0016] Collect power frequency current waveform data and obtain classic fault waveform data;

[0017] Calculating the similarity between the power frequency current waveform data and the classic fault waveform data;

[0018] When the similarity reaches a preset similarity threshold, the power frequency current waveform data is determined as first fault waveform data.

[0019] Optionally, the step of extracting the first latent space variable of the input data includes:

[0020] An autoencoder is used to reduce the length of the input data to obtain a first latent space variable of the input data.

[0021] Optionally, the step of extracting suspected fault wave head data from the second fault waveform data includes:

[0022] Inputting the second fault waveform data into a preset second recurrent neural network and outputting a fault time window;

[0023] The fault time window is matched in the second fault waveform data to obtain suspected fault wave head data.

[0024] Optionally, the step of determining whether the second latent space variable belongs to a fault waveform includes:

[0025] Input the second latent space variable into the preset random forest model and output the classification result;

[0026] It is determined whether the second latent space variable belongs to a fault waveform according to the classification result.

[0027] The present invention also provides a transmission line fault wave head positioning device, comprising:

[0028] An acquisition module, configured to acquire first fault waveform data;

[0029] a normalization module, configured to perform data normalization processing on the first fault waveform data to obtain input data;

[0030] A first latent space variable extraction module, used to extract the first latent space variable of the input data;

[0031] a classification module, configured to classify the first latent space variables and obtain second fault waveform data;

[0032] A suspected fault wave front data extraction module, configured to extract suspected fault wave front data from the second fault waveform data;

[0033] A second latent space variable extraction module, used to extract the second latent space variables of the suspected fault wave head data;

[0034] a judgment module, configured to judge whether the second latent space variable belongs to a fault waveform;

[0035] The fault wave head position identification module is used to identify the fault wave head position in the fault waveform by presetting a first recurrent neural network model.

[0036] Optionally, the acquisition module includes:

[0037] The acquisition submodule is used to collect power frequency current waveform data and obtain classic fault waveform data;

[0038] A similarity calculation submodule, configured to calculate the similarity between the power frequency current waveform data and the classic fault waveform data;

[0039] The first fault waveform data determining submodule is configured to determine the power frequency current waveform data as first fault waveform data when the similarity reaches a preset similarity threshold.

[0040] Optionally, the first latent space variable extraction module includes:

[0041] The first latent space variable extraction submodule is used to reduce the length of the input data by using an automatic encoder to obtain the first latent space variable of the input data.

[0042] The present invention further provides an electronic device, comprising a processor and a memory:

[0043] The memory is used to store program code and transmit the program code to the processor;

[0044] The processor is configured to execute the transmission line fault wave front location method as described above according to the instructions in the program code.

[0045] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store program code, and the program code is used to execute the transmission line fault wave head locating method as described in any one of the above items.

[0046] The above technical solution shows that the present invention has the following advantages: the present invention collects first fault waveform data; performs data normalization processing on the power frequency current waveform data to obtain input data; extracts the first latent space variable of the input data; classifies the latent space variable to obtain second fault waveform data; extracts suspected fault wave head data from the second fault waveform data; extracts the second latent space variable of the suspected fault wave head data; determines whether the second latent space variable belongs to the fault waveform; and if so, identifies the fault wave head position in the fault waveform by presetting a first recurrent neural network model. This reduces the false alarm rate of the fault waveform and improves the wave head positioning accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0048] Figure 1 A flowchart of a method for locating a fault wave head in a power transmission line provided by an embodiment of the present invention;

[0049] Figure 2 A flowchart of a method for locating a fault wave head in a power transmission line provided by an embodiment of the present invention;

[0050] Figure 3 A process for extracting first latent space variables by an autoencoder provided in an embodiment of the present invention;

[0051] Figure 4 This is a structural block diagram of a transmission line fault wave front locating device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The embodiments of the present invention provide a method, apparatus, device and storage medium for locating a fault wave head of a transmission line, which are used to solve the technical problems of a high false alarm rate for fault waveforms and low wave head positioning accuracy in existing traveling wave wave head positioning algorithms.

[0053] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0054] See also Figure 1 , Figure 1 A flowchart of the steps of a method for locating a fault wave head of a transmission line provided by an embodiment of the present invention.

[0055] The present invention provides a method for locating a fault wave head of a transmission line, which may specifically include the following steps:

[0056] Step 101, collecting first fault waveform data;

[0057] In the embodiment of the present invention, the first fault waveform data refers to the collected data that may contain fault waveforms after preliminary screening.

[0058] Step 102: performing data normalization processing on the first fault waveform data to obtain input data;

[0059] Normalization is a way to simplify calculations, that is, to transform a dimensioned expression into a dimensionless expression, which becomes a scalar.

[0060] After the first fault waveform data is collected, it can be subjected to data normalization processing to summarize the statistical distribution of the first fault waveform data and obtain input data for the automatic encoder.

[0061] In an embodiment of the present invention, min-max normalization can be used to perform data normalization. The process formula is as follows:

[0062]

[0063] Among them, v' is the target vector obtained after normalization, that is, the input data of the autoencoder, v is the input vector to be converted, that is, the first fault waveform data, min(A) is the minimum value of all original features, max(A) is the maximum value of all original features, new_max(A) is the maximum boundary value of the target vector to be generated, and new_min(A) is the minimum boundary value of the target vector to be generated.

[0064] Step 103, extracting the first latent space variable of the input data;

[0065] A latent space variable is a variable in a multidimensional space. The multidimensional features it contains are usually less explanatory, but after a certain mapping, they can be converted into complete multidimensional vector features with real meaning.

[0066] After the input data is obtained, the first latent space variables can be extracted through the autoencoder to achieve dimensionality reduction and noise reduction processing of the input first fault waveform data.

[0067] Step 104: classify the first latent space variables to obtain second fault waveform data;

[0068] In an embodiment of the present invention, the first latent space variables can be classified through a deep neural network, waveforms with a judgment result of non-fault can be filtered out, and waveforms with a judgment result of fault can be retained as second fault waveform data.

[0069] This process can further remove the interference of clutter on fault waveform judgment and improve the recognition rate of fault waveform.

[0070] Step 105, extracting suspected fault wave head data from the second fault waveform data;

[0071] After obtaining the second fault waveform data, the pre-trained recurrent neural network can be used to extract suspected fault wave head data therefrom.

[0072] Step 106, extracting the second latent space variables of the suspected fault wave head data;

[0073] After extracting the suspected fault wave head data, its second latent space variables can be extracted through the autoencoder to achieve dimensionality reduction and noise reduction of the suspected fault wave head data.

[0074] Step 107, determining whether the second latent space variable belongs to a fault waveform;

[0075] Step 108: If yes, identify the fault wave head position in the fault waveform by using a preset first recurrent neural network model.

[0076] Next, the random forest model is used to determine whether the waveform of the second latent space variable is a fault waveform, thereby reducing misjudgments of fault waveforms. When the second latent space variable is a fault waveform, the first recurrent neural network model can be used to identify the location of the fault wave head in the fault waveform.

[0077] A random forest is a classifier that contains multiple decision trees, and the class it outputs is determined by the mode of the classes output by the individual trees.

[0078] In practical applications, identifying the position of the fault wave head helps to locate the fault point.

[0079] The present invention collects first fault waveform data; performs data normalization on the power frequency current waveform data to obtain input data; extracts first latent space variables from the input data; classifies the latent space variables to obtain second fault waveform data; extracts suspected fault wave head data from the second fault waveform data; extracts second latent space variables from the suspected fault wave head data; determines whether the second latent space variables belong to the fault waveform; and if so, identifies the fault wave head location in the fault waveform using a preset first recurrent neural network model. This reduces the false alarm rate for fault waveforms and improves wave head location accuracy.

[0080] See also Figure 2 , Figure 2 This is a flowchart of a method for locating a fault wave head in a transmission line provided by an embodiment of the present invention. Specifically, the method may include the following steps:

[0081] Step 201: Collect power frequency current waveform data and obtain classic fault waveform data;

[0082] Step 202, calculating the similarity between the power frequency current waveform data and the classic fault waveform data;

[0083] Step 203 , when the similarity reaches a preset similarity threshold, determining the power frequency current waveform data as the first fault waveform data;

[0084] In an embodiment of the present invention, the power frequency current data collected at the device end can be matched with the classic fault waveform data. If the similarity reaches a preset similarity threshold, the power frequency current data can be determined as suspected fault current data and reported as the first fault waveform data to the main station for the next step of screening and identification.

[0085] By matching the power frequency current with the classic fault waveform data to screen the first fault waveform data, the quality of the reported data can be improved, the computational burden of the master station can be reduced, and the speed of the master station in locating the fault wave head can be accelerated.

[0086] Step 204: performing data normalization processing on the first fault waveform data to obtain input data;

[0087] After the first fault waveform data is collected, it can be subjected to data normalization processing to summarize the statistical distribution of the first fault waveform data and obtain input data for the automatic encoder.

[0088] Step 205, extracting the first latent space variable of the input data;

[0089] After the input data is obtained, an autoencoder may be used to reduce the length of the input data to obtain a first latent space variable of the input data, so as to achieve dimensionality reduction and noise reduction processing of the input first fault waveform data.

[0090] An autoencoder is an artificial neural network used for efficient encoding in unsupervised learning. The purpose of an autoencoder is to learn a representation (encoding) of a set of data, typically for dimensionality reduction.

[0091] In the specific implementation, such as Figure 3 As shown in Figure 2, the process of extracting the first latent space variables through the autoencoder can include the following steps:

[0092] A. Input the first fault waveform data with a length of 2048;

[0093] B. Use a multi-layer convolutional neural network C1 to reduce the input of length 2048 to an array of length 1024;

[0094] C. Use a multi-layer convolutional neural network C2 to reduce the input of the previous layer with a length of 1024 to an array of length 512;

[0095] D. Use a multi-layer convolutional neural network C3 to reduce the input of the previous layer with a length of 512 to an array of length 256;

[0096] E. Use a multi-layer convolutional neural network C4 to expand the input of the previous layer with a length of 256 to an array of length 512;

[0097] F. Use a multi-layer convolutional neural network C5 to expand the input of the previous layer with a length of 256 to an array of length 1024;

[0098] G. Use a multi-layer convolutional neural network C6 to expand the input of the previous layer with a length of 256 to an array of length 2048;

[0099] H. At this point, the output data length is the same as the initial input length. Calculate the least squares error between the original input first fault waveform data and the output, and use the backpropagation method to update the parameter matrices of all the above convolutional neural networks; until the least squares error between the original input first fault waveform data and the data output by C6 meets the preset threshold range, the training is completed;

[0100] I. When the training is completed and the data is actually used, the array with a length of 256 output in step D is obtained as the first latent space variable of the first fault waveform data.

[0101] Step 206: classify the first latent space variables to obtain second fault waveform data;

[0102] In an embodiment of the present invention, the first latent space variables can be classified through a deep neural network, waveforms with a judgment result of non-fault can be filtered out, and waveforms with a judgment result of fault can be retained as second fault waveform data.

[0103] This process can further remove the interference of clutter on fault waveform judgment and improve the recognition rate of fault waveform.

[0104] Step 207, extracting suspected fault wave head data from the second fault waveform data;

[0105] In this embodiment of the present invention, step 207 may include the following sub-steps:

[0106] S11, inputting the second fault waveform data into a preset second recurrent neural network and outputting a fault time window;

[0107] S12, matching the fault time window in the second fault waveform data to obtain suspected fault wave head data.

[0108] In an embodiment of the present invention, the second fault waveform data can be input into a preset second recurrent neural network to output a fault time window for making a preliminary judgment on the wave head position.

[0109] The second recurrent neural network can be trained through fault waveform data and its corresponding fault time window.

[0110] After obtaining the fault time window, the suspected fault wave head data can be found in the second fault waveform data.

[0111] Step 208, extracting the second latent space variables of the suspected fault wave head data;

[0112] After extracting the traveling wave data corresponding to the suspected fault wave head data, its second latent space variable can be extracted to perform dimensionality reduction and noise reduction on the suspected fault wave head data.

[0113] Step 209, determining whether the second latent space variable belongs to a fault waveform;

[0114] In this embodiment of the present invention, step 209 may include:

[0115] S21, input the second latent space variable into the preset random forest model and output the classification result;

[0116] S22: Determine whether the second latent space variable belongs to a fault waveform based on the classification result.

[0117] In practice, multiple latent space variables can be used to train a random forest model, with two classification results set: a non-fault waveform and a fault waveform. After the random forest model is trained, the second latent space variable is input into the trained random forest model to obtain the corresponding classification result. This step aims to further filter the fault waveforms and eliminate interference from non-fault waveforms (clutter) on fault waveform identification.

[0118] Step 210: If yes, identify the fault wave head position in the fault waveform by using a preset first recurrent neural network model.

[0119] When the second latent space variable is a fault waveform, it is input into the first recurrent neural network model to identify the fault wave head position.

[0120] The first recurrent neural network can be trained using multiple latent space variables and corresponding fault wave head positions to establish a correlation between the latent space variables and the fault wave head positions. Thus, when the second latent space variable is obtained, the corresponding fault wave head position can be inferred.

[0121] The present invention collects first fault waveform data; performs data normalization on the power frequency current waveform data to obtain input data; extracts first latent space variables from the input data; classifies the latent space variables to obtain second fault waveform data; extracts suspected fault wave head data from the second fault waveform data; extracts second latent space variables from the suspected fault wave head data; determines whether the second latent space variables belong to the fault waveform; and if so, identifies the fault wave head location in the fault waveform using a preset first recurrent neural network model. This reduces the false alarm rate for fault waveforms and improves wave head location accuracy.

[0122] See also Figure 4 , Figure 4 This is a structural block diagram of a transmission line fault wave front locating device provided by an embodiment of the present invention.

[0123] An embodiment of the present invention provides a transmission line fault wave front location device, comprising:

[0124] An acquisition module 401 is used to acquire first fault waveform data;

[0125] A normalization module 402 is configured to perform data normalization processing on the first fault waveform data to obtain input data;

[0126] A first latent space variable extraction module 403 is used to extract the first latent space variable of the input data;

[0127] A classification module 404 is used to classify the first latent space variables and obtain second fault waveform data;

[0128] The suspected fault wave front data extraction module 405 is used to extract the suspected fault wave front data from the second fault waveform data;

[0129] The second latent space variable extraction module 406 is used to extract the second latent space variables of the suspected fault wave head data;

[0130] A judgment module 407 is used to judge whether the second latent space variable belongs to a fault waveform;

[0131] The fault wave head position identification module 408 is used to identify the fault wave head position in the fault waveform by using a preset first recurrent neural network model.

[0132] In this embodiment of the present invention, the acquisition module 401 includes:

[0133] The acquisition submodule is used to collect power frequency current waveform data and obtain classic fault waveform data;

[0134] Similarity calculation submodule, used to calculate the similarity between power frequency current waveform data and classic fault waveform data;

[0135] The first fault waveform data determining submodule is configured to determine the power frequency current waveform data as the first fault waveform data when the similarity reaches a preset similarity threshold.

[0136] In this embodiment of the present invention, the first latent space variable extraction module 403 includes:

[0137] The first latent space variable extraction submodule is used to reduce the length of the input data by using an automatic encoder to obtain the first latent space variable of the input data.

[0138] In the embodiment of the present invention, the suspected fault wave head data extraction module 405 includes:

[0139] A fault time window output submodule is used to input the second fault waveform data into a preset second recurrent neural network and output a fault time window;

[0140] The suspected fault wave head data extraction submodule is used to match the fault time window in the second fault waveform data to obtain the suspected fault wave head data.

[0141] In this embodiment of the present invention, the determination module 407 includes:

[0142] The classification submodule is used to input the second latent space variable into the preset random forest model and output the classification result;

[0143] The judgment submodule is used to judge whether the second latent space variable belongs to the fault waveform according to the classification result.

[0144] An embodiment of the present invention further provides an electronic device, the device including a processor and a memory:

[0145] The memory is used to store program codes and transmit the program codes to the processor;

[0146] The processor is configured to execute the method for locating a fault wave head of a power transmission line according to the instructions in the program code.

[0147] An embodiment of the present invention further provides a computer-readable storage medium, which is used to store program code, and the program code is used to execute the transmission line fault wave head location method of the embodiment of the present invention.

[0148] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0149] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0150] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, apparatus, or computer program products. Thus, embodiments of the present invention may take the form of a fully hardware embodiment, a fully software embodiment, or an embodiment combining software and hardware. Furthermore, embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0151] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the process in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0152] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0153] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0154] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0155] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or terminal device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or terminal device that includes the element.

[0156] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for locating a fault wave head of a transmission line, characterized in that: include: collecting first fault waveform data; performing data normalization processing on the first fault waveform data to obtain input data; Extracting a first latent space variable of the input data; classifying the first latent space variables to obtain second fault waveform data; extracting suspected fault wave head data from the second fault waveform data; extracting a second latent space variable of the suspected fault wave head data; determining whether the second latent space variable belongs to a fault waveform; If so, identifying the fault wave head position in the fault waveform by presetting a first recurrent neural network model; Wherein, the step of collecting the first fault waveform data includes: Collect power frequency current waveform data and obtain classic fault waveform data; Calculating the similarity between the power frequency current waveform data and the classic fault waveform data; When the similarity reaches a preset similarity threshold, the power frequency current waveform data is determined as first fault waveform data.

2. The method according to claim 1, characterized in that The step of extracting the first latent space variable of the input data comprises: An autoencoder is used to reduce the length of the input data to obtain a first latent space variable of the input data.

3. The method according to claim 1, characterized in that The step of extracting suspected fault wave head data from the second fault waveform data includes: Inputting the second fault waveform data into a preset second recurrent neural network and outputting a fault time window; The fault time window is matched in the second fault waveform data to obtain suspected fault wave head data.

4. The method according to claim 1, wherein The step of determining whether the second latent space variable belongs to a fault waveform includes: Input the second latent space variable into the preset random forest model and output the classification result; It is determined whether the second latent space variable belongs to a fault waveform according to the classification result.

5. A transmission line fault wave head positioning device, characterized in that: include: An acquisition module, configured to acquire first fault waveform data; a normalization module, configured to perform data normalization processing on the first fault waveform data to obtain input data; A first latent space variable extraction module, used to extract the first latent space variable of the input data; a classification module, configured to classify the first latent space variables and obtain second fault waveform data; A suspected fault wave front data extraction module, configured to extract suspected fault wave front data from the second fault waveform data; A second latent space variable extraction module, used to extract the second latent space variables of the suspected fault wave head data; a judgment module, configured to judge whether the second latent space variable belongs to a fault waveform; a fault wave head position identification module, configured to identify the fault wave head position in the fault waveform by using a preset first recurrent neural network model; Wherein, the acquisition module includes: The acquisition submodule is used to collect power frequency current waveform data and obtain classic fault waveform data; A similarity calculation submodule, configured to calculate the similarity between the power frequency current waveform data and the classic fault waveform data; The first fault waveform data determining submodule is configured to determine the power frequency current waveform data as first fault waveform data when the similarity reaches a preset similarity threshold.

6. The device according to claim 5, characterized in that The first latent space variable extraction module includes: The first latent space variable extraction submodule is used to reduce the length of the input data by using an automatic encoder to obtain the first latent space variable of the input data.

7. An electronic device, characterized in that: The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the transmission line fault wave head locating method according to any one of claims 1 to 4 according to the instructions in the program code.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium is used to store program code, and the program code is used to execute the transmission line fault wave head locating method according to any one of claims 1-4.

Images