A knowledge picture-based output line distributed fault diagnosis method

Abstract

This application discloses a distributed fault diagnosis method for transmission lines based on knowledge images, comprising the following steps: dividing the transmission line into several continuous monitoring intervals using monitoring terminals marked in a pre-constructed line topology knowledge image as boundary points; superimposing the current direction and amplitude information collected by the monitoring terminals onto the corresponding terminal positions in the line topology knowledge image to determine the target monitoring interval where the fault point is located; obtaining the location information of the fault point within the target monitoring interval; obtaining the traveling wave current waveform characteristics of the fault point and obtaining the fault cause corresponding to the fault point according to a pre-constructed mapping library; identifying the probability of occurrence of the fault cause at the current fault point location; determining whether the probability of occurrence is greater than a preset probability threshold, and if so, outputting the location information of the fault point and the fault cause. This application has the advantage of improving the accuracy of fault diagnosis.

Description

Technical Field

[0001] This application relates to the field of power transmission line fault diagnosis technology, and in particular to a distributed fault diagnosis method for output lines based on knowledge images. Background Technology

[0002] Transmission lines are a core component of the power system, and their safe and stable operation is directly related to the reliability of power supply. Transmission lines are widely distributed and operate in complex environments, making them susceptible to faults caused by various factors such as lightning strikes, tree obstructions, wildfires, wind deflection, and insulator breakdown. Among these, high-resistance grounding faults are particularly challenging to diagnose due to their inconspicuous fault characteristics and difficulty in identification.

[0003] Among existing transmission line fault diagnosis technologies, traveling wave-based diagnostic methods are widely used due to their fast response speed and high positioning accuracy. However, in existing methods, transmission line fault location and fault cause identification are independent of each other, failing to form a complete collaborative diagnostic process. Furthermore, the core information of each link is not integrated through knowledge images, making it difficult to further improve diagnostic efficiency and accuracy. Summary of the Invention

[0004] The main purpose of this application is to provide a distributed fault diagnosis method for output lines based on knowledge images, which aims to solve the technical problem of low diagnostic accuracy of existing transmission line fault diagnosis methods.

[0005] To achieve the above objectives, this application provides a distributed fault diagnosis method for output lines based on knowledge images, comprising the following steps: Using the monitoring terminals marked in the pre-constructed line topology knowledge image as the dividing points, the transmission line is divided into several continuous monitoring intervals; each monitoring interval is covered by two adjacent monitoring terminals. The current direction and amplitude information collected by the monitoring terminal are superimposed onto the corresponding terminal location in the line topology knowledge image to determine the target monitoring range where the fault point is located. Based on the knowledge image of the line topology, obtain the location information of the fault point within the target monitoring section; The traveling wave current waveform characteristics of the fault point are obtained, and the fault cause corresponding to the fault point is obtained according to the pre-built mapping library; wherein, the mapping library is a database of mapping relationships between waveform characteristics and fault causes; Identify the probability of the cause of the fault occurring at the current fault location; Determine if the probability of occurrence is greater than a preset probability threshold. If yes, output the location information of the fault point and the cause of the fault. If no, return to superimpose the current direction and amplitude information collected by the monitoring terminal onto the terminal location corresponding to the line topology knowledge image to determine the target monitoring range where the fault point is located.

[0006] Optionally, the step of acquiring the traveling wave current waveform characteristics of the fault point and acquiring the fault cause corresponding to the fault point according to a preset mapping library includes: Obtain the morphological characteristics of the traveling wave current waveform at the fault point; these morphological characteristics include polarity P, rise edge steepness k, half-peak time T, amplitude I, and pre-discharge coefficient K. P The polarity is determined by the initial positive and negative directions of the traveling wave current waveform, defined as +1 for positive and -1 for negative. The rising edge steepness k represents the average rate of change of the traveling wave current from 0 to 10% to 90% of its peak amplitude. The half-peak time T represents the time required for the traveling wave current to decrease from its peak amplitude to 50% of its peak amplitude. The amplitude I is taken as the maximum peak amplitude of the traveling wave current, and the pre-discharge coefficient K... P Indicates the intensity of the pre-discharge phenomenon; Based on morphological features, generate real-time fault waveform images; The similarity of real-time fault waveform images with pre-built waveform standard knowledge images is compared to obtain the fault cause corresponding to the fault point.

[0007] Optionally, the step of comparing the real-time fault waveform image with a pre-constructed waveform standard knowledge image to obtain the fault cause corresponding to the fault point includes: Combining morphological features into feature vectors =[P, k, T, I, K P ]; Obtaining feature vectors Compared with the standard feature vectors of various faults in the mapping library Euclidean distance d i d i The expression is: ; In the formula, P i k i T i I i K Pi These are the standard characteristic parameters corresponding to the fault type in the waveform standard knowledge image; According to the Euclidean distance d i To obtain the specific cause of the fault at the fault point.

[0008] Optionally, the method based on Euclidean distance d i To obtain the specific cause of the fault at the fault point, including: Select multiple Euclidean distances d i The minimum distance value d in the data min ; Obtain the preset threshold σ marked in the fault feature mapping knowledge image, and set the minimum distance value d. min Compare with a preset threshold σ; If d min If d ≤ σ, the corresponding fault type will be output as the specific fault cause, and the real-time fault waveform image will be overlaid and compared with the waveform standard knowledge image of the corresponding fault type to generate a feature difference knowledge image, and the difference points will be marked; if d min If the value is greater than σ, then an unknown fault will be output.

[0009] Optionally, the probability of identifying the cause of the fault occurring at the current fault location includes: Obtain the line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient corresponding to the current fault location; Input the line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient into the preset probability prediction model to obtain the probability of the fault cause occurring at the current fault location.

[0010] Optionally, the step of inputting the line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient into a preset probability prediction model to obtain the probability of the fault cause occurring at the current fault location includes: The line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient are input into a preset probability prediction model to obtain the initial probability R; the expression of the probability prediction model is: R = R0·α / 5·β·γ; In the formula, R0 is the inherent probability of the fault type, α is the line tower type coefficient, with a value range of 1 to 5, β is the environmental adaptability coefficient, with a value range of 0 to 1, and γ is the equipment aging coefficient, with a value range of 1 to 2. Let the initial probability R and the minimum distance d be... min Input a preset probability correction model to obtain the final probability of the fault cause occurring at the current fault location.

[0011] Optionally, the expression for the probability correction model is: R'=R(1-d min / σ); In the formula, R' is the probability of the final occurrence.

[0012] Optionally, the expression for the equipment aging factor γ is: γ=W1·γ1+W2·γ2+W3·γ3+W4·γ4; In the formula, W1, W2, W3, and W4 are the first weighting coefficient, the second weighting coefficient, the third weighting coefficient, and the fourth weighting coefficient, respectively; γ1 is the line service life correction coefficient, with a value range of 1 to 2; γ2 is the equipment maintenance frequency correction coefficient, with a value range of 0.8 to 1.2; γ3 is the environmental corrosion degree correction coefficient, with a value range of 0.9 to 1.3; and γ4 is the line load fluctuation coefficient correction coefficient, with a value range of 0.9 to 1.1.

[0013] Optionally, the pole and tower types include suspended vertical line poles, suspended angle poles, tension straight line poles, tension angle poles, terminal poles, and Z-type poles, wherein the line pole and tower type coefficient α corresponding to suspended vertical line poles is 1, the line pole and tower type coefficient α corresponding to suspended angle poles is 2, the line pole and tower type coefficient α corresponding to tension straight line poles is 3, the line pole and tower type coefficient α corresponding to tension angle poles is 4, and the line pole and tower type coefficient α corresponding to terminal poles and Z-type poles is 5.

[0014] Optionally, obtaining the location information of the fault point within the target monitoring interval based on the line topology knowledge image includes: Obtain the line length L between two monitoring terminals corresponding to the target monitoring interval marked in the line topology knowledge image; The times t1 and t2 corresponding to the first arrival of the traveling wave current at the target monitoring interval are obtained respectively; Obtain the distance x between the fault point and one of the monitoring terminals in the target monitoring interval; where x = (Lv·Δt) / 2, v is the propagation speed of the traveling wave in the transmission line, and Δt = |t1-t2|. Based on the distance x and the coordinates of the corresponding monitoring terminal in the target monitoring interval marked in the line topology knowledge image, the coordinate location information of the fault point is obtained.

[0015] The beneficial effects that this application can achieve are as follows: This application, by combining visualization aids with line topology knowledge images, employs a two-step positioning strategy: first interval positioning, then precise positioning. This allows for rapid location of fault points within the target monitoring interval. Then, based on the traveling wave current waveform characteristics of the fault point and a constructed "waveform feature-fault cause" mapping library, multi-dimensional waveform feature extraction accurately distinguishes various fault causes. Furthermore, it verifies the probability of the identified fault cause occurring at the current fault point location. By determining if the probability exceeds a preset probability threshold, the location information and fault cause are output; otherwise, a secondary detection is triggered, indicating a detection error. This avoids misjudgments where waveform features match correctly, but the probability of the fault occurring at that location is extremely low. Therefore, this application, by combining line topology knowledge images, improves the accuracy of fault point location and, by adding a verification step, enhances the accuracy of fault cause detection, thus comprehensively improving the accuracy of fault diagnosis. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0017] Figure 1 This is a flowchart illustrating a distributed fault diagnosis method for output lines based on knowledge images, as described in an embodiment of this application.

[0018] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0020] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationship and movement of each component in a specific posture. If the specific posture changes, the directional indication will also change accordingly.

[0021] In this application, unless otherwise expressly specified and limited, the terms "connection," "fixed," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral part; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal communication of two components or the interaction between two components, unless otherwise expressly limited. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0022] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the meaning of "and / or" throughout the text includes three parallel solutions; for example, "A and / or B" includes solution A, solution B, or a solution where both A and B are satisfied simultaneously. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0023] Example Reference Figure 1 This embodiment provides a distributed fault diagnosis method for output lines based on knowledge images, including the following steps: Step S10: Using the monitoring terminals marked in the pre-constructed line topology knowledge image as the dividing points, the transmission line is divided into several continuous monitoring intervals; each monitoring interval is covered by two adjacent monitoring terminals. In this step, a pre-constructed line topology knowledge image is presented visually, showing the overall route of the transmission line, tower locations, monitoring terminal deployment points and numbers, and satellite synchronization clock parameters (such as GPS and BeiDou synchronization clocks) for each terminal. It also labels the line length and section number between adjacent monitoring terminals, thus forming a standardized line topology knowledge image, which is stored on each monitoring terminal and the remote platform. Based on the auxiliary role of the line topology knowledge image, section positioning errors can be reduced, and positioning accuracy improved. This avoids problems in existing technologies, such as terminal number confusion and misjudgment of section divisions due to the lack of visual references, leading to section positioning deviations (such as misjudging adjacent sections as fault sections), affecting subsequent diagnostic processes. In this step, the line topology knowledge image clearly labels the number, location, distance between adjacent terminals, and section number of each terminal. Current information corresponds one-to-one with terminal locations, effectively avoiding number confusion and section misjudgment. The section positioning accuracy is extremely high, ensuring that subsequent precise positioning is only performed on fault sections, avoiding invalid calculations and improving overall diagnostic efficiency.

[0024] Step S20: Overlay the current direction and amplitude information collected by the monitoring terminal onto the terminal location corresponding to the line topology knowledge image to determine the target monitoring range where the fault point is located; After a fault occurs, each monitoring terminal collects fault current or traveling wave current signals in real time, extracting the direction (forward / reverse, based on the line power flow direction) and amplitude information of the current. The collected current direction and amplitude information are superimposed onto the corresponding terminal position in the line topology knowledge image. By comparing the differences in current direction and amplitude between adjacent monitoring terminals in the line topology knowledge image, and combining the line interval division rules marked in the image, the specific monitoring interval where the fault point is located can be quickly determined. At the same time, the fault interval range is initially marked in the topology knowledge image to facilitate subsequent accurate positioning.

[0025] The core principle of this step is that the traveling wave current generated at the fault point propagates to both ends, and the monitoring terminals at both ends of the monitoring interval will collect obvious current changes. The current direction is related to the position of the fault point relative to the terminal. Combined with the amplitude attenuation law (the amplitude attenuates with distance during the propagation of the traveling wave), and with the visualization assistance of the line topology knowledge image, the monitoring interval where the fault point is located can be quickly located.

[0026] Step S30: Based on the line topology knowledge image, obtain the location information of the fault point within the target monitoring section, specifically including: Obtain the line length L between two monitoring terminals corresponding to the target monitoring interval marked in the line topology knowledge image; The times t1 and t2 corresponding to the first arrival of the traveling wave current at the target monitoring interval are obtained respectively; Obtain the distance x between the fault point and one of the monitoring terminals in the target monitoring interval; where x = (Lv·Δt) / 2, v is the propagation speed of the traveling wave in the transmission line (generally taken as 0.295 km / μs), and Δt = |t1-t2|. Based on the distance x and the coordinates of the corresponding monitoring terminal in the target monitoring interval marked in the line topology knowledge image, the coordinate location information of the fault point is obtained.

[0027] In this step, the precise location of the fault point can be calculated using the propagation characteristics of traveling waves. For example, consider a 100km long 220kV high-voltage transmission line with three monitoring terminals (Terminal A, Terminal B, and Terminal C) deployed along the line. A line topology knowledge image is pre-constructed, marking the locations and numbers of Terminal A (starting point, coordinates: N30°12′, E118°45′), Terminal B (midpoint, 50km, coordinates: N30°15′, E118°50′), and Terminal C (ending point, 100km, coordinates: N30°18′, E118°55′), and the lengths of each interval (50km for interval AB and 50km for interval BC). If the fault point is detected to be within interval AB, the first arrival time of the traveling wave at Terminal A is extracted (t1 = 100μs), and the first arrival time of the traveling wave at Terminal B is extracted (t2 = 150μs). The time difference Δt = |100 - 150| = 50μs is calculated, and the lengths of interval AB marked in the line topology knowledge image are retrieved. Substituting the above, we can calculate x = (50 - 0.295 · 50) / 2 = 17.625 km. Then, by combining this distance x with the coordinate information of terminal A, we can convert it into the coordinate location information of the corresponding fault point and achieve precise positioning.

[0028] Step S40: Obtain the traveling wave current waveform characteristics at the fault point, and obtain the corresponding fault cause based on the pre-built mapping library; wherein, the mapping library is a database of mapping relationships between waveform characteristics and fault causes; specifically including: Obtain the morphological characteristics of the traveling wave current waveform at the fault point; these morphological characteristics include polarity P, rise edge steepness k, half-peak time T, amplitude I, and pre-discharge coefficient K. P The polarity is determined by the initial positive and negative directions of the traveling wave current waveform, defined as +1 for positive and -1 for negative. The rising edge steepness k represents the average rate of change of the traveling wave current from 0 to 10% to 90% of its peak amplitude. The half-peak time T represents the time required for the traveling wave current to decrease from its peak amplitude to 50% of its peak amplitude. The amplitude I is taken as the maximum peak amplitude of the traveling wave current, and the pre-discharge coefficient K... P Indicates the intensity of the pre-discharge phenomenon; Based on morphological features, generate real-time fault waveform images; The similarity of real-time fault waveform images with pre-built waveform standard knowledge images is compared to obtain the fault cause corresponding to the fault point.

[0029] In this step, the morphological features of the traveling wave current waveform at the fault point are extracted. These features include the core and typical polarity P, rise steepness k, half-peak time T, amplitude I, and pre-discharge coefficient K. P Features such as the pre-discharge coefficient K P=I' / I, where I' is the maximum current value during the pre-discharge phase (when there is no pre-discharge phenomenon, K...). P =0), pre-discharge coefficient K P To characterize the intensity of the pre-discharge phenomenon, real-time fault waveform images can be generated based on the extracted five types of core waveform features. By comparing the similarity with pre-constructed waveform standard knowledge images, the fault cause corresponding to the fault point can be accurately identified.

[0030] It should be noted that the above-mentioned waveform standard knowledge images are designed for various faults such as lightning strikes, tree obstructions, wildfires, wind deflection, and insulator breakdown. Through a large number of engineering measurements and simulation experiments, standard traveling wave current waveforms for various faults were collected. The standard waveforms and waveform characteristic parameters (polarity, rise steepness, half-peak time, amplitude, standard values ​​and deviation ranges of pre-discharge coefficient) for each type of fault were integrated into a visual image, labeled with fault type and characteristic parameter descriptions, forming a traveling wave waveform standard knowledge image, which is stored according to fault type.

[0031] It should also be noted that pre-discharge refers to the current fluctuations caused by weak ionization and partial discharge in the fault medium (such as trees or the surface of insulators) before the fault occurs (i.e., before the peak value of the traveling wave appears). Its waveform characteristics are significantly different from line noise, and the core criteria for judgment are as follows: Time correlation: Pre-discharge only occurs in a "specific time period before the peak of the traveling wave", that is, from the initial moment of the fault to the peak moment of the traveling wave, and is concentrated in 5~20μs before the peak moment of the traveling wave (the specific range can be calibrated according to the line voltage level). Fluctuations outside this time period are not considered pre-discharge. Amplitude correlation: The amplitude of the pre-discharge current fluctuation is 1% to 5% of the peak amplitude of the traveling wave (below 1%, it can be judged as noise; above 5%, it belongs to the initial stage of fault discharge and is still included in the pre-discharge), and the fluctuation is continuous (at least 3 consecutive fluctuation cycles occur, and the amplitude fluctuation of each cycle does not exceed 20% of its own amplitude). Waveform correlation: The pre-discharge waveform exhibits the characteristics of "small amplitude, continuous, and smooth fluctuations", without sharp pulses, and the direction of fluctuation is consistent with the polarity of the subsequent traveling wave peak (e.g., if the traveling wave peak is positive, the pre-discharge fluctuation is also positive); while noise is mostly random sharp pulses, without a fixed direction, and without a continuous fluctuation pattern.

[0032] As an optional implementation, the step of comparing the real-time fault waveform image with a pre-constructed waveform standard knowledge image to obtain the fault cause corresponding to the fault point includes: Combining morphological features into feature vectors =[P, k, T, I, K P ]; Obtaining feature vectors Compared with the standard feature vectors of various faults in the mapping library Euclidean distance d i d i The expression is: ; In the formula, P i k i T i I i K Pi These are the standard feature parameters corresponding to the fault type in the waveform standard knowledge image (which can be quickly queried through the waveform standard knowledge image); According to the Euclidean distance d i To obtain the specific cause of the fault at the fault point.

[0033] In this embodiment, when identifying the cause of a fault, morphological features can be combined into a feature vector. =[P, k, T, I, K P The system matches corresponding waveform standard knowledge images based on a mapping library, and then extracts the corresponding fault standard feature vectors. =[P i k i T i I i K Pi The Euclidean distance d is obtained using the Euclidean distance calculation method. i This is used to indirectly characterize the degree of difference between the fault waveform feature vector extracted in real time and the standard feature vectors of various faults, i.e., the Euclidean distance d. i The smaller the value, the greater the matching degree between the two, which means that the fault point can be accurately matched with the corresponding fault cause.

[0034] As an optional implementation, the method based on Euclidean distance d i To obtain the specific cause of the fault at the fault point, including: Select multiple Euclidean distances d i The minimum distance value d in the data min ; Obtain the preset threshold σ marked in the fault feature mapping knowledge image, and set the minimum distance value d. min Compare with a preset threshold σ; If d min If d ≤ σ, the corresponding fault type will be output as the specific fault cause, and the real-time fault waveform image will be overlaid and compared with the waveform standard knowledge image of the corresponding fault type to generate a feature difference knowledge image, and the difference points will be marked; if d min If the value is greater than σ, then an unknown fault will be output.

[0035] In this embodiment, since it is necessary to distinguish between various faults such as lightning strikes, tree obstructions, wildfires, and wind-induced deviation (including high-resistance grounding faults), the waveform characteristics of these various faults overlap but have fundamental differences. Therefore, the minimum distance value d is used. min The system can filter out the type that best matches the real-time fault characteristics from multiple candidate fault types, avoiding misjudgments caused by feature overlap, and is particularly effective in distinguishing high-resistance grounding faults from conventional faults (such as tree obstruction and wind deflection); then, based on the preset threshold σ marked in the fault feature mapping knowledge image, when d min If the value is less than or equal to σ, the corresponding fault type will be output as the specific fault cause. Otherwise, an unknown fault will be output. Manual review can be triggered, and the manual reviewer can compare the waveform standard knowledge images and fault feature mapping knowledge images to help determine the fault cause.

[0036] It should be noted that the above-mentioned fault feature mapping knowledge image integrates the standard waveform feature vectors of various faults, Euclidean distance calculation rules, and preset threshold σ (calibrated according to engineering measured data, generally with a value of 0.1~0.3) into a visual image, presenting the correspondence between "fault type - waveform feature - difference judgment standard", thus forming a fault feature mapping knowledge image, which can be used as an intuitive reference for feature matching.

[0037] In summary, this step employs multi-dimensional feature vector matching, combined with Euclidean distance to calculate the degree of difference, and uses waveform standard knowledge images for intuitive comparison, providing double verification of the fault cause, which can improve the fault identification accuracy to ≥95%. Furthermore, by using a preset threshold σ, suspected faults can be manually reviewed, further reducing the false positive rate and ensuring the reliability of the diagnostic results. The fault feature mapping knowledge image visually presents the correspondence between "fault type - waveform features - degree of difference judgment criteria," and the waveform standard knowledge image annotates the standard waveforms and feature parameters of various faults, allowing for quick querying and comparison without tedious manual calculations and comparisons. Feature matching time is ≤50ms, significantly improving fault identification efficiency.

[0038] Step S50: Identify the probability of the fault cause occurring at the current fault location, specifically including: Obtain the line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient corresponding to the current fault location; Input the line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient into the preset probability prediction model to obtain the probability of the fault cause occurring at the current fault location.

[0039] In this step, the line tower type coefficient can represent the different fault susceptibility weights corresponding to different tower types. The higher the weight, the stronger the susceptibility. The environmental adaptability coefficient is matched according to the fault type. For example, tree obstruction faults correspond to the surrounding tree density coefficient, and wildfire faults correspond to the surrounding vegetation coverage coefficient. The larger the coefficient, the higher the probability of fault occurrence. The equipment aging coefficient represents the degree of equipment aging. The higher the degree of aging, the higher the probability of fault. Therefore, this embodiment combines the three core parameters of line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient to input the preset probability prediction model, so as to quantitatively calculate the probability of the fault cause occurring at the current fault location, which improves the data reference basis for accurately verifying the fault cause.

[0040] As an optional implementation, the step of inputting the line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient into a preset probability prediction model to obtain the probability of the fault cause occurring at the current fault location includes: The line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient are input into a preset probability prediction model to obtain the initial probability R; the expression of the probability prediction model is: R = R0·α / 5·β·γ; In the formula, R0 is the inherent probability of the fault type, α is the line tower type coefficient, with a value range of 1 to 5, β is the environmental adaptability coefficient, with a value range of 0 to 1, and γ is the equipment aging coefficient, with a value range of 1 to 2. Let the initial probability R and the minimum distance d be... min Input a preset probability correction model to obtain the final probability of the fault cause occurring at the current fault location; the expression for the probability correction model is: R'=R(1-d min / σ); In the formula, R' is the probability of the final occurrence.

[0041] In this embodiment, the inherent probability R0 of the fault type is obtained from historical fault data statistics (e.g., the inherent probability of lightning strike fault R0 = 0.35, the inherent probability of tree obstacle fault R0 = 0.28, which can be queried through fault feature mapping knowledge images, with a value range of 0.2~0.5). α / 5 can standardize the tower type weight to a coefficient between 0 and 1 to eliminate the influence of dimensions. The environmental adaptability coefficient β reflects the degree of adaptability between the fault point environment and the fault type, and can be calibrated by combining the line topology knowledge image with on-site measured data. The equipment aging coefficient γ reflects the impact of the operating status of the line equipment at the fault point on the occurrence of the fault. After collecting the above parameters and substituting them into the probability prediction model, the initial probability R can be calculated. For example, taking tree obstacle as an example, the inherent probability of tree obstacle fault R0 = 0.28. The core parameters α = 3, β = 0.8, and γ = 1.5 are collected. Substituting them into the above formula, R≈0.2 can be calculated. At the same time, this embodiment also incorporates the minimum distance value d. min The calculation results are corrected (at this time d) min ≤σ), which further improves the accuracy of the calculation of the occurrence probability R'.

[0042] It should be noted that the aforementioned mapping library is constructed as follows: through extensive engineering measurements and simulations, traveling wave waveform data of various faults such as lightning strikes, tree obstructions, wildfires, wind deflection, and insulator breakdown are collected. Standard waveform feature parameters (mean ± deviation) for each type of fault are extracted, and a standardized feature mapping table is established. The waveform standard knowledge image and the fault feature mapping knowledge image are updated synchronously, supporting subsequent online updates (adding fault types, optimizing feature parameters, and knowledge image content). At the same time, historical occurrence data of various faults are statistically analyzed to determine the inherent probability R0 of different fault types, which is then added to the fault feature mapping knowledge image to provide basic data support for fault location probability analysis.

[0043] As an optional implementation method, the expression for the equipment aging factor γ is: γ=W1·γ1+W2·γ2+W3·γ3+W4·γ4; In the formula, W1, W2, W3, and W4 are the first weighting coefficient, the second weighting coefficient, the third weighting coefficient, and the fourth weighting coefficient, respectively; γ1 is the line service life correction coefficient, with a value range of 1 to 2; γ2 is the equipment maintenance frequency correction coefficient, with a value range of 0.8 to 1.2; γ3 is the environmental corrosion degree correction coefficient, with a value range of 0.9 to 1.3; and γ4 is the line load fluctuation coefficient correction coefficient, with a value range of 0.9 to 1.1.

[0044] In this embodiment, the equipment aging coefficient γ comprehensively considers the effects of line operating years, equipment maintenance frequency, environmental corrosion level, and line load fluctuations. These influencing factors are quantified and converted into corresponding correction coefficients, balancing scientific rigor with engineering feasibility and improving calculation accuracy. Among them, the line operating years correction coefficient γ1, as a core parameter, reflects the impact of line operating time on equipment aging. Its corresponding first weighting coefficient W1 can be set to 0.5. The longer the operating years, the more severe the aging of the equipment (insulators, conductors, tower accessories, etc.), and the higher the value of γ1. For example, specific values ​​are as follows: Operating years ≤ 10 years (newly commissioned / young and middle-aged lines): γ1=1 (equipment in good condition, low degree of aging). 10 years < service life ≤ 20 years (middle-aged line): γ1 = 1.4 (the equipment shows slight aging, and the risk of failure increases slightly). 20 years < service life ≤ 30 years (old lines): γ1 = 1.8 (significant equipment aging, significantly increased risk of failure). For lines that have been in operation for more than 30 years (overdue lines): γ1=2 (the equipment is severely aged and the risk of failure is extremely high).

[0045] The aforementioned equipment maintenance frequency correction coefficient γ2 (whose corresponding second weighting coefficient W2 can be 0.2) reflects the inhibitory effect of routine equipment maintenance on aging. The higher the maintenance frequency, the slower the equipment aging, and the lower the value of γ2. For example, the specific values ​​are as follows: Maintenance ≥ 2 times per year (high-frequency maintenance): γ2 = 0.8 (effectively inhibits aging and reduces the risk of failure); Annual maintenance (routine maintenance): γ2=1 (normal maintenance, aging rate follows normal pattern); Maintenance every 2 years (low-frequency maintenance): γ2=1.1 (untimely maintenance accelerates aging); Maintenance every 3 years or more (very infrequent maintenance): γ2=1.2 (severely insufficient maintenance, significantly accelerated aging).

[0046] The aforementioned environmental corrosion correction coefficient γ3 (whose corresponding third weighting coefficient W3 can be 0.2) reflects the corrosive effect of the surrounding environment on the equipment. The higher the corrosion level, the faster the equipment ages, and the higher the value of γ3. This can be combined with the environmental type marked in the line topology knowledge image. For example, the specific values ​​are as follows: Open, non-corrosive environments (e.g., plains, suburbs, free from industrial pollution, salinity, humidity, etc.): γ3=0.9 (extremely weak corrosion, slow aging). In general environments (e.g., ordinary towns, non-industrial areas, etc.): γ3=1 (slight corrosion, normal aging rate). Moderately corrosive environments (e.g., suburban industrial areas, slightly saline-alkali land, humid mountainous areas, etc.): γ3=1.2 (significant corrosion, accelerated aging); Severely corrosive environments (such as coastal areas, heavy industrial areas, high salinity and alkalinity areas): γ3=1.3 (strong corrosion and extremely fast aging).

[0047] The aforementioned load fluctuation correction factor γ4 (whose corresponding fourth weighting factor W4 can be 0.1) reflects the impact of line load fluctuations on equipment aging. The greater the load fluctuation, the more severe the equipment wear and tear, and the faster the aging process. Therefore, a higher value for γ4 is required. For example, specific values ​​are as follows: Load fluctuation ≤20% (stable load): γ4=0.9 (stable load, low equipment wear and tear, slow aging); 20% < load fluctuation ≤ 40% (normal fluctuation): γ4 = 1 (load fluctuation is normal, aging rate is in line with normal). Load fluctuation > 40% (severe fluctuation): γ4 = 1.1 (severe load fluctuation, high equipment wear and tear, accelerated aging).

[0048] As an optional implementation, the pole types include suspended vertical line poles, suspended angle poles, tension straight line poles, tension angle poles, terminal poles, and Z-type poles. Among them, the line pole type coefficient α corresponding to the suspended vertical line pole is 1, the line pole type coefficient α corresponding to the suspended angle pole is 2, the line pole type coefficient α corresponding to the tension straight line pole is 3, the line pole type coefficient α corresponding to the tension angle pole is 4, and the line pole type coefficient α corresponding to the terminal pole and Z-type pole is 5.

[0049] In this embodiment, when setting the coefficient α according to the tower type, since suspended vertical line towers are mainly used on straight sections of transmission lines, they have a simple structure, uniform stress, and no special protective design. They are mostly used in flat, open areas (without trees or hills blocking the view) and areas with weak lightning activity, so their fault susceptibility is the lowest. The main fault type is insulator aging and breakdown, and the probability of faults such as lightning strikes, tree obstruction, and wind deflection is extremely low. Therefore, the corresponding coefficient α is set to the minimum value of 1. Suspended angle towers are used at the corners of transmission lines (angle ≤ 30°), and the stress is slightly lower. Compared to suspended vertical transmission line towers, tension straight-line towers are more complex. The line route involves slight bends, and there may be some tree obstruction in the surrounding area. The risk of lightning strikes is slightly higher than that of suspended vertical towers, therefore their fault susceptibility is lower. The main fault types are wind deflection and minor tree obstruction. The probability of lightning strikes is low, and its corresponding coefficient α can be taken as 2. Tension straight-line towers are used on straight sections of transmission lines. They can withstand greater longitudinal tension and have higher structural strength than suspended towers. They are mostly used in areas where lines cross small rivers or roads. The surrounding environment is of moderate complexity, and there is a certain risk of tree obstruction and wind deflection. Its fault susceptibility is moderate, including tree obstruction, wind deflection, and insulator breakdown. The risk of lightning strikes is moderate, and its corresponding coefficient α can be taken as 3. Tension angle towers are used at the corners of transmission lines (angles of 30°~60°), with a double-circuit arrangement. They have a complex structure and concentrated stress, and are often used in areas with complex terrain (hilly or mountainous edges), where there are many trees and hills nearby, and frequent lightning activity. The probability of wind deflection, tree obstruction, and lightning strikes is high, resulting in a high fault susceptibility. The double-circuit arrangement increases the probability of fault occurrence, and the risk of wind deflection at the corner is significantly increased, as is the risk of lightning strikes. The probability of lightning strikes is higher than that of single-circuit towers, so its corresponding coefficient α can be taken as 4; terminal towers are used at the beginning and end of transmission lines, connecting substations, and are subject to the greatest stress and have the most complex structure; Z-type towers are mostly used in UHV lines in mountainous areas. Although their lightning protection performance is better than that of traditional T-type towers, the mountainous environment is complex, lightning activity is frequent, and the risk of tree obstruction and wildfire is high. Both types of towers are in areas with high failure incidence; the fault susceptibility is the highest, and various faults such as lightning strikes, tree obstruction, wildfire, wind deflection, and insulator breakdown can occur, and the consequences of the faults are more serious. Therefore, its corresponding coefficient α can be taken as the maximum value of 5.

[0050] In summary, this step, in addition to identifying the cause of the fault, adds a verification step to assess the probability of the fault occurring at the current fault location. This avoids potential misjudgments where "the waveform feature matches correctly, but the probability of the fault occurring at that location is extremely low" (e.g., misjudging a fault as a tree obstruction in a treeless area). Therefore, by calculating the probability of the fault cause occurring at the fault location, this step can effectively verify the reliability of the diagnostic results. If the probability is too low, it can trigger secondary matching and manual review to avoid misjudgments and further improve the accuracy of the diagnostic results, making it particularly suitable for fault diagnosis in complex environments. Furthermore, the parameters required for calculating the fault location probability (tower type, relevant environmental parameters, etc.) can be extracted from line topology knowledge images and fault feature mapping knowledge images, eliminating the need for additional manual data collection and making parameter extraction convenient.

[0051] Step S60: Determine whether the probability of occurrence is greater than the preset probability threshold. If yes, output the location information of the fault point and the cause of the fault. If no, return to superimpose the current direction and amplitude information collected by the monitoring terminal onto the terminal location corresponding to the line topology knowledge image to determine the target monitoring range where the fault point is located.

[0052] In this step, by setting a probability threshold, if the calculated probability of occurrence is greater than the probability threshold, it can be determined that the probability of the cause of the fault occurring at this location is high, and the diagnosis result is reliable. Otherwise, the test is repeated. If the second test result still determines that the probability of the cause of the fault occurring at this location is low, the manual review procedure is initiated.

[0053] In summary, this application has the following advantages over the prior art: 1. Significantly improved diagnostic accuracy and reliability: Through dual positioning of "interval positioning + precise positioning" and dual identification of "multi-dimensional feature matching + probability verification", combined with the intuitive assistance and verification mechanism of knowledge images, the positioning error and fault misjudgment rate are effectively reduced. The positioning error is ≤50m, the fault identification accuracy rate is ≥95%, and the probability verification further improves the reliability of the diagnostic results, meeting the actual needs of engineering.

[0054] 2. Strong engineering adaptability and wide applicability: This method does not require large-scale modification of existing transmission lines. The monitoring terminal adopts a satellite synchronous clock, which is compatible with the existing monitoring system. Various knowledge images and mapping libraries support online updates and can be adapted to transmission lines of different levels, terrains and operating environments. It can be widely used for various fault diagnosis of high voltage and ultra-high voltage transmission lines, including high resistance grounding faults. The engineering deployment cost is low and the compatibility is strong.

[0055] 3. High ease of use and reduced operation and maintenance costs: The visualization feature of knowledge images lowers the professional threshold for operation and maintenance personnel. Each step is simple to operate and does not require complex professional knowledge and operation skills. The remote output and display function reduces the workload of on-site operation and maintenance, shortens the emergency repair time, reduces operation and maintenance costs, and improves the level of intelligence of transmission line operation and maintenance.

[0056] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A distributed fault diagnosis method for output lines based on knowledge images, characterized in that, Includes the following steps: Using the monitoring terminals marked in the pre-constructed line topology knowledge image as the dividing points, the transmission line is divided into several continuous monitoring intervals; each monitoring interval is covered by two adjacent monitoring terminals. The current direction and amplitude information collected by the monitoring terminal are superimposed onto the corresponding terminal location in the line topology knowledge image to determine the target monitoring range where the fault point is located. Based on the knowledge image of the line topology, obtain the location information of the fault point within the target monitoring section; The traveling wave current waveform characteristics of the fault point are obtained, and the fault cause corresponding to the fault point is obtained according to the pre-built mapping library; wherein, the mapping library is a database of mapping relationships between waveform characteristics and fault causes; Identify the probability of the cause of the fault occurring at the current fault location; Determine if the probability of occurrence is greater than a preset probability threshold. If yes, output the location information of the fault point and the cause of the fault. If no, return to superimpose the current direction and amplitude information collected by the monitoring terminal onto the terminal location corresponding to the line topology knowledge image to determine the target monitoring range where the fault point is located.

2. The distributed fault diagnosis method for output lines based on knowledge images as described in claim 1, characterized in that, The process of acquiring the traveling wave current waveform characteristics of the fault point and obtaining the corresponding fault cause according to a preset mapping library includes: Obtain the morphological characteristics of the traveling wave current waveform at the fault point; these morphological characteristics include polarity P, rise edge steepness k, half-peak time T, amplitude I, and pre-discharge coefficient K. P The polarity is determined by the initial positive and negative directions of the traveling wave current waveform, defined as +1 for positive and -1 for negative. The rising edge steepness k represents the average rate of change of the traveling wave current from 0 to 10% to 90% of its peak amplitude. The half-peak time T represents the time required for the traveling wave current to decrease from its peak amplitude to 50% of its peak amplitude. The amplitude I is taken as the maximum peak amplitude of the traveling wave current, and the pre-discharge coefficient K... P Indicates the intensity of the pre-discharge phenomenon; Based on morphological features, generate real-time fault waveform images; The similarity of real-time fault waveform images with pre-built waveform standard knowledge images is compared to obtain the fault cause corresponding to the fault point.

3. The distributed fault diagnosis method for output lines based on knowledge images as described in claim 2, characterized in that, The step of comparing the real-time fault waveform image with a pre-constructed waveform standard knowledge image to obtain the fault cause corresponding to the fault point includes: Combining morphological features into feature vectors =[P, k, T, I, K P ]; Obtaining feature vectors Compared with the standard feature vectors of various faults in the mapping library Euclidean distance d i d i The expression is: ； In the formula, P i k i T i I i K Pi These are the standard characteristic parameters corresponding to the fault type in the waveform standard knowledge image; According to the Euclidean distance d i To obtain the specific cause of the fault at the fault point.

4. The distributed fault diagnosis method for output lines based on knowledge images as described in claim 3, characterized in that, The Euclidean distance d i To obtain the specific cause of the fault at the fault point, including: Select multiple Euclidean distances d i The minimum distance value d in the data min ; Obtain the preset threshold σ marked in the fault feature mapping knowledge image, and set the minimum distance value d. min Compare with a preset threshold σ; If d min If d ≤ σ, the corresponding fault type will be output as the specific fault cause, and the real-time fault waveform image will be overlaid and compared with the waveform standard knowledge image of the corresponding fault type to generate a feature difference knowledge image, and the difference points will be marked; if d min If the value is greater than σ, then an unknown fault will be output.

5. The distributed fault diagnosis method for output lines based on knowledge images as described in claim 4, characterized in that, The probability of identifying the cause of the fault occurring at the current fault location includes: Obtain the line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient corresponding to the current fault location; Input the line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient into the preset probability prediction model to obtain the probability of the fault cause occurring at the current fault location.

6. The distributed fault diagnosis method for output lines based on knowledge images as described in claim 5, characterized in that, The process of inputting the line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient into a preset probability prediction model to obtain the probability of the fault cause occurring at the current fault location includes: The line tower type coefficient, environmental adaptability coefficient, and equipment aging coefficient are input into a preset probability prediction model to obtain the initial probability R; the expression of the probability prediction model is: R = R0·α / 5·β·γ; In the formula, R0 is the inherent probability of the fault type, α is the line tower type coefficient, with a value range of 1 to 5, β is the environmental adaptability coefficient, with a value range of 0 to 1, and γ is the equipment aging coefficient, with a value range of 1 to 2. Let the initial probability R and the minimum distance d be... min Input a preset probability correction model to obtain the final probability of the fault cause occurring at the current fault location.

7. The distributed fault diagnosis method for output lines based on knowledge images as described in claim 6, characterized in that, The expression for the probability correction model is: R'=R(1-d min / s); In the formula, R' is the probability of the final occurrence.

8. The distributed fault diagnosis method for output lines based on knowledge images as described in claim 6, characterized in that, The expression for the equipment aging factor γ is: γ=W1·γ1+W2·γ2+W3·γ3+W4·γ4; In the formula, W1, W2, W3, and W4 are the first weight coefficient, the second weight coefficient, the third weight coefficient, and the fourth weight coefficient, respectively. γ1 is the correction factor for the service life of the line, with a value range of 1 to 2; γ2 is the correction factor for the frequency of equipment maintenance, with a value range of 0.8 to 1.2; γ3 is the correction factor for the degree of environmental corrosion, with a value range of 0.9 to 1.3; γ4 is the correction factor for the line load fluctuation coefficient, with a value range of 0.9 to 1.

1.

9. The distributed fault diagnosis method for output lines based on knowledge images as described in claim 6, characterized in that, The types of poles and towers include suspended vertical line poles, suspended angle poles, tension straight line poles, tension angle poles, terminal poles, and Z-type poles. Among them, the line pole type coefficient α corresponding to suspended vertical line poles is 1, the line pole type coefficient α corresponding to suspended angle poles is 2, the line pole type coefficient α corresponding to tension straight line poles is 3, the line pole type coefficient α corresponding to tension angle poles is 4, and the line pole type coefficient α corresponding to terminal poles and Z-type poles is 5.

10. A distributed fault diagnosis method for output lines based on knowledge images as described in any one of claims 1-9, characterized in that, The method of obtaining location information of fault points within the target monitoring interval based on line topology knowledge images includes: Obtain the line length L between two monitoring terminals corresponding to the target monitoring interval marked in the line topology knowledge image; The times t1 and t2 corresponding to the first arrival of the traveling wave current at the target monitoring interval are obtained respectively; Obtain the distance x between the fault point and one of the monitoring terminals in the target monitoring interval; where x = (Lv·Δt) / 2, v is the propagation speed of the traveling wave in the transmission line, and Δt = |t1-t2|. Based on the distance x and the coordinates of the corresponding monitoring terminal in the target monitoring interval marked in the line topology knowledge image, the coordinate location information of the fault point is obtained.

Images