Power transmission line high-resistance grounding fault identification method and system

Abstract

The invention relates to the technical field of power transmission line fault identification, and discloses a power transmission line high-resistance grounding fault identification method and system, and the method comprises the steps: sequentially judging and calculating the fault type and fault distance of a line through the recording data after a power transmission line fault and line parameters; and calculating the resistance values of the time-varying transition resistors through the self-adaptive sliding data window, comparing the mean value of the resistance values of the time-varying transition resistors with a preset threshold value, and judging whether a high-resistance grounding fault occurs in the line or not. And forming a time-varying transition resistance image based on the resistance value of the time-varying transition resistance, and inputting the time-varying transition resistance image into a pre-constructed fault reason identification model to output a fault occurrence reason and a prediction probability thereof. According to the method, the problems that the resistance value of the steady-state transition resistor cannot reflect the nonlinearity and randomness of the fault equivalent resistor and the fault characteristics cannot be completely reflected are solved; meanwhile, the method can reflect the essential characteristics of the high-resistance grounding fault, enhance the function of mechanism analysis in the decision making process, and improve the interpretability of an algorithm parameter structure.

Description

Technical Field

[0001] The present invention relates to the technical field of transmission line fault identification, and in particular to a method and system for identifying a high-resistance grounding fault in a transmission line. Background Art

[0002] The safe and reliable operation of transmission lines is an important condition for ensuring the stable operation of the power grid. However, due to the wide distribution range of transmission lines, the large area they cross, the complex line structure, the changing operating environment, and their susceptibility to harsh natural environments, various types of faults are prone to occur during operation. However, the fault analysis devices currently used in power grid operation and maintenance have limited functions and accuracy, and the reclosing of lines is relatively blind, often resulting in multiple trips on the same line in a short period of time, or incorrect protection actions. In addition, the determination of the nature of the fault mainly relies on manual line inspections, which cannot accurately determine the nature of the fault and provide recovery measures in a timely and accurate manner. Therefore, quickly and accurately identifying the cause of the fault is of great significance for guiding protection and safety automatic devices to quickly remove the fault, subsequently restore power supply, and ensure the reliability of the power supply of the power grid.

[0003] Ground faults are a common type of fault in power system operation. High-impedance ground faults (HIFs) have attracted considerable attention due to their unique characteristics. High-impedance ground faults typically occur when overhead lines come into contact with high-impedance media (such as branches, gravel, and concrete). Their equivalent resistance can reach hundreds or even thousands of ohms and exhibits nonlinear characteristics. The fault process often involves unstable ground arcs, resulting in weak fault currents with nonlinear and random variations. The characteristics of this fault current closely resemble transient disturbances during normal operation (such as capacitor switching and load switching), making it difficult for traditional protection devices to effectively detect and eliminate them. Furthermore, high-impedance ground faults typically persist for extended periods. The unstable ground arcs and high temperatures around the grounding point can potentially cause fires, leading to serious system equipment and personal safety incidents. Therefore, effective solutions are urgently needed to accurately detect high-impedance ground faults to mitigate risks and avoid potential hazards.

[0004] Currently, mainstream high-resistance ground fault detection methods are divided into two main categories: threshold-based methods and artificial intelligence-based methods. Threshold-based methods extract specific characteristic quantities from the fault signal based on mechanism analysis or further calculate new characteristic quantities. Then, based on the characteristic quantities and empirical formulas, a reasonable threshold is set to achieve HIF detection. This includes methods based on steady-state transition resistance values. These methods extract the voltage and current phasors after the fault transient process stabilizes, calculate the steady-state transition resistance value, and set a threshold for HIF detection. These methods have clear physical meaning and are highly interpretable. However, due to the significant differences in the nonlinear characteristics of high-resistance ground faults at different arcing stages, intermittent arcing can produce irregular distortion that is difficult to filter out, and the possible offset of the arc current distortion center, in practical applications, relying solely on the steady-state transition resistance value is difficult to account for different fault scenarios. This leads to problems such as incomplete representation of fault characteristics and difficulty in threshold setting.

[0005] With the development of artificial intelligence (AI) technology, algorithms such as neural networks, expert systems, decision trees, and fuzzy theory have been widely applied to high-resistance ground fault detection. These methods train models using large amounts of data to learn the behavior patterns of high-resistance ground fault characteristics, thereby enabling detection and addressing the difficulty of threshold setting. However, these methods rely on the statistical distribution of data, neglecting the role of mechanism analysis in the decision-making process, and the algorithm parameter structure lacks interpretability. In the safety-sensitive task of high-resistance ground fault detection, models lacking interpretability can lead to misjudgments, exacerbating system operational risks and even triggering major safety incidents such as fires and electric shocks. Therefore, to effectively address the shortcomings of existing high-resistance ground fault detection methods and improve the safety and reliability of power systems, it is urgent to develop a high-resistance ground fault detection method that comprehensively considers the fault mechanism and combines it with artificial intelligence. Summary of the Invention

[0006] In view of the above existing problems, the present invention is proposed.

[0007] Therefore, the present invention provides a method and system for identifying high-resistance grounding faults in transmission lines to solve the problems of existing high-resistance grounding fault detection technologies, such as the difficulty of threshold-based methods in adapting to different fault scenarios and the difficulty in threshold setting, and the lack of interpretability of artificial intelligence-based methods, which may lead to safety hazards such as misjudgment.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] In a first aspect, the present invention provides a method for identifying a high-resistance grounding fault in a transmission line, comprising:

[0010] Obtaining a first result by performing a first judgment calculation on line parameters and recorded wave data of the fault line;

[0011] Adaptively sliding the data window based on the first result to calculate the time-varying transition resistance value during the fault period;

[0012] Comparing the average value of the time-varying transition resistance with the first fault threshold to obtain a second result;

[0013] Based on the second result, a grayscale image formed according to the resistance value of the time-varying transition resistor is input into a pre-built first identification model, and the cause of the fault and its predicted probability are output.

[0014] As a preferred solution of the method for identifying a high-resistance grounding fault of a transmission line according to the present invention, the first result includes at least a first fault type and a first fault distance of the fault line.

[0015] As a preferred solution of the method for identifying a high-resistance grounding fault in a transmission line according to the present invention, the adaptive sliding data window for calculating the time-varying transition resistance during the fault period includes:

[0016] determining a first sampling frequency and a first number of sliding cycles;

[0017] Calculate the number of sampling points of data on both sides of the line within one system cycle based on the first sampling frequency;

[0018] Calculate the total sliding length and sliding step length of the data windows on both sides of the line based on the first sliding cycle number and the number of sampling points;

[0019] The transition resistance value at the corresponding moment of each data window is calculated, and the time-varying transition resistance value is obtained through the calculation results of all data windows.

[0020] As a preferred solution of the method for identifying a high-resistance grounding fault in a transmission line according to the present invention, obtaining the second result includes:

[0021] Setting a first fault threshold;

[0022] When the average value of the time-varying transition resistance is greater than or equal to the first fault threshold, it is determined that a high-resistance grounding fault occurs in the line; otherwise, it is determined that no high-resistance grounding fault occurs in the line.

[0023] As a preferred solution of the method for identifying high-resistance grounding faults in transmission lines described in the present invention, the time-varying transition resistance value is normalized to obtain a time-varying transition resistance curve, and the color RGB three-channel values of the time-varying transition resistance curve are corresponded to the time-varying transition resistance value scale to achieve RGB normalization, thereby obtaining a grayscale image with a standard size.

[0024] As a preferred solution of the method for identifying a high-resistance grounding fault of a transmission line according to the present invention, wherein: the pre-constructed first identification model includes an input layer, an intermediate hidden layer and an output layer;

[0025] The input layer is used to receive and process the grayscale image of the standard size and pass it to the intermediate hidden layer, and the input layer includes a convolution layer, a normalization layer and an activation function layer;

[0026] The intermediate hidden layer includes multiple convolutional layers and pooling layers for extracting image features and performing nonlinear transformations. The output of the input layer is converted into a feature map through the residual block of ResNet18. The feature map is pooled using a global average pooling layer. The height and width of each feature map are reduced and then expanded into a single row tensor form based on the principle of parameter invariance.

[0027] The output layer takes the output of the intermediate hidden layer as input and outputs the cause of the fault and its predicted probability; the output layer includes a fully connected layer and an activation layer, the fully connected layer multiplies the single-row tensor input tensor output by the intermediate hidden layer by the weight matrix and adds a bias term; the activation layer applies a Softmax activation function to the output of the fully connected layer to output the cause of the fault and its predicted probability.

[0028] As a preferred solution of the method for identifying a high-resistance grounding fault in a transmission line according to the present invention, the calculation of the transition resistance value at the corresponding moment of each data window includes:

[0029] Calculate the equivalent zero-sequence impedance of the power supplies on both sides of the line;

[0030] Determining a special phase according to a first fault type of the fault line, and calculating new voltage and current phasors based on the acquired voltage and current phasors;

[0031] According to the equivalent zero-sequence impedance of the power sources on both sides of the line and the calculated new voltage and current phasors, the transition resistance values are calculated respectively according to different first fault types of the fault line.

[0032] In a second aspect, the present invention provides a system for identifying a high-resistance grounding fault in a transmission line, comprising:

[0033] A first calculation module is used to obtain a first result by performing a first judgment calculation on the line parameters and the recorded wave data of the fault line;

[0034] A second calculation module, configured to calculate the time-varying transition resistance value during the fault period based on the first result adaptively sliding the data window;

[0035] A first comparison module is configured to compare an average value of the time-varying transition resistance with a first fault threshold to obtain a second result;

[0036] The fault identification module is used to input the grayscale image formed according to the resistance value of the time-varying transition resistor into a pre-built first identification model based on the second result, and output the cause of the fault and its predicted probability.

[0037] In a third aspect, the present invention provides an electronic device, comprising:

[0038] memory and processor;

[0039] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method for identifying a high-resistance grounding fault of a transmission line are implemented.

[0040] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the method for identifying a high-resistance grounding fault of a transmission line.

[0041] Compared with the prior art, the present invention has the following advantages: The present invention provides a method and system for identifying high-resistance grounding faults in power transmission lines. The method uses recorded waveform data and line parameters after a transmission line fault to sequentially determine and calculate the fault type and fault distance of the line. The fault type and fault distance are then used as known conditions, and an adaptive sliding data window is used to calculate the time-varying transition resistance during the fault period. Finally, the mean of the time-varying transition resistance is compared with a pre-set threshold to determine whether a high-resistance grounding fault has occurred on the line. This solves the problem that the steady-state transition resistance cannot reflect the nonlinearity and randomness of the equivalent resistance of a high-resistance grounding fault, and thus the fault characteristics cannot be fully reflected. The present invention utilizes a deep neural network unit, including an input layer, an intermediate hidden layer, and an output layer, to perform deep learning and recognition on the time-varying transition resistance image of the fault sample, thereby determining the specific cause of the high-resistance grounding fault. The image input of this image recognition method is the time-varying transition resistance of the fault sample, which can reflect the essential characteristics of the high-resistance grounding fault, enhance the role of mechanism analysis in the decision-making process, and improve the interpretability of the algorithm parameter structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of 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 paying any creative work.

[0043] Figure 1 A logical diagram of the overall process of the method according to an embodiment of the present invention;

[0044] Figure 2 This is a structural diagram of a cause identification model of the method according to an embodiment of the present invention;

[0045] Figure 3 This is a schematic diagram of a typical time-varying transition resistance image of a lightning fault according to the method of an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of a typical time-varying transition resistance image of a wildfire fault according to the method of an embodiment of the present invention;

[0047] Figure 5 This is a schematic diagram of a typical time-varying transition resistance image of a pollution flashover fault according to the method of an embodiment of the present invention;

[0048] Figure 6 A schematic diagram of a typical time-varying transition resistance image of a foreign body fault according to the method of an embodiment of the present invention;

[0049] Figure 7 This is a schematic diagram of a typical time-varying transition resistance image of an icing fault according to the method described in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0051] Example 1, with reference to Figure 1 As an embodiment of the present invention, a method for identifying high-resistance grounding faults in transmission lines is provided, which solves the problem that the steady-state transition resistance value cannot reflect the nonlinearity and randomness of the equivalent resistance of the high-resistance grounding fault and the fault characteristics cannot be fully reflected. Figure 1 The specific steps shown include:

[0052] S100: Obtaining a first result by performing a first judgment calculation on line parameters and recorded wave data of the faulty line;

[0053] S200: Calculating the time-varying transition resistance during the fault period based on the first result through an adaptive sliding data window;

[0054] S300: Compare the average value of the time-varying transition resistance with the first fault threshold to obtain a second result;

[0055] S400: Based on the second result, a grayscale image formed according to the time-varying transition resistance is input into a pre-built first identification model, and the cause of the fault and its predicted probability are output.

[0056] It should be noted that the present invention provides a method and system for identifying high-resistance grounding faults on power transmission lines. This method uses recorded waveform data and line parameters after a transmission line fault to sequentially determine and calculate the fault type and fault distance of the line. The fault type and fault distance are then used as known conditions, and an adaptive sliding data window is used to calculate the time-varying transition resistance during the fault period. Finally, the mean of the time-varying transition resistance is compared with a pre-set threshold to determine whether a high-resistance grounding fault has occurred on the line. This method addresses the problem that the steady-state transition resistance cannot reflect the nonlinearity and randomness of the equivalent resistance of a high-resistance grounding fault, and thus the fault characteristics cannot be fully reflected. The present invention utilizes a deep neural network unit, including an input layer, an intermediate hidden layer, and an output layer, to perform deep learning and recognition on the time-varying transition resistance image of the fault sample, thereby determining the specific cause of the high-resistance grounding fault. The image input of this image recognition method is the time-varying transition resistance of the fault sample, which can reflect the essential characteristics of the high-resistance grounding fault, enhance the role of mechanism analysis in the decision-making process, and improve the interpretability of the algorithm parameter structure.

[0057] Example 2, reference Figures 2 to 7 According to an embodiment of the present invention, a method for identifying a high-resistance grounding fault in a transmission line is provided.

[0058] In the embodiment of the present application, the first result obtained by performing a first judgment calculation on the line parameters and the recorded data of the faulty line in step S100 includes:

[0059] Specifically, this embodiment defines the two sides of the fault line as the M side and the N side, that is, the two ends of the fault line. By defining the M side and the N side, the position of the fault relative to the two ends of the line can be clearly determined, thereby quickly locating the fault range, improving the search efficiency of the staff, and facilitating the analysis of fault characteristics and the changing patterns of electrical quantities.

[0060] Specifically, the line parameters of the fault line refer to the total positive-sequence impedance, negative-sequence impedance, and zero-sequence impedance of the fault line, which are represented by Z1, z2, and z0, respectively. The recorded data of the fault line includes three-phase and zero-sequence voltage and current data extracted from the fault recording files on both sides of the fault line, wherein the reference direction of the voltage data is the node-to-ground voltage, and the reference direction of the current data is from both sides of the line to the line. That is, the fault recorded data in the present invention refers to the three-phase and zero-sequence voltage and current data of the fault line.

[0061] Preferably, the recorded data of the fault line is extracted using a data window of two cycles before and after the fault starting point;

[0062] Specifically, calculating and obtaining voltage and current phasors based on fault recording data includes: performing Fourier transform on data of a system cycle data window length before and after the fault starting point, and then performing time synchronization to obtain three-phase and zero-sequence voltage and current phasors in the load phase and the fault phase;

[0063] Preferably, the data window after the fault starting point is taken from the time period when the fault transient process tends to be stable.

[0064] In an optional embodiment, the first judgment calculation can be based on a preliminary analysis of current and voltage, detecting whether a fault exists by comparing the current difference at both ends of the fault line; the first judgment calculation can be based on an analysis of time domain / frequency domain characteristics; the first judgment calculation can also be based on a preliminary classification based on machine learning, that is, using historical fault data to train a model, and then using this model to perform preliminary classification on new recorded data.

[0065] In an embodiment of the present application, the first judgment calculation includes: judging and obtaining the first fault type of the line based on the current phasor; calculating and obtaining the first fault distance of the line based on the line parameters and the voltage and current phasors; wherein the first result includes at least the first fault type and the first fault distance of the fault line.

[0066] In an optional embodiment, the first fault type of the line includes single-phase grounding of phase A, single-phase grounding of phase B, single-phase grounding of phase C, AB two-phase short circuit grounding, BC two-phase short circuit grounding, CA two-phase short circuit grounding, AB two-phase short circuit, BC two-phase short circuit, CA two-phase short circuit, and ABC three-phase short circuit, which are represented by AN, BN, CN, ABN, BCN, CAN, AB, BC, CA, and ABC respectively.

[0067] Specifically, the process of determining the first fault type of the line includes:

[0068] The current mutation amount is calculated based on the acquired current phasor. The formula is:

[0069]

[0070] in, is the three-phase current phasor in the load phase, is the three-phase current phasor at the fault stage;

[0071] The zero-sequence current mutation coefficient is calculated based on the acquired current phasor. The formula is:

[0072]

[0073] The zero-sequence current mutation coefficient is compared with the preset threshold k1 to determine whether the fault type is a ground fault. Specifically, if the zero-sequence current mutation coefficient is satisfied, Then the fault type is judged to be a ground fault; otherwise, The fault type is determined to be a non-grounding fault. Among them, the fault types of grounding faults include AN, BN, CN, ABN, BCN, and CAN, and the fault types of non-grounding faults include AB, BC, CA, and ABC;

[0074] If the fault type is determined to be a ground fault, it is further determined whether the fault type is a single-phase ground fault, where the single-phase ground fault includes the fault types AN, BN, and CN. The specific judgment process includes: and If both conditions are met, the fault type is judged to be AN; and If both conditions are met, the fault type is judged to be BN; and If all of the above conditions are met, the fault type is determined to be CN; if none of the above conditions are met, the fault type is determined to be a two-phase short-circuit grounding fault, where the two-phase short-circuit grounding fault includes the fault types ABN, BCN, and CAN, where k2 is a preset threshold;

[0075] If the fault type is determined to be a two-phase short-circuit grounding fault, further judgment is performed. If ΔI max =|ΔI AB |, then the fault type is judged to be ABN; if ΔI max =|ΔI BC |, the fault type is determined to be BCN; if none of the above conditions are met, the fault type is determined to be BCN, where ΔImax represents the maximum current difference when determining the two-phase short-circuit grounding fault type;

[0076] If the fault type is determined to be a non-ground fault, it is further determined whether the fault type is a two-phase short circuit fault, wherein the two-phase short circuit fault includes the fault types AB, BC, and CA. The specific judgment process includes: and If both conditions are met, the fault type is judged to be AB; and If both conditions are met, the fault type is judged to be BC; and If all of the above conditions are met at the same time, the fault type is judged to be CA; if none of the above conditions are met, the fault type is judged to be ABC three-phase short circuit.

[0077] In an optional embodiment, the first fault distance of the line refers to the ratio of the distance between the fault point of the line and the M side thereof to the total length of the line, that is, the fault distance in the present invention is a relative distance.

[0078] Specifically, the calculation process of the first fault distance of the line includes:

[0079] Based on the obtained voltage and current phasors, the negative sequence voltage and negative sequence current phasors are calculated and obtained. The formula is:

[0080]

[0081] in, is the three-phase voltage phasor at the fault stage on the M side of the fault line, is the three-phase current phasor at the fault stage on the M side of the fault line, is the three-phase voltage phasor at the fault stage on the N side of the fault line, is the three-phase current phasor at the fault stage on the N side of the fault line;

[0082] The equivalent negative-sequence impedance of the power supplies on both sides of the line is calculated based on the negative-sequence voltage and negative-sequence current phasors. The formula is:

[0083]

[0084] Further calculation of the quadratic equation Ax 2 +Bx+C=0 coefficient, the formula is:

[0085]

[0086] Solve the linear equation Ax 2 +Bx+C=0. Further, the root with a value range of 0-1 is taken as the fault distance of the line and represented by d.

[0087] It should be noted that the above step S100 provides the necessary input conditions for the subsequent accurate calculation of the time-varying transition resistance value, lays the foundation for identifying the type and location of the high-resistance grounding fault, effectively utilizes existing data resources, and ensures the accuracy and reliability of subsequent analysis.

[0088] In the embodiment of the present application, the above step S200 of calculating the time-varying transition resistance value during the fault period based on the first result adaptive sliding data window includes:

[0089] determining a first sampling frequency and a first number of sliding cycles;

[0090] Calculate the number of sampling points of data on both sides of the line within a system cycle based on the first sampling frequency;

[0091] Calculate the total sliding length and sliding step of the data windows on both sides of the line based on the first sliding cycle number and the sampling point number;

[0092] The transition resistance value at the corresponding moment of each data window is calculated, and the time-varying transition resistance value is obtained through the calculation results of all data windows.

[0093] In an optional embodiment, according to the Nyquist sampling theorem, the first sampling frequency should be at least twice the highest frequency component of the signal. However, in practical applications, in order to more accurately analyze signal characteristics, the sampling frequency is often much higher than this minimum requirement. For power systems, the first sampling frequency may be in the range of several kilohertz (kHz), such as 4kHz, 8kHz, or even higher. The selection of the first sliding period number mainly affects the size and overlap of the data window, and thus the temporal resolution and stability of the calculation results. Selecting a smaller sliding period number means higher temporal resolution but may increase the computational burden. A larger sliding period number may result in reduced temporal resolution but helps smooth the results and reduce the impact of noise. The specific value needs to be determined based on actual needs. For example, if the system period is 50Hz (i.e., each period is 20 milliseconds), one possible option is to set the sliding period number to 1 to 5 periods, which means adjusting the data window size between 20 milliseconds and 100 milliseconds.

[0094] Specifically, the steps of adaptively sliding the data window in this embodiment include:

[0095] Determine the sampling frequency of the fault recording data on both sides of the line, and based on this, calculate the number of sampling points of the fault recording data on both sides of the line within one system cycle. The formula is:

[0096]

[0097] Where T is the system period, f SM ,f SN are the sampling frequencies of the fault recording data on the M and N sides of the line, respectively. SM ,N SN are the number of sampling points of the fault recording data on the M side and the N side of the line in one system cycle respectively;

[0098] Determine the system cycle number of the data window sliding and calculate the total length of the data window sliding on both sides of the line. The formula is:

[0099]

[0100] Among them, k is the number of system cycles of data window sliding, l M ,l N are the total sliding lengths of the discrete data windows on the M and N sides of the line, respectively;

[0101] Preferably, to make the calculated time-varying transition resistance data more continuous, the step length of the data window sliding should be minimized as much as possible, and the step length of the data window sliding should be a positive integer, and the step length of the data window sliding is 1. At the same time, since the total sliding length of the data windows on both sides of the line is not necessarily the same, in order to keep the data windows on both sides of the line consistent in time during the sliding process, only the side with the smaller total sliding length of the data window can take a sliding step length of 1, and the other side takes a corresponding value based on the multiple of the total sliding length of the data window. The formula is expressed as:

[0102]

[0103] Among them, S M ,S N are the sliding steps of the fault recording data window on the M side and the N side of the line, respectively. S min=min{N SM ,N SN}, indicating N SM ,N SN the smaller of

[0104] After determining the sliding step size of the data window, a series of continuously sliding data windows can be obtained. The transition resistance value at the corresponding moment of each data window can be calculated. Through all data windows, the time-varying transition resistance value can be calculated and obtained.

[0105] In the embodiment of the present application, calculating the transition resistance value at the corresponding moment of each data window includes:

[0106] Calculate the equivalent zero-sequence impedance of the power supply on both sides of the line. The formula is:

[0107]

[0108] in, are the zero-sequence voltage and current phasors at the fault stage on the M side of the fault line, are the zero-sequence voltage and current phasors at the fault stage on the N side of the fault line;

[0109] The special phase is determined according to the first fault type of the fault line, and new voltage and current phasors are calculated based on the obtained voltage and current phasors. The formula is expressed as follows:

[0110]

[0111] in, are the fault-specific phase voltage and current phasors at the fault stage on the M side of the fault line, are the voltage phasors of the special phases of the fault phase lagging and leading faults on the M side of the fault line, are the current phasors of the special phases of the lagging and leading faults at the fault stage on the M side of the faulted line, respectively;

[0112] According to the equivalent zero-sequence impedance of the power sources on both sides of the line and the calculated new voltage and current phasors, the transition resistance values are calculated respectively according to the different first fault types of the fault line.

[0113] Specifically, when the fault type belongs to {AN, BN, CN}, the formula for calculating the transition resistance is as follows:

[0114]

[0115] Specifically, when the fault type belongs to {ABN, BCN, CAN}, the formula for calculating the transition resistance is as follows:

[0116]

[0117] Specifically, when the fault type belongs to {AB, BC, CA}, the formula for calculating the transition resistance is as follows:

[0118]

[0119] Specifically, when the fault type is ABC, the formula for calculating the transition resistance is as follows:

[0120]

[0121] It should be noted that the above step S200 can dynamically adjust the analysis window to adapt to different fault conditions, accurately capture the change process of the transition resistance, and thus more accurately reflect the nonlinear and random characteristics of the high-resistance grounding fault, providing key parameter support for subsequent fault identification.

[0122] In the embodiment of the present application, the above step S300 compares the average value of the time-varying transition resistor with the first fault threshold, and the second result obtained includes:

[0123] Setting a first fault threshold;

[0124] When the average value of the time-varying transition resistance is greater than or equal to the first fault threshold, it is determined that a high-resistance grounding fault occurs in the line; otherwise, it is determined that no high-resistance grounding fault occurs in the line.

[0125] In an optional embodiment, the determination of the first fault threshold includes: ① analyzing data from past high-resistance ground faults to determine a reasonable threshold; ② considering how different operating environments may affect the measured transition resistance; ③ typically adding a safety margin to the calculated base threshold to ensure system safety and avoid missed faults (i.e., actual faults not being detected); and ④ referencing relevant technical literature and industry-recommended practice guidelines. The setting of the first fault threshold is not fixed but should be adjusted and optimized based on actual conditions and continuously updated and improved with the accumulation of more data and technological developments.

[0126] It should be noted that the above step S300 can effectively identify whether a high-resistance grounding fault occurs, provide a clear judgment standard, simplify the fault detection process, and make it possible to quickly distinguish between a normal state and a fault state.

[0127] In the embodiment of the present application, the above step S400 inputs the grayscale image formed according to the resistance value of the time-varying transition resistor into the pre-built first identification model based on the second result, and outputs the cause of the fault and its predicted probability, including:

[0128] Specifically, obtaining a grayscale image formed according to the resistance value of the time-varying transition resistor includes: normalizing the resistance value of the time-varying transition resistor to obtain a time-varying transition resistance curve, and corresponding the color RGB three-channel values of the time-varying transition resistance curve to the resistance scale of the time-varying transition resistor to achieve RGB normalization, thereby obtaining a grayscale image with a standard size.

[0129] Specifically, the calculation formula for normalizing the time-varying transition resistance is:

[0130]

[0131] in, are the time-varying transition resistance values before and after normalization, x max ,x min are the maximum and minimum values of the time-varying transition resistance before normalization.

[0132] It should be noted that the time-varying transition resistance image only includes the time-varying transition resistance curve; it lacks scale axes and labels. The time-varying transition resistance values resulting from different fault causes can differ by orders of magnitude. Therefore, the time-varying transition resistance curve requires RGB normalization. For the time-varying transition resistance curve, the resistance values are first normalized. Then, the three RGB channels of the curve color are assigned values from [0,0,0] to [255,255,255] corresponding to the event transition resistance scale to achieve RGB normalization.

[0133] In an optional embodiment, the first identification model may be a classification model based on machine learning, which learns different types of fault characteristics through a training data set and is capable of effectively classifying them. The first identification model may be a model based on deep learning, which is suitable for processing image data and can automatically extract complex features for classification tasks. The first identification model may also be a system based on fuzzy logic, which can process imprecise or uncertain information and is suitable for fault diagnosis in complex environments.

[0134] In the embodiments of this application, Figure 2 The first identification model shown is constructed by a deep neural network unit, including an input layer, an intermediate hidden layer, and an output layer;

[0135] Specifically, the input layer includes a convolutional layer Conv with an input size of 224×224×3, a convolution kernel size of 7×7×1, a sliding step of 1, and a padding length of 3; a normalization layer Bacth Norm scales pixel values from [0,0,0] to [255,255,255] to [0,0,0] to [1,1,1] to accelerate training and improve model performance; and an activation function layer ReLU. The input layer receives and preprocesses image data and passes it to the subsequent intermediate hidden layers.

[0136] Specifically, the middle hidden layer includes multiple convolutional layers and pooling layers, which are used to extract image features and perform nonlinear transformations. The process includes: the output of the input layer is converted into a feature map through 8 ResNet18 residual blocks. The two convolutional layers in each residual block use a 3x3 convolution kernel with a stride of 1 and a padding length of 1 to keep the feature map size unchanged. The activation function is the ReLU function; the feature map is pooled using a global average pooling layer, reducing the height and width of each feature map to 1. The feature map is then expanded into a single row tensor based on the principle of parameter invariance.

[0137] Specifically, the output layer receives the output of the middle hidden layer as input and outputs the cause of the fault and its predicted probability. The process includes: the output layer includes a fully connected layer FC, which multiplies the single row tensor input tensor output by the middle hidden layer by the weight matrix and adds a bias term; the Softmax activation layer applies the Softmax activation function to the output of the fully connected layer to output the cause of the fault and its predicted probability, where the Softmax activation function is as follows:

[0138]

[0139] Among them, z j is the element in the input tensor;

[0140] Preferably, the fault cause identification model outputs the fault cause including but not limited to wildfire, external force damage, foreign matter, tree flash, ice flash, bird flash, etc. Figures 3 to 7 The following are typical time-varying transition resistance images for lightning strike, wildfire, pollution flashover, foreign object fault, and icing fault. Based on the characteristics of the input image, the first identification model can be used to obtain the cause of the fault and its predicted probability.

[0141] In the embodiment of the present application, pre-building and training a first recognition model includes the following steps:

[0142] Constructing a first recognition model, which is built through a deep neural network unit, including an input layer, an intermediate hidden layer, and an output layer;

[0143] Obtain historical fault samples, including obtaining line parameters of multiple historical fault line samples and fault recording data of the fault line;

[0144] Calculate and obtain the time-varying transition resistance value during line fault period;

[0145] Normalizing the time-varying transition resistance value and converting it into a time-varying transition resistance image as an image feature;

[0146] The time-varying transition resistance image of a single fault line sample is taken as a fault sample, where the fault label is the cause of the fault;

[0147] The acquired historical fault samples are divided into a training data set and a test set to train and test the original recognition model until convergence, thereby obtaining a first recognition model.

[0148] Preferably, the historical time-varying transition resistance image of a single fault is used as a fault sample, and the label is the cause of the fault. Several fault samples form a training data set and a test set to train and test the first identification model, preferably 70% and 30% to form a trained first identification model.

[0149] It should be noted that the above step S400 uses deep learning to perform in-depth analysis of fault characteristics, which can not only accurately identify the specific cause of the high-resistance grounding fault, but also enhance the accuracy and explainability of fault diagnosis, and provide strong support for operation and maintenance decisions.

[0150] Embodiment 3: This embodiment provides a system for identifying a high-resistance grounding fault in a transmission line, including:

[0151] A first calculation module is used to obtain a first result by performing a first judgment calculation on the line parameters and the recorded wave data of the fault line;

[0152] A second calculation module is used to calculate the time-varying transition resistance value during the fault period using an adaptive sliding data window based on the first result;

[0153] A first comparison module is used to compare the average value of the time-varying transition resistance with the first fault threshold to obtain a second result;

[0154] The fault identification module is used to input the grayscale image formed according to the time-varying transition resistance into the pre-built first identification model based on the second result, and output the cause of the fault and its predicted probability.

[0155] It should be noted that the technical solution of the system for identifying high-resistance grounding faults in transmission lines and the technical solution of the above-mentioned method for identifying high-resistance grounding faults in transmission lines belong to the same concept. For details not described in detail in the technical solution of the system for identifying high-resistance grounding faults in transmission lines in this embodiment, please refer to the description of the technical solution of the above-mentioned method for identifying high-resistance grounding faults in transmission lines.

[0156] The above-mentioned unit modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the corresponding operations of the above-mentioned modules.

[0157] This embodiment also provides an electronic device, comprising a processor, memory, a communication interface, a display screen, and an input device connected via a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The communication interface of the computer device is configured to communicate with an external terminal via wired or wireless communication, where the wireless communication may be achieved via Wi-Fi, a carrier network, NFC (near-field communication), or other technologies. When executed by the processor, the computer program implements a method for identifying a high-resistance grounding fault in a power transmission line. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen. The input device of the computer device may be a touch layer covering the display screen, or may be buttons, a trackball, or a touchpad provided on the computer device housing, or may be an external keyboard, touchpad, or mouse.

[0158] This embodiment further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method proposed in the above embodiment is implemented.

[0159] The storage medium proposed in this embodiment and the method proposed in the above embodiment belong to the same inventive concept. For technical details not fully described in this embodiment, please refer to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0160] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general-purpose hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this 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 (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the method of the embodiment of the present invention.

[0161] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. 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.

[0162] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiment of the present application can be implemented in various computer languages.

[0163] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes 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 device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0164] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising 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.

[0165] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing 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.

[0166] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. 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 present application.

[0167] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A method for identifying a high-resistance grounding fault in a transmission line, characterized in that: include: Obtaining a first result by performing a first judgment calculation on line parameters and recorded wave data of the fault line; Adaptively sliding the data window based on the first result to calculate the time-varying transition resistance value during the fault period; Comparing the average value of the time-varying transition resistance with the first fault threshold to obtain a second result; Based on the second result, a grayscale image formed according to the resistance value of the time-varying transition resistor is input into a pre-built first identification model, and the cause of the fault and its predicted probability are output.

2. The method for identifying a high-resistance grounding fault in a power transmission line according to claim 1, wherein: The first result includes at least a first fault type and a first fault distance of the fault line.

3. The method for identifying a high-resistance grounding fault in a power transmission line according to claim 2, wherein: The adaptive sliding data window calculates the time-varying transition resistance value during the fault period, including: determining a first sampling frequency and a first number of sliding cycles; Calculate the number of sampling points of data on both sides of the line within one system cycle based on the first sampling frequency; Calculate the total sliding length and sliding step length of the data windows on both sides of the line based on the first sliding cycle number and the number of sampling points; The transition resistance value at the corresponding moment of each data window is calculated, and the time-varying transition resistance value is obtained through the calculation results of all data windows.

4. The method for identifying a high-resistance grounding fault in a power transmission line according to claim 3, wherein: Acquiring the second result includes: Setting a first fault threshold; When the average value of the time-varying transition resistance is greater than or equal to the first fault threshold, it is determined that a high-resistance grounding fault occurs in the line; otherwise, it is determined that no high-resistance grounding fault occurs in the line.

5. The method for identifying a high-resistance grounding fault in a power transmission line according to claim 4, wherein: The time-varying transition resistor value is normalized to obtain a time-varying transition resistance curve, and the color RGB three-channel values of the time-varying transition resistance curve are corresponded to the time-varying transition resistor resistance scale to achieve RGB normalization, thereby obtaining a grayscale image with a standard size.

6. The method for identifying a high-resistance grounding fault in a power transmission line according to claim 5, wherein: The pre-built first recognition model includes an input layer, an intermediate hidden layer and an output layer; The input layer is used to receive and process the grayscale image of the standard size and pass it to the intermediate hidden layer, and the input layer includes a convolution layer, a normalization layer and an activation function layer; The intermediate hidden layer includes multiple convolutional layers and pooling layers for extracting image features and performing nonlinear transformations. The output of the input layer is converted into a feature map through the residual block of ResNet18. The feature map is pooled using a global average pooling layer. The height and width of each feature map are reduced and then expanded into a single row tensor form based on the principle of parameter invariance. The output layer takes the output of the intermediate hidden layer as input and outputs the cause of the fault and its predicted probability; the output layer includes a fully connected layer and an activation layer, the fully connected layer multiplies the single-row tensor input tensor output by the intermediate hidden layer by the weight matrix and adds a bias term; the activation layer applies a Softmax activation function to the output of the fully connected layer to output the cause of the fault and its predicted probability.

7. The method for identifying a high-resistance grounding fault in a power transmission line according to claim 3, wherein: Calculating the transition resistance value at the corresponding moment of each data window includes: Calculate the equivalent zero-sequence impedance of the power supplies on both sides of the line; Determining a special phase according to a first fault type of the fault line, and calculating new voltage and current phasors based on the acquired voltage and current phasors; According to the equivalent zero-sequence impedance of the power sources on both sides of the line and the calculated new voltage and current phasors, the transition resistance values are calculated respectively according to different first fault types of the fault line.

8. A transmission line high-resistance grounding fault identification system, applying the method according to any one of claims 1 to 7, characterized in that: include: A first calculation module is used to obtain a first result by performing a first judgment calculation on the line parameters and the recorded wave data of the fault line; A second calculation module, configured to calculate the time-varying transition resistance value during the fault period based on the first result adaptively sliding the data window; A first comparison module is configured to compare an average value of the time-varying transition resistance with a first fault threshold to obtain a second result; The fault identification module is used to input the grayscale image formed according to the resistance value of the time-varying transition resistor into a pre-built first identification model based on the second result, and output the cause of the fault and its predicted probability.

9. An electronic device comprising a memory and a processor, characterized in that: The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having computer-executable instructions stored thereon, characterized in that: When the computer executable instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images