Method and system for estimating open fire distribution of overhead conductor tree touch fault

Abstract

The invention discloses a method and a system for estimating open fire distribution of an overhead conductor tree contact fault. The method comprises the following steps: acquiring zero-sequence current data of a conductor and carbonization data of a test tree in a process from conductor contact to carbonization fire of the test tree; according to the zero-sequence current data, extracting electrical characteristic data influencing the open fire distribution range, and obtaining electrical characteristic vector data; inputting the carbonized data and the electrical feature vector data into a neural network model for learning and training to obtain an open fire distribution development prediction model; and inputting the electrical characteristic data extracted in real time into the open fire distribution development prediction model to obtain an open fire distribution prediction result. According to the method, the open fire distribution development prediction model can be established according to the corresponding relation between the zero-sequence current and the carbonization process of the tree, and real-time estimation of the open fire distribution trend of the wire tree touching fault is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of conductor tree-touching fault analysis, and in particular to a method and system for estimating the open flame distribution of an overhead conductor tree-touching fault. Background Art

[0002] With the continuous increase in the length of overhead lines and the increasing complexity of the power grid structure, the problem of wire contact tree faults and the fires caused by them will become more and more prominent. Therefore, accurate prediction of open fire distribution is of great significance to ensure the stability of power supply and improve the safety and reliability of the power grid.

[0003] During the development of the wire-to-tree fault, the zero-sequence current and its harmonic components will gradually change significantly, especially before the formation of an open flame, the expansion of the carbonization path has a key impact on the current signal. As the degree of carbonization deepens, the moisture in the tree gradually decreases, which can easily cause combustion and form an open flame. The formation and distribution of this open flame are affected by the carbonization path, and the high-temperature arc distributed along the carbonization path provides a heat source for the generation of the open flame. Therefore, the zero-sequence current in the wire-to-tree fault is closely related to the development of the carbonization path. The changing trend of the zero-sequence current can characterize the extension of the carbonization path, thereby reflecting the distribution range of the open flame on the tree. By detecting the changing trend of the zero-sequence current, the appearance time and distribution trend of the open flame on the tree can be reasonably estimated. Summary of the invention

[0004] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] In view of the above existing problems, the present invention is proposed.

[0006] Therefore, the present invention provides a method and system for estimating the open flame distribution of an overhead conductor tree-touching fault to solve the problem of low accuracy in estimating the open flame distribution range in existing methods.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] In a first aspect, the present invention provides a method for estimating the distribution of open flames in an overhead conductor tree-touching fault, comprising:

[0009] Obtain the zero-sequence current data of the wire from the time when the test tree contacts the wire to the time when it carbonizes and catches fire, as well as the carbonization data of the test tree;

[0010] According to the zero-sequence current data, the electrical characteristic data affecting the distribution range of the open flame is extracted, and the electrical characteristic vector data is obtained;

[0011] The carbonization data and electrical characteristic vector data are input into the neural network model for learning and training to obtain the open fire distribution development prediction model;

[0012] The electrical characteristic data extracted in real time is input into the open flame distribution development prediction model to obtain the open flame distribution estimation results.

[0013] As a preferred solution of the method for estimating the distribution of open flames of overhead wire tree-touching faults described in the present invention, wherein:

[0014] The zero-sequence current data of the wire and the carbonization data of the test tree from the time the test tree contacts the wire to the time it carbonizes and catches fire are obtained by multiple wire-contact-tree simulation tests involving at least two or more test trees.

[0015] As a preferred solution of the method for estimating the distribution of open flames of overhead wire tree-touching faults described in the present invention, wherein:

[0016] The method of obtaining zero-sequence current data of the wire during the process from when the test tree contacts the wire to when it carbonizes and catches fire comprises the following steps:

[0017] Obtain the sampling equipment in the wire-on-tree simulation test to record the entire test process from the test tree contacting the wire to carbonization and ignition;

[0018] According to the data recorded by the sampling device, the instantaneous value of the zero-sequence current under the preset period and preset sampling frequency in the sampling device is obtained.

[0019] As a preferred solution of the method for estimating the distribution of open flames of overhead wire tree-touching faults described in the present invention, wherein:

[0020] The step of obtaining carbonization data of the test tree during the process from contact with the wire to carbonization and fire includes the following steps:

[0021] Obtain the whole process of the test tree from contacting the wire to carbonization and ignition recorded by the camera equipment in multiple wire-contact tree simulation tests;

[0022] According to the video data recorded by the camera device, carbonization data based on time series is obtained.

[0023] As a preferred solution of the method for estimating the distribution of open flames of overhead wire tree-touching faults described in the present invention, wherein:

[0024] According to the zero-sequence current data, the electrical characteristic data affecting the distribution range of the open flame is extracted, and the electrical characteristic vector data obtained includes:

[0025] Calculate the electrical characteristic data affecting the open flame distribution range in each preset cycle according to the zero-sequence current data;

[0026] According to the electrical characteristic data of each preset period, a plurality of electrical characteristic vectors are obtained;

[0027] According to multiple electrical characteristic vectors, a complete electrical characteristic matrix is ​​obtained for each conductor contact tree simulation test;

[0028] The electrical characteristic matrix is ​​obtained from each test, and the electrical characteristic vector data is obtained through multiple wire contact tree simulation tests.

[0029] As a preferred solution of the method for estimating the distribution of open flames of overhead wire tree-touching faults described in the present invention, wherein:

[0030] The electrical characteristic data include the effective value, average value and variance of the current, the fundamental wave, the average value and variance of the fundamental and even harmonics, and the fundamental and even harmonic factors.

[0031] As a preferred solution of the method for estimating the distribution of open flames of overhead wire tree-touching faults described in the present invention, wherein:

[0032] Before the carbonization data and electrical characteristic vector data are input into the neural network model for learning and training to obtain the open fire distribution development prediction model, the method further includes:

[0033] The carbonization data and the electrical characteristic vector data are matched in time series.

[0034] In a second aspect, the present invention provides a system for estimating the distribution of open flames in an overhead wire tree-touching fault, comprising:

[0035] A test data acquisition module, used to acquire the zero-sequence current data of the wire from the test tree contacting the wire to the carbonization and ignition process and the carbonization data of the test tree;

[0036] A data processing module is used to extract electrical characteristic data affecting the distribution range of open flames according to zero-sequence current data, and obtain electrical characteristic vector data;

[0037] A model training module is used to input carbonization data and electrical characteristic vector data into a neural network model for learning and training to obtain an open fire distribution development prediction model;

[0038] The open flame distribution prediction module is used to input the real-time extracted electrical characteristic data into the open flame distribution development prediction model to obtain the open flame distribution prediction result.

[0039] In a third aspect, the present invention provides a computing device, comprising:

[0040] Memory, used to store programs;

[0041] A processor is used to execute the computer executable instructions, which, when executed by the processor, implement the steps of the method for estimating the distribution of open flames due to overhead wire tree-touching faults.

[0042] In a fourth aspect, the present invention provides a computer-readable storage medium, comprising: when the program is executed by a processor, the steps of implementing the method for estimating the open flame distribution of an overhead wire tree-touching fault are implemented.

[0043] Beneficial effects of the present invention: The present invention obtains zero-sequence current data of the wire and carbonization data of the test tree during the process from contacting the wire to carbonization and ignition in multiple wire-contact-tree simulation tests; extracts electrical characteristic data that affect the open flame distribution range based on the zero-sequence current data, and obtains electrical characteristic vector data; inputs the carbonization data and electrical characteristic vector data into a neural network model for learning and training, and obtains a trained open flame distribution development prediction model based on electrical characteristics; inputs the electrical characteristic data extracted in real time into the trained open flame distribution development prediction model, and obtains an open flame distribution estimation result. The present invention can obtain a trained open flame distribution development prediction model based on electrical characteristics according to the correspondence between zero-sequence current and the carbonization process of the tree, and realize real-time detection of the open flame distribution trend of the wire-contact-tree fault. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. 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 creative labor. Among them:

[0045] Figure 1 A basic flow chart of a method for estimating the distribution of open flames for overhead conductor tree-touching faults provided by one embodiment of the present invention;

[0046] Figure 2 A basic flow chart of a method for estimating the distribution of open flames for overhead conductor tree-touching faults provided by one embodiment of the present invention;

[0047] Figure 3 A schematic diagram of the basic flow of a method for estimating the open flame distribution of an overhead conductor tree-touching fault provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0049] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0050] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0051] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0052] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0053] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0054] Example 1

[0055] Reference Figure 1, is an embodiment of the present invention, and provides a method for estimating the distribution of open flames of overhead wire tree-touching faults, such as Figure 1 As shown, the following steps are included:

[0056] S1: Obtain the zero-sequence current data of the wire from the time when the test tree contacts the wire to the time when it carbonizes and catches fire, as well as the carbonization data of the test tree;

[0057] In an embodiment of the present application, the zero-sequence current data of the wire and the carbonization data of the test tree during the process from contacting the wire to carbonization and catching fire are obtained by multiple wire-contact-tree simulation tests involving at least two or more test trees.

[0058] In the embodiment of the present application, the test tree samples of the multiple test processes include different tree species, and the different tree species include different tree diameters and different tree heights. By testing samples of different tree species, diameters and heights multiple times, richer data changes can be provided. In the data analysis stage, more variable combinations can allow the model to better explore the potential laws that affect the distribution of open flames. For example, when studying the relationship between zero-sequence current data and open flame distribution, the characteristics of different trees will lead to different zero-sequence current change laws.

[0059] In the embodiment of the present application, by combining the carbonization data of different tree heights and types, a more accurate open flame distribution model related to electrical characteristics can be established. By learning these complex situations, the model can comprehensively consider multiple factors when making predictions, improve the accuracy and reliability of the prediction results, and more accurately estimate the open flame distribution of overhead wire tree-touching faults related to various trees encountered in actual environments.

[0060] In the embodiment of the present application, the zero-sequence current data of the wire and the carbonization data of the test tree during the process from contacting the wire to carbonization and ignition of the test tree in multiple wire-contact-tree simulation tests are obtained, specifically including:

[0061] In the embodiment of the present application, the sampling device in the wire-contact-tree simulation test records the entire test process from the test tree contacting the wire to carbonization and ignition;

[0062] In the embodiment of the present application, by selecting and installing a suitable sampling device. For the sampling of zero-sequence current, it is necessary to select a current transformer with high precision, high bandwidth and anti-interference ability and a matching data acquisition card, etc., to ensure that the weak zero-sequence current changes in the wire can be accurately captured. The current transformer is correctly installed in the appropriate position of the wire to ensure that it is in good contact with the wire and can sense the current signal in real time.

[0063] In the embodiment of the present application, after the test begins, the sampling device is started to work uninterruptedly throughout the whole process, and the changes in the electrical parameters related to the test tree are recorded from the moment the test tree just touches the wire, and the whole process is completed until the tree is completely carbonized and catches fire. At the same time, the environmental parameters of the test site should also be recorded synchronously, because these environmental factors may have a certain impact on the carbonization process of the test tree and the zero-sequence current of the wire.

[0064] In the embodiment of the present application, through precise sampling equipment and standardized operation, all relevant data in the critical process from the start to the end of the test are completely and reliably recorded, providing an original and comprehensive data basis for the subsequent accurate extraction of the instantaneous value of the zero-sequence current. At the same time, the recording of environmental parameters also helps to analyze the test results more comprehensively.

[0065] In an embodiment of the present application, based on the data recorded by the sampling device, the instantaneous value of the zero-sequence current at a preset period and a preset sampling frequency in the sampling device is obtained.

[0066] In an embodiment of the present application, after the test is completed, the raw current data collected by the sampling device is first preliminarily cleaned and sorted to remove abnormal data points that may be caused by temporary equipment failure, electromagnetic interference, etc. The preliminary cleaning of the data can be achieved by setting a reasonable data threshold range and adopting a filtering algorithm (such as mean filtering, Kalman filtering, etc.).

[0067] In the embodiment of the present application, the sorted current data is extracted according to the preset sampling period and sampling frequency. For example, if the preset sampling period is one period of 20 ms and the sampling frequency is 50 Hz, the corresponding instantaneous current value is extracted from the original continuous current data. This extraction process can be achieved by writing a special data extraction program or using the relevant functions in professional data processing software (such as Matlab, etc.).

[0068] In the embodiment of the present application, the extracted zero-sequence current instantaneous value is marked and recorded in detail, and the marked content should include the corresponding timestamp, the test round to which it belongs, and other key information, so as to facilitate the query and analysis of subsequent data and match and integrate with other relevant data (such as carbonization data), and these instantaneous values ​​can be organized into a standardized data table form for storage in a waveform file. When stored as a waveform file, it is possible to restore the curve of the current amplitude over time through suitable software tools (such as oscilloscope software, data analysis software, etc.), and the waveform file will contain parameter information of the sampling device in the file header or other data areas, such as sampling frequency, sampling accuracy, trigger mode, etc. For example, when stored in a simple CSV format, the data column in the file may have a timestamp column and a corresponding current amplitude column. By drawing the timestamp as the horizontal axis and the current amplitude as the vertical axis in a two-dimensional coordinate system, an intuitive current amplitude waveform curve can be obtained, which shows the dynamic change of the zero-sequence current during the process from the occurrence of the fault to the spread of the open flame (or other set time range).

[0069] In the embodiment of the present application, the instantaneous value of the zero-sequence current is accurately extracted according to the preset requirements, so that the subsequent analysis can focus on these representative and standardized data, which is convenient for mining the current change characteristics closely related to the tree carbonization and ignition process, and provides a key electrical characteristic data basis for building a prediction model.

[0070] In the embodiment of the present application, the camera equipment records the whole process of the test tree from contacting the wire to carbonization and ignition in multiple wire-contacting-tree simulation tests;

[0071] In an embodiment of the present application, before the start of the experiment, the camera equipment is debugged and appropriate shooting parameters are set, such as adjusting the resolution to a value that can clearly distinguish the boundaries of the carbonized area of ​​the tree, setting the frame rate to a value that is sufficient to capture the rapid changes in the carbonization process, and ensuring that the storage capacity of the equipment is large enough to fully store the video data of the entire test process, or setting up a connection with an external storage device to achieve real-time data storage and backup, until the carbonization and fire of the test trees are over, completing the video recording of the entire test process.

[0072] In the embodiment of the present application, carbonization data based on time series is obtained according to the video data recorded by the camera device.

[0073] In the embodiment of the present application, professional video analysis software is used to analyze the captured video data frame by frame. First, the video is split in chronological order to obtain each frame of the image, and then the boundary of the carbonized area of ​​the tree in each frame of the image is identified by an image recognition algorithm (such as a recognition algorithm based on color features, texture features, etc.). The carbonization data in this embodiment is the ratio of the carbonization distribution length of the test tree to the total length of the test tree, which is used to indicate the degree of carbonization.

[0074] In an embodiment of the present application, these calculated carbonization degree parameters are organized according to time series. For example, the change in carbonization degree at the corresponding moment (the ratio of the carbonization distribution length of the test trees to the total length of the test trees) is recorded at a time interval of one second to form a complete carbonization data table based on time series, which clearly shows the dynamic development process of the carbonization degree of trees over time.

[0075] In the embodiment of the present application, time series-based carbonization data is successfully extracted from the video data, which can intuitively and quantitatively reflect the carbonization development law of the test trees during the tree-touching failure process, so that in the subsequent steps it can be combined with the electrical characteristic data and input into the neural network model for learning and training to more accurately construct a model that can predict the distribution of open flames.

[0076] S2: extracting electrical characteristic data affecting the distribution range of open flames according to zero-sequence current data, and obtaining electrical characteristic vector data;

[0077] In the embodiment of the present application, the electrical characteristic data affecting the open flame distribution range in each preset cycle is calculated based on the zero-sequence current data, and the electrical characteristic data includes the effective value, average value and variance of the current, the fundamental wave, the average value and variance of the fundamental and even harmonics, and the fundamental and even harmonic factors;

[0078] In the embodiment of the present application, the zero-sequence current data is processed in segments with a period of 20 ms. For the zero-sequence current signal in each data segment, the effective value RMS, average value Mean and variance Var of the current in the segment are first calculated by the following formulas:

[0079]

[0080] Where I0(i) is the value of the signal at the i-th sampling point; N is the total number of sampling points of the signal.

[0081] In the embodiment of the present application, a fast Fourier transform (FFT) is performed on the zero-sequence current signal in each data segment to obtain a frequency domain representation of the signal:

[0082]

[0083] in, is the complex representation of the signal in the frequency domain; j is the imaginary unit; f n is the frequency (n=1,2,3,4,5).

[0084] In the embodiment of the present application, the amplitude of each frequency point can be obtained by calculating the amplitude of the FFT result:

[0085]

[0086] In the embodiment of the present application, based on the calculated harmonic amplitudes, the 2nd, 3rd, 4th, and 5th harmonic factors are calculated, and the calculation formula is as follows:

[0087]

[0088] In the embodiment of the present application, each frequency signal is subjected to an inverse Fourier transform (IFFT) to obtain a time domain reconstructed signal I of each harmonic. n (t), calculate the variance of these reconstructed signals:

[0089]

[0090] Among them, μ n is the mean value of the signal at this frequency.

[0091] In the embodiment of the present application, the amplitude of the fundamental wave and each harmonic is obtained by discrete Fourier transform (DFT). For the fundamental wave amplitude mean value and variance, the fundamental wave is the main component of the waveform, and its amplitude change reflects the change of the basic characteristics of the waveform. For the mean value and variance of the 2nd-5th harmonic amplitude, these harmonics may have different performances in an asymmetric waveform. For example, in some asymmetric cases, even harmonics (such as 2nd and 4th even harmonics) may be enhanced, and the changes in their amplitude mean value and variance can serve as an important clue to detect waveform asymmetry.

[0092] In the embodiment of the present application, a plurality of electrical characteristic vectors are obtained according to the electrical characteristic data of each preset period;

[0093] In an embodiment of the present application, by calculating the electrical characteristic data of each preset cycle, a 1×17 characteristic vector will be obtained for each data segment. Assuming that the total amount of data covers T seconds, the total number of time intervals processed is T / 0.02. Each time interval corresponds to a characteristic vector, and multiple electrical characteristic vectors are generated.

[0094] In the embodiment of the present application, a complete electrical characteristic matrix is ​​obtained for each wire contact tree simulation test based on multiple electrical characteristic vectors;

[0095] In the embodiment of the present application, a plurality of electrical characteristic vectors are spliced ​​row by row to form a complete characteristic matrix corresponding to each wire contact tree simulation test process.

[0096] In the embodiment of the present application, an electrical characteristic matrix is ​​obtained according to each test, and electrical characteristic vector data is obtained through multiple wire contact tree simulation tests.

[0097] In the embodiment of the present application, the data of each experiment is processed in the same way to generate a corresponding characteristic matrix, and all characteristic matrices are saved in a specified file to obtain electrical characteristic vector data.

[0098] In the embodiment of the present application, the characteristics of the current waveform are described from different angles according to the 17 extracted features. By extracting these features for each waveform, a relatively comprehensive feature vector is constructed. When the waveform is asymmetric, these sub-features may change individually or in combination. For example, waveform asymmetry may cause the current average value to deviate from zero, while the average value and variance of the second harmonic amplitude increase, and the harmonic factor may also change.

[0099] In the embodiment of the present application, in order to ensure the alignment of carbonization data and feature vector data, this study interpolated the manually recorded carbonization data. Specifically, the number of points in the carbonization data in each experiment will be consistent with the number of rows of the extracted electrical feature vector, thereby ensuring that the carbonization data can correspond one-to-one with the electrical feature vector data in the time series, which is convenient for subsequent model training based on the number of experiments.

[0100] S3: Input the carbonization data and electrical characteristic vector data into the neural network model for learning and training to obtain the open fire distribution development prediction model;

[0101] In the embodiment of the present application, before the carbonization data and the electrical characteristic data are input into the neural network model for learning and training, and the trained open flame distribution development prediction model based on electrical characteristics is obtained, the following is also included:

[0102] The carbonization data and electrical characteristic vector data are matched in time series.

[0103] In an embodiment of the present application, after data processing is completed, the processed feature vector and carbonization data are used as the input and output of the neural network respectively, and the aforementioned neural network model is used for training to capture the complex nonlinear relationship between the feature vector and the carbonization data, and to construct an open flame distribution development prediction model based on electrical characteristics.

[0104] It should be noted that since the electrical feature vector data includes 17 features, its feature dimension is high and the data set is large, the single-layer neural network structure in the existing technology may not be able to effectively find the global optimal solution, resulting in poor prediction results.

[0105] In an embodiment of the present application, by improving the hidden layer structure of the neural network, the number of neurons in each layer is set to a decreasing manner layer by layer: the first layer is 50 neurons, the second layer is 40 neurons, and the third layer is 30 neurons. In this structure, the front layer contains more neurons, which helps to extract high-dimensional features and capture complex patterns and relationships in the input data. In particular, for data with high-dimensional features, a large number of neurons in the front layer can more fully represent the different dimensional information in the input space. As the network level deepens, the number of neurons in the back layer gradually decreases. This strategy helps to compress features, thereby removing redundant information and retaining the core features of the data. Through this improved neural network structure, it is possible to reduce network complexity and optimize the training process while ensuring efficient feature extraction, thereby improving the computational efficiency and generalization ability of the model.

[0106] In the embodiment of the present application, the established model aims to reveal the nonlinear relationship between the open flame distribution range and various electrical quantity parameters. By selecting the Tansig activation function, it has a stronger nonlinear expression ability and can fit complex nonlinear relationships more accurately, especially when dealing with complex patterns such as the open flame distribution range. This can further optimize the performance of the model and improve the fitting effect of the nonlinear relationship.

[0107] S4: Input the electrical characteristic data extracted in real time into the open flame distribution development prediction model to obtain the open flame distribution estimation result.

[0108] In an embodiment of the present application, in an actual overhead line operation scenario, the current electrical characteristic data is extracted in real time by real-time monitoring equipment (such as current transformers and oscilloscopes, etc.) deployed at the corresponding location. The real-time data must also be feature extracted and sorted according to the data processing specifications determined in step S2 to ensure that the data format is consistent with the electrical characteristic data input when training the model.

[0109] In an embodiment of the present application, these electrical characteristic data extracted and processed in real time are input into a trained open flame distribution development prediction model. The model will quickly output the corresponding open flame distribution prediction results based on the internally learned mapping relationship. The results can be presented to relevant operation and maintenance personnel in a visual form (generating an image of the open flame distribution range, the carbonization degree of the ratio of the carbonized distribution length of the tree to the total length of the tree).

[0110] It should be noted that the present invention can obtain a trained open flame distribution development prediction model based on electrical characteristics according to the correspondence between zero-sequence current and the carbonization process of trees. The method of estimating the open flame distribution of wire-touching-tree faults based on multi-dimensional zero-sequence current characteristics is closely related to the physical mechanism of the development of open flames in wire-touching-tree faults in principle. Compared with the existing estimation methods, the characteristics of even harmonics are added, which effectively avoids the false detection and missed detection caused by waveform asymmetry in the existing methods. The present invention realizes the real-time prediction of the open flame distribution of overhead wire-touching-tree faults in actual operation using the prediction model established in the early stage, which provides timely and valuable reference for power operation and maintenance personnel to take effective fire extinguishing, isolation and other preventive measures in advance, and helps to reduce the losses and safety risks caused by faults.

[0111] This embodiment also provides a system for estimating the distribution of open flames in an overhead wire tree-touching fault, including:

[0112] A test data acquisition module, used to acquire the zero-sequence current data of the wire from the test tree contacting the wire to the carbonization and ignition process and the carbonization data of the test tree;

[0113] A data processing module is used to extract electrical characteristic data affecting the distribution range of open flames according to zero-sequence current data, and obtain electrical characteristic vector data;

[0114] A model training module is used to input carbonization data and electrical characteristic vector data into a neural network model for learning and training to obtain an open fire distribution development prediction model;

[0115] The open flame distribution prediction module is used to input the real-time extracted electrical characteristic data into the open flame distribution development prediction model to obtain the open flame distribution prediction result.

[0116] Furthermore, it also includes:

[0117] Memory, used to store programs;

[0118] A processor is used to load the program to execute the method for estimating the distribution of open flames caused by overhead wire tree-touching faults.

[0119] This embodiment also provides a computer-readable storage medium storing a program, which, when executed by a processor, implements the method for estimating the distribution of open flames due to overhead wire tree-touching faults.

[0120] The storage medium proposed in this embodiment and the method for estimating the open flame distribution of overhead wire tree-touching faults proposed in the above embodiment belong to the same inventive concept. The technical details not described in detail in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0121] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ReadOnly, Memory, ROM), random access memory (RandomAccess Memory, RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform the methods of various embodiments of the present invention.

[0122] Example 2

[0123] This is an embodiment of the present invention, which provides a method for estimating the distribution of open flames in an overhead conductor tree-touching fault. In order to verify its beneficial effects, comparison results of two schemes are provided.

[0124] In order to better verify the superiority of the model proposed in the method for estimating the distribution of open flames in overhead wire tree faults proposed in the present invention (the extracted features include odd and even harmonics, and odd and even harmonic factors), it is compared with the model of the prior art (the extracted features only include odd harmonics and odd harmonic factors). Figure 2 and Figure 3 It can be seen that the model proposed by the method for estimating the distribution of open flames of overhead wire tree faults in the present invention is better than the existing model in terms of prediction effect. In the existing model, the maximum absolute error is 21%, the minimum is 0%, and the average absolute error is 7%; while the maximum absolute error of the model proposed in this application is 4%, the minimum is 0%, and the average absolute error is 1%. The results show that the model proposed in this application has obvious advantages in accuracy and has good practicality and application value.

[0125] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for estimating the distribution of open flames in overhead wire tree-touching faults, characterized in that: include: Obtain the zero-sequence current data of the wire from the time when the test tree contacts the wire to the time when it carbonizes and catches fire, as well as the carbonization data of the test tree; According to the zero-sequence current data, the electrical characteristic data affecting the distribution range of the open flame is extracted, and the electrical characteristic vector data is obtained; The carbonization data and electrical characteristic vector data are input into the neural network model for learning and training to obtain the open fire distribution development prediction model; The electrical characteristic data extracted in real time is input into the open flame distribution development prediction model to obtain the open flame distribution estimation results.

2. The method for estimating the distribution of open flames due to an overhead wire tree-touching fault according to claim 1, characterized in that: The zero-sequence current data of the wire and the carbonization data of the test tree from the time the test tree contacts the wire to the time it carbonizes and catches fire are obtained by multiple wire-contact-tree simulation tests involving at least two or more test trees.

3. The method for estimating the distribution of open flames due to an overhead wire tree-touching fault according to claim 1 or 2, characterized in that: The method of obtaining zero-sequence current data of the wire during the process from when the test tree contacts the wire to when it carbonizes and catches fire comprises the following steps: Obtain the sampling equipment in the wire-on-tree simulation test to record the entire test process from the test tree contacting the wire to carbonization and fire; According to the data recorded by the sampling device, the instantaneous value of the zero-sequence current under the preset period and preset sampling frequency in the sampling device is obtained.

4. The method for estimating the distribution of open flames due to an overhead wire tree-touching fault according to claim 3, characterized in that: The step of obtaining carbonization data of the test tree from contact with the wire to carbonization and fire includes the following steps: Obtain the whole process of the test tree from contacting the wire to carbonization and ignition recorded by the camera equipment in multiple wire-contact tree simulation tests; According to the video data recorded by the camera device, carbonization data based on time series is obtained.

5. The method for estimating the distribution of open flames due to an overhead wire tree-touching fault according to claim 4, characterized in that: According to the zero-sequence current data, the electrical characteristic data affecting the distribution range of the open flame is extracted, and the electrical characteristic vector data obtained includes: Calculate the electrical characteristic data affecting the open flame distribution range in each preset cycle according to the zero-sequence current data; According to the electrical characteristic data of each preset period, a plurality of electrical characteristic vectors are obtained; According to multiple electrical characteristic vectors, a complete electrical characteristic matrix is ​​obtained for each conductor contact tree simulation test; The electrical characteristic matrix is ​​obtained from each test, and the electrical characteristic vector data is obtained through multiple wire contact tree simulation tests.

6. The method for estimating the distribution of open flames due to an overhead wire tree-touching fault according to claim 5, characterized in that: The electrical characteristic data include the effective value, average value and variance of the current, the fundamental wave, the average value and variance of the fundamental and even harmonics, and the fundamental and even harmonic factors.

7. The method for estimating the distribution of open flames due to an overhead wire tree-touching fault according to claim 6, characterized in that: Before the carbonization data and electrical characteristic vector data are input into the neural network model for learning and training to obtain the open fire distribution development prediction model, the method further includes: The carbonization data and the electrical characteristic vector data are matched in time series.

8. A system based on the method for estimating the distribution of open flames in overhead wire tree-touching faults according to claim 1, characterized in that: A test data acquisition module, used to acquire the zero-sequence current data of the wire from the test tree contacting the wire to the carbonization and ignition process and the carbonization data of the test tree; A data processing module is used to extract electrical characteristic data affecting the distribution range of open flames according to zero-sequence current data, and obtain electrical characteristic vector data; A model training module is used to input carbonization data and electrical characteristic vector data into a neural network model for learning and training to obtain an open fire distribution development prediction model; The open flame distribution prediction module is used to input the real-time extracted electrical characteristic data into the open flame distribution development prediction model to obtain the open flame distribution prediction result.

9. A computing device, characterized in that include: Memory, used to store programs; A processor is used to load the program to execute the steps of the method for estimating the distribution of open flames due to an overhead wire tree-touching fault as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a program, characterized in that: When the program is executed by a processor, the steps of the method for estimating the distribution of open flames due to an overhead wire tree-touching fault as described in any one of claims 1 to 7 are implemented.

Images