Fault identification and early warning method and system for power transmission line

Abstract

The invention discloses a fault identification early warning method and system for a power transmission line, and the method comprises the steps: receiving a fault flashover signal, determining a generation region of the fault flashover signal, and obtaining a target image; determining whether the power transmission line in the occurrence area has abnormal discharge or not; if yes, performing image recognition to obtain a to-be-detected fault source; obtaining a tower gradient, an electromagnetic field and a radar change of a tower area in the occurrence area; determining a motion path of a current fault source, determining the type of the fault source through image snapshot, and determining a target fault source; obtaining meteorological data through a micro-meteorological sensor and network meteorological weather, and obtaining future and real-time galloping swing and galloping frequency of the power transmission line; and starting image snapshot to determine a target probability and an image of a secondary fault caused by a target fault source on the power transmission line by combining a galloping amplitude, a galloping frequency, a tree barrier, an ultrahigh engineering vehicle, a floating object, a lightning stroke and a value greater than an early warning value, determining a high-risk interval section, and sending an early warning picture and a video to a background in real time.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of fault warning technology, and in particular to a method and system for fault identification and warning of a power transmission line. Background Art

[0002] Transmission lines usually cross a wide geographical area, including complex terrains such as mountains, forests, and rivers. These environmental conditions pose great challenges to the daily maintenance and inspection of transmission lines. First, the complexity of the line corridor means that inspectors need to face various natural obstacles, such as steep slopes, dense vegetation, and unstable ground conditions. This leads to a long inspection cycle for manual inspections, making it difficult to detect and handle potential faults in a timely manner, resulting in poor results for manual inspections, and in turn poor safety in transmission line maintenance.

[0003] Therefore, in the current maintenance of transmission lines, relying solely on traditional manual inspection methods is difficult to meet the safety requirements of modern power systems for transmission lines. Summary of the invention

[0004] The embodiments of the present application provide a method and system for fault identification and early warning of a power transmission line, which are used to effectively improve the safety of power transmission line maintenance.

[0005] To achieve the above objectives, the embodiments of the present application adopt the following technical solutions:

[0006] In a first aspect, a method for fault identification and early warning of a power transmission line is provided, the method comprising:

[0007] In response to receiving a fault flashover signal, determining an occurrence area of ​​the fault flashover signal, and acquiring a target image of the occurrence area;

[0008] Determining whether there is abnormal discharge in the power transmission line in the occurrence area;

[0009] In the case where abnormal discharge exists in the power transmission line in the occurrence area, image recognition is performed on the target image to obtain a first image recognition result, wherein the first image recognition result includes at least one fault source to be detected;

[0010] In response to receiving the target parameters sent by the laser radar, obtaining the inclination of the tower of the power transmission line in the occurrence area and the electromagnetic field change of the power transmission line at the current moment;

[0011] Determine a movement path of a current fault source according to the target parameter, and determine a fault source category of the current fault source according to the movement path, the target parameter, the inclination of the tower, and the electromagnetic field change;

[0012] According to the fault source category, determining a target fault source among the fault sources to be tested, wherein the target fault source is used to characterize the current fault source;

[0013] Acquire meteorological data for a future time period, and predict the dancing amplitude and dancing frequency of the power transmission line for the future time period based on the meteorological data for the future time period;

[0014] Determine the target probability of the target fault source causing a secondary fault on the transmission line in combination with the dancing swing amplitude, the dancing frequency and the motion path, and determine the high-risk section of the transmission line;

[0015] The target probability and the high-risk section of the transmission line are output.

[0016] In a possible implementation manner of the first aspect, determining whether there is abnormal discharge in the power transmission line in the occurrence area includes:

[0017] In response to receiving an electromagnetic field signal sent by an electromagnetic field sensor, performing a fast Fourier transform on the electromagnetic field signal to obtain a spectrum diagram;

[0018] Extracting spectrum features of the spectrum graph, wherein the spectrum features include amplitude distribution, spectrum graph symmetry, phase distribution, spectrum energy distribution, bandwidth, fundamental frequency harmonic components, and spectrum tilt;

[0019] In the discharge type feature library, a target spectrum feature matching the amplitude distribution, the spectrum graph symmetry and the fundamental frequency harmonic component is determined according to a preset similarity algorithm, and a target discharge type corresponding to the target spectrum feature is determined;

[0020] Acquiring a reference feature of the target discharge type, and determining whether the target discharge type is an abnormal discharge according to the reference feature;

[0021] Among them, the benchmark features include benchmark phase distribution, benchmark spectral energy distribution, benchmark frequency bandwidth and benchmark spectral inclination. When there are non-fundamental harmonic components and at least two of the benchmark features are inconsistent with the corresponding spectral features, the target discharge type is determined to be abnormal discharge.

[0022] In another possible implementation manner of the first aspect, determining a movement path of a current fault source according to the target parameter, and determining a fault source category of the current fault source according to the movement path, the target parameter, the inclination of the tower, and the electromagnetic field change, includes:

[0023] Obtain the line height and line location of the transmission line in the occurrence area;

[0024] Determine a movement path of the current fault source according to the target parameter, and determine an influence factor and influence range of the current fault source on the transmission line according to the movement path, the target parameter, the inclination of the tower, the line height and the line position, wherein the target parameter includes height data, length data and speed data of the current fault source;

[0025] Determine at least one category of a fault source to be tested by combining the influencing factor and the influencing range;

[0026] According to the change in the electromagnetic field, a fault source category of a current fault source is determined in at least one of the fault source categories to be detected.

[0027] In another possible implementation manner of the first aspect, determining a movement path of a current fault source according to the target parameter, and determining an influence factor and an influence range of the current fault source on a transmission line according to the movement path, the target parameter, the inclination of the tower, the line height, and the line position, includes:

[0028] Constructing a motion model of the current fault source according to the height data, length data and speed data of the current fault source;

[0029] Using the motion model, simulating the position change of the current fault source within a preset time period;

[0030] Determine the spatial occupied area at each time point by combining the height data and the length data;

[0031] Determine the movement path of the current fault source according to the position change of the current fault source within a preset time period and the spatial occupied area at each time point;

[0032] Simulating the movement path of the current fault source, the line height and the line position to obtain the intersection and overlapping area between the current fault source and the transmission line, the length of the overlapping area being the contact length, and the area of ​​the overlapping area being used to characterize the impact range;

[0033] Using a contact time calculation formula, according to the contact length and the speed data, the contact time between the current fault source and the transmission line is calculated;

[0034] The direction of the force applied by the current fault source to the transmission line is determined according to the inclination of the tower and the speed data, wherein the contact time and the direction of the force are used to characterize the influencing factor.

[0035] In another possible implementation manner of the first aspect, the determining at least one category of a to-be-detected fault source in combination with the impact factor and the impact range includes:

[0036] Determine the nature of the current fault source according to the contact time and the direction of the force, wherein the nature of the current fault source includes instantaneous impact and continuous interference;

[0037] In the case where the nature of the current fault source is an instantaneous impact, determining at least one type of fault source to be tested in combination with the impact range; or

[0038] In the case where the nature of the current fault source is continuous interference, determining at least one category of the fault source to be tested in combination with the impact range;

[0039] The categories of fault sources to be tested include over-height vehicles, cranes, tree obstacles and foreign objects.

[0040] In another possible implementation manner of the first aspect, determining a fault source category in at least one of the fault source categories to be detected according to the change in the electromagnetic field includes:

[0041] Determining the properties of the electromagnetic field change, the properties of the electromagnetic field change including instantaneous properties, slow properties and continuous properties;

[0042] In the case where the property of the electromagnetic field change is instantaneous, determining that the current fault source is related to metal, and determining a fault source category related to metal in at least one of the fault source categories to be tested;

[0043] In the case that the property of the electromagnetic field change is slow and / or continuous, it is determined that the current fault source is related to non-metal, and a fault source category related to non-metal is determined in at least one of the fault source categories to be tested.

[0044] In another possible implementation manner of the first aspect, the meteorological data includes wind direction data and wind speed data, and predicting the dancing amplitude and dancing frequency of the power transmission line in the future time period according to the meteorological data in the future time period includes:

[0045] The wind direction data and the wind speed data of the future time period are input into a preset dynamic equation, and the dynamic equation is solved to obtain the dancing amplitude and dancing frequency of the conductor of the transmission line at each time point in the future time period.

[0046] In another possible implementation manner of the first aspect, combining the dancing swing amplitude, the dancing frequency, and the motion path to determine a target probability that the target fault source causes a secondary fault on the transmission line, and determining a high-risk section of the transmission line, includes:

[0047] Determine a stress concentration area on the transmission line within a target area according to the dancing swing amplitude and the dancing frequency, wherein the target area includes the occurrence area, and the stress concentration area is used to characterize a high-risk section of the transmission line;

[0048] Determining the amount of overlap between the stress concentration area on the transmission line and the movement path of the target fault source;

[0049] A value obtained by dividing the number of overlaps by the length of the power transmission line in the target area is used as the target probability of the target fault source.

[0050] In another possible implementation manner of the first aspect, determining the stress concentration area on the transmission line in the target area according to the dancing swing amplitude and the dancing frequency includes:

[0051] Using a preset stress calculation formula to calculate the stress value of the conductor of the power transmission line at different dancing swing amplitudes and dancing frequencies;

[0052] The position where the stress value is the largest is regarded as the stress concentration area.

[0053] In a second aspect, the present application provides a transmission line fault identification and early warning system, comprising:

[0054] a memory configured to store instructions; and

[0055] The processor is configured to call the instructions from the memory and implement the above-mentioned transmission line fault identification and early warning method when executing the instructions.

[0056] Through the above technical scheme, the fault flashover signal on the transmission line is monitored in real time to locate the fault area, so as to quickly lock the fault area and greatly shorten the fault response time; by detecting abnormal discharge and performing image recognition on the target image, automatic inspection in all-weather and complex terrain can be realized, which effectively improves the accuracy and efficiency of fault identification; through the target parameters, tower inclination and electromagnetic field changes detected by the lidar, combined with the movement path of the current fault source, the fault source category is determined, the type of fault source can be located more accurately and false alarms can be reduced; the dancing frequency and dancing swing amplitude are predicted through meteorological data, and the target probability of the target fault source causing a secondary fault on the transmission line is determined based on the dancing swing amplitude, dancing frequency and movement path, and the high-risk section of the transmission line is determined, so that fault early warning can be realized, overcoming the problems of long manual inspection cycle and poor effect, and effectively improving the safety and maintenance efficiency of the transmission line.

[0057] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 A schematic diagram of a flow chart of a method for identifying and warning a fault of a power transmission line provided in an embodiment of the present application;

[0059] Figure 2 A flowchart of the steps of determining the impact factor and impact range of the current fault source on the transmission line provided in an embodiment of the present application. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the specific implementation methods described herein are only used to illustrate and explain the embodiments of the present application, and are not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0061] It should be noted that if the embodiments of the present application involve directional indications (such as up, down, left, right, front, back...), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0062] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present application, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0063] Figure 1 The following is a schematic diagram showing a flow chart of a method for identifying and warning a fault of a power transmission line according to an embodiment of the present application. Figure 1 As shown, an embodiment of the present application provides a method for fault identification and early warning of a power transmission line, which may include the following steps.

[0064] S110, in response to receiving a fault flashover signal, determining an occurrence area of ​​the fault flashover signal, and acquiring a target image of the occurrence area;

[0065] S120, determining whether there is abnormal discharge in the power transmission line in the occurrence area;

[0066] S130, when abnormal discharge exists in the power transmission lines in the occurrence area, performing image recognition on the target image to obtain a first image recognition result, wherein the first image recognition result includes at least one fault source to be detected;

[0067] S140, in response to receiving the target parameters sent by the laser radar, obtaining the inclination of the tower of the transmission line in the occurrence area at the current moment and the change of the electromagnetic field of the transmission line;

[0068] S150, determining a movement path of the current fault source according to the target parameters, and determining a fault source category of the current fault source according to the movement path, the target parameters, the inclination of the tower, and the change of the electromagnetic field;

[0069] S160, determining a target fault source from the fault sources to be tested according to the fault source category, where the target fault source is used to characterize the current fault source;

[0070] S170, obtaining meteorological data for a future time period, and predicting the dancing amplitude and dancing frequency of the power transmission line for the future time period based on the meteorological data for the future time period;

[0071] S180, determining the target probability of the target fault source causing a secondary fault on the transmission line based on the dancing swing amplitude, dancing frequency and movement path, and determining the high-risk section of the transmission line;

[0072] S190. Output target probability and high-risk sections of transmission lines.

[0073] In the first embodiment, first respond to the received fault flashover signal. The fault flashover signal can be captured by a high-speed camera, that is, the visual image can be quickly captured by the high-speed camera when a fault flashover occurs in the transmission line. The fault flashover signal indicates that an electrical fault may have occurred at a certain position of the transmission line. After receiving the fault flashover signal, the occurrence area of ​​the fault flashover signal can be determined by GPS positioning, and the target image of the occurrence area can be captured by a high-speed camera. The high-altitude camera is fixedly installed on the tower of the transmission line. After determining the occurrence area of ​​the fault flashover signal, the high-speed camera in the occurrence area is quickly called to shoot and obtain the target image. The occurrence area refers to the transmission line area where the fault may occur as indicated by the fault flashover signal, including the transmission line, tower and other facilities within the monitoring range of the high-speed camera.

[0074] After determining the occurrence area, first determine whether there is abnormal discharge in the occurrence area. Under normal circumstances, there is weak corona discharge on the transmission line. Abnormal discharge refers to discharge phenomena beyond the normal range, which may be caused by insulator damage, line slack or approaching external objects. Whether there is abnormal discharge can be determined by electromagnetic signals, sound waves or electrical parameters. Taking electromagnetic field detection as an example, the steps to determine whether there is abnormal discharge can be: 1. Collect electromagnetic signals through broadband antennas and high-speed sampling equipment. The sampling rate usually needs to reach hundreds of MHz or even GHz levels to capture fast discharge pulses. 2. Perform pre-processing such as filtering and denoising on the collected original electromagnetic signals. 3. Extract the spectral characteristics of the electromagnetic signal, including amplitude, rise time, duration, etc. 4. Compare the spectral characteristics with the corresponding normal baseline characteristics to determine whether there is abnormal discharge.

[0075] When it is confirmed that there is abnormal discharge in the occurrence area, the target image is subjected to image recognition to obtain a first image recognition result including at least one fault source to be detected. The fault source to be detected may be an object, equipment failure, or environmental factor. The object may be an over-height vehicle, a tree barrier, a kite, a bird, etc. The equipment failure may be a broken insulator, a loose line, corrosion of metal parts, etc. The environmental factor may be snow accumulation, ice, and pollutant deposition, etc.

[0076] The specific implementation steps of image recognition can be: first pre-process the target image, and use a pre-trained deep convolutional neural network to extract the image features of the target image, then use a preset target detection algorithm to locate and classify possible fault sources to be tested, and finally output at least one fault source to be tested and the category of each fault source to be tested. Specifically, image pre-processing includes denoising, contrast enhancement and image segmentation. Denoising is used to reduce the noise of the target image; contrast enhancement is used to improve the clarity of the target image; image segmentation is used to segment the target image into regions of interest (ROI) to focus on monitoring the transmission lines in the area where the fault occurs.

[0077] In this embodiment, when a fault flashover signal is received, the laser radar in the occurrence area simultaneously sends the target parameters of the current fault source and is received by the current execution subject. After receiving the target parameters, the inclination of the transmission line tower in the occurrence area at the current moment and the electromagnetic field change of the transmission line are obtained. The target parameters include the height data, length data and speed data of the current fault source. It should be noted that the current fault source mentioned in this embodiment is an unknown fault source, that is, the category of the current fault source is unknown, and only the relevant target parameters of the current fault source can be obtained through the laser radar.

[0078] Specifically, the laser radar can accurately measure the height, length and speed of the current fault source by emitting laser pulses and measuring the return time of the reflected light. The laser radar can also obtain the point cloud data of the transmission line tower, which can be used to build a three-dimensional model of the tower.

[0079] The inclination of the transmission line tower can be calculated by the preset RANSAC algorithm. Specifically, first, the RANSAC algorithm is used to fit the central axis in the tower point cloud data. Then, the central axis is compared with the vertical line to determine the angle θ. Finally, the angle θ is substituted into the following inclination calculation formula to calculate the inclination. The inclination calculation formula is:

[0080] Inclination = tan(θ)*100%.

[0081] For example, if θ=2° is calculated, the inclination is ≈3.49%.

[0082] The change in the electromagnetic field can be determined by an electromagnetic sensor, that is, the change in the electromagnetic field can be determined by the electric field strength E and the magnetic field strength H detected by the electromagnetic sensor.

[0083] This embodiment uses laser radar technology to detect physical deformations such as tower tilt and conductor sag with millimeter-level accuracy. At the same time, electromagnetic field monitoring can capture electrical anomalies of the line in real time.

[0084] Since the target parameters include the height data, speed data and length data of the current fault source, the movement path of the current fault source can be determined according to the target parameters. After determining the movement path, the category of the current fault source can be determined based on the movement path, target parameters, the inclination of the tower and the electromagnetic field changes.

[0085] Specifically, the motion path of the fault source can be predicted by the real-time coordinates and speed data of the current fault source. For example, when the three-dimensional coordinates of the current fault source are (x, y, z), the speed vector is (vx, vy, vz), and the timestamp is t, the preset target tracking algorithm can be used to track the target, predict its future position, and obtain a smooth motion trajectory. The preset target tracking algorithm can be a Kalman filter algorithm.

[0086] Next, the fault source is classified based on the motion path, target parameters, the inclination of the tower, and the change of the electromagnetic field. In one implementation, the motion characteristics of the current fault source can be determined by analyzing the speed and direction of the motion path, such as fast and irregular or slow and regular. After that, the size of the current fault source can be determined by the target parameters; the inclination of the tower can be used to determine whether the current fault source has physically collided with the wire or tower of the transmission line; the change of the electromagnetic field can be used to determine whether the current fault source is metal or non-metal. In summary, the fault source category of the current fault source can be determined.

[0087] In another embodiment, the motion path, target parameters, inclination of the tower, and electromagnetic field changes may be input into a pre-trained classifier to obtain a probability distribution of the fault source category.

[0088] After determining the fault source category, the best matching target fault source can be determined among the fault sources to be tested. The best matching target fault source can be determined by comparing motion features, size, collision, and metal. The similarity of category vectors can also be calculated using cosine similarity to determine the best matching target fault source among the fault sources to be tested. It should be noted that since the category of the target fault source is known, the target fault source is a known fault source.

[0089] In order to accurately predict the dynamic behavior of the transmission line under future meteorological conditions, first, obtain the meteorological data for the future time period, including wind speed and direction (usually expressed as a 10-minute average), etc. The meteorological data includes forecast data for the next 24-72 hours, with a time resolution of every hour or every 3 hours.

[0090] Next, the dynamic behavior (dancing amplitude and dancing frequency) of the transmission line is predicted by meteorological data. The steps can be: 1. Use the finite element method to build a dynamic model of the conductor. 2. Calculate the wind force based on the meteorological data. 3. Use the numerical integration method to solve the dynamic equation and calculate the dancing amplitude and frequency.

[0091] Since the current fault source may not be instantaneous, it is necessary to combine the dancing swing amplitude, dancing frequency and fault source movement path to determine the target probability of the target fault source causing secondary faults on the transmission line and determine the high-risk interval sections to quantify potential risks and identify high-risk line sections that need to be focused on. Secondary faults may include direct contact between the fault source and the transmission line, electrical discharge caused by the fault source, mechanical stress caused by the fault source exceeding the line's tolerance, etc.

[0092] In one embodiment, the steps for determining the target probability may be: 1. Superimposing the motion path of the target fault source with the predicted dancing area of ​​the transmission line. 2. For each time step, calculate the minimum distance between the fault source and the transmission line. 3. Calculate the probability of electrical discharge based on the electric field strength model and the material properties of the fault source. 4. Combine the dancing swing amplitude and frequency to calculate the dynamic stress of the line, and use the fatigue accumulation model to evaluate the probability of failure. 5. Integrate the risk over the entire prediction time period to obtain the target probability. 7. Use a density clustering algorithm (such as DBSCAN) to identify high-risk areas. 8. Determine the high-risk interval segment based on the clustering results and the preset risk threshold. Finally, use a heat map to visualize the risk distribution and intuitively display the high-risk interval segment.

[0093] In the second embodiment, a fault identification and early warning method for a power transmission line may further include: receiving a fault flashover signal, determining the occurrence area of ​​the fault flashover signal, and obtaining a target image of the occurrence area; determining whether there is abnormal discharge in the transmission line in the occurrence area; if there is abnormal discharge, performing image recognition on the target image to obtain a first image recognition result; receiving target parameters, obtaining the inclination of the tower area of ​​the transmission line in the occurrence area at the current moment, and the changes in electromagnetic field and radar monitoring; determining the movement path of the current fault source according to the target parameters, and determining the fault source category through image capture; determining the target fault source among the fault sources to be tested according to the fault source category; obtaining future meteorological data monitored and compared with the sensor module through micro-meteorological sensors and network meteorological weather, and predicting and obtaining the dancing swing amplitude and dancing frequency of the power transmission line in the future and real-time time periods in real time; combining the dancing swing amplitude, dancing frequency and movement path, tree obstacles, super-high engineering vehicles, floating objects, lightning strikes, which are greater than the set warning value, automatically starting image capture to determine the target probability and image data of the target fault source causing secondary faults to the transmission line, determining the high-risk section of the transmission line, and sending the warning pictures and videos to the background in real time.

[0094] Specifically, compared with the first embodiment, micro-meteorological sensors and network meteorological weather can be used to obtain future meteorological data for comparison with the sensor module monitoring, and the inclination of the tower and the electromagnetic field changes can be obtained through radar monitoring changes, so that the state of the transmission line can be monitored in real time. In addition, the fault source category can be determined through image capture, and tree barriers, super-high engineering vehicles, floating objects, lightning strikes and other fault source categories can be considered. Finally, image capture can be automatically started, and after determining the target probability and image data of the target fault source causing secondary faults to the transmission line, the warning pictures and videos can be sent to the background in real time.

[0095] In specific implementation, tree obstacles such as tree growth and tree falling caused by strong winds may come into contact with transmission lines, causing short circuits or grounding faults; if the height of an engineering vehicle working near a transmission line exceeds the safe distance, it may collide with the line, causing line damage; strong winds may blow lightweight floating objects such as plastic bags and kites onto the transmission line, causing short circuits or grounding faults; direct lightning strikes on the transmission line may cause line damage or tripping, etc. Therefore, factors such as tree obstacles, super-high engineering vehicles, floating objects, and lightning strikes need to be considered in the fault identification and early warning method of the transmission line, so as to more comprehensively evaluate the potential risks of the transmission line, issue early warnings and take corresponding protective measures, improve the safety and reliability of the transmission line, and ensure the stable operation of the transmission line in various complex environments.

[0096] This embodiment monitors the fault flashover signal on the transmission line in real time to locate the fault area, so as to quickly lock the fault area and greatly shorten the fault response time; by detecting abnormal discharge and performing image recognition on the target image, automatic inspection in all-weather and complex terrain can be realized, effectively improving the accuracy and efficiency of fault identification; by detecting the target parameters, tower inclination and electromagnetic field changes detected by the lidar, combined with the movement path of the current fault source, the fault source category is determined, the type of fault source can be located more accurately, and false alarms can be reduced; the dancing frequency and dancing swing amplitude are predicted through meteorological data, and the target probability of the target fault source causing a secondary fault on the transmission line is determined based on the dancing swing amplitude, dancing frequency and movement path, and the high-risk section of the transmission line is determined, so that fault early warning can be realized, overcoming the problems of long manual inspection cycle and poor effect, and effectively improving the safety and maintenance efficiency of the transmission line.

[0097] In one implementation of this embodiment, determining whether there is abnormal discharge in the power transmission line in the occurrence area includes the following steps:

[0098] S210, in response to receiving the electromagnetic field signal sent by the electromagnetic field sensor, performing a fast Fourier transform on the electromagnetic field signal to obtain a spectrum diagram;

[0099] S220, extracting spectrum features of the spectrum graph, where the spectrum features include amplitude distribution, spectrum graph symmetry, phase distribution, spectrum energy distribution, bandwidth, fundamental frequency harmonic components, and spectrum tilt;

[0100] S230, in the discharge type feature library, determining a target spectrum feature that matches the amplitude distribution, the symmetry of the spectrum graph, and the fundamental frequency harmonic components according to a preset similarity algorithm, and determining a target discharge type corresponding to the target spectrum feature;

[0101] S240, obtaining a reference feature of a target discharge type, and determining whether the target discharge type is an abnormal discharge according to the reference feature;

[0102] Among them, the benchmark features include benchmark phase distribution, benchmark spectral energy distribution, benchmark frequency bandwidth and benchmark spectral tilt. When there are non-fundamental harmonic components and at least two benchmark features are inconsistent with the corresponding spectral features, the target discharge type is determined to be abnormal discharge.

[0103] In this embodiment, the electromagnetic field signal is first subjected to a fast Fourier transform (FFT) to obtain a spectrum diagram. Specifically, the electromagnetic field sensor receives an electromagnetic field signal in the time domain, which is a continuous analog signal and needs to be converted into a discrete digital signal through sampling and quantization. The sampling frequency should not be less than twice the highest frequency of the signal to avoid spectrum aliasing. For the electromagnetic field signal of the transmission line, the sampling frequency can be in the range of several kHz to several MHz. Next, the FFT algorithm (fast Fourier transform algorithm) is applied to these discrete time series data. FFT is a method for calculating discrete Fourier transform (DFT). The core idea is to decompose a DFT of length N into several shorter DFTs, and then recursively calculate the short DFT. In specific implementation, the base-2FFT algorithm can be used, requiring the input sequence length to be a power of 2. If the length of the original data does not meet the requirements, it can be adjusted by zero padding or truncation. The output of FFT is a complex array, which contains the amplitude and phase information of the signal at different frequencies. By calculating the modulus of the complex number, the amplitude information in the spectrum diagram can be obtained; by calculating the radix of the complex number, the phase information can be obtained. The resulting spectrogram shows frequency on the horizontal axis and amplitude or power spectral density on the vertical axis.

[0104] Next, key spectrum features are extracted from the spectrum graph, including amplitude distribution, spectrum graph symmetry, phase distribution, spectrum energy distribution, bandwidth, fundamental harmonic components, and spectrum tilt. Amplitude distribution reflects the distribution of signal energy on each frequency, which can be quantified by calculating the sum or average of amplitudes in different frequency intervals. The symmetry of the spectrum graph can be evaluated by comparing the similarity of the positive and negative frequency parts. A completely symmetrical spectrum graph indicates that the original signal may be a real-valued signal. Phase distribution is used to provide relative time delay information of each frequency component of the signal, which can be obtained by calculating the phase angle of each frequency point. Spectral energy distribution can be obtained by calculating the sum of squares of the amplitudes of each frequency point in the spectrum graph, which is used to reflect the distribution of total signal energy in the frequency domain. Bandwidth is the frequency range where signal energy is concentrated, which can be measured using indicators such as 3dB bandwidth or equivalent noise bandwidth. Fundamental harmonic components refer to the amplitudes of the fundamental frequency and its integer multiple frequencies that appear in the spectrum, which can be extracted by identifying the amplitudes of specific frequency points. Spectral tilt reflects the energy distribution of high-frequency components relative to low-frequency components, which can be quantified by calculating the ratio of energy in the high-frequency area to that in the low-frequency area. Feature extraction can use digital signal processing technology, such as filtering, interpolation, peak detection, etc. For example, for the extraction of fundamental frequency harmonic components, the peak detection algorithm can be used to find the significant peaks in the spectrum, and then verify whether these peaks form a harmonic relationship. The bandwidth can be calculated using the cumulative energy method, that is, the energy is accumulated from the low frequency until the frequency range corresponding to a certain percentage (such as 95%) of the total energy is reached.

[0105] The discharge type feature library includes various discharge type spectrum features, and each discharge type has its own feature vector, including amplitude distribution, spectrum symmetry, and fundamental frequency harmonic components. The matching process uses a preset similarity algorithm, such as the Euclidean distance algorithm. Specifically, when using the Euclidean distance, the extracted spectrum features and the features in the database can be used as points in a multidimensional space, and the distance between them can be calculated. The database feature with the smallest distance is the matching result.

[0106] It should be noted that the matched target spectrum features may correspond to normal discharge or abnormal discharge. Since the features of different discharge types may overlap, this implementation alone may not be able to classify them completely accurately. Therefore, other information needs to be combined to further determine whether the discharge is abnormal.

[0107] After determining the target discharge type, it is further determined whether the discharge is an abnormal discharge. First, the benchmark features of the target discharge type are obtained, including the benchmark phase distribution, the benchmark spectrum energy distribution, the benchmark bandwidth and the benchmark spectrum inclination. The benchmark features are used to represent the typical features of this type of discharge under normal circumstances. The judgment process mainly includes two aspects: one is to check whether there are non-fundamental harmonic components, and the other is to compare the consistency of the actual features with the benchmark features. The presence of non-fundamental harmonic components may mean that an abnormality has occurred during the discharge process, which can be judged by analyzing whether there are significant peaks in the spectrum that do not belong to integer multiples of the fundamental frequency. The comparison of feature consistency can be done by statistical methods, such as calculating the degree of difference between the actual feature and the benchmark feature. For example, the mean square error or standard deviation can be used to quantify the difference, and if the mean square error exceeds a preset threshold, the feature is considered inconsistent. When there are non-fundamental harmonic components and at least two features are inconsistent with the benchmark, it is determined to be an abnormal discharge. Please refer to the prior art for the calculation of the mean square error or standard deviation, and this application will not go into details.

[0108] This implementation method determines whether there is abnormal discharge through the spectrum characteristics of the spectrum diagram, thereby achieving accurate identification of abnormal discharge, and adopts a multi-feature comprehensive judgment method to improve the accuracy of identification, reduce the possibility of false alarms and missed alarms, and help improve the safety and reliability of transmission lines and reduce the risk of faults.

[0109] In one implementation of this embodiment, determining the movement path of the current fault source according to the target parameter, and determining the fault source category of the current fault source according to the movement path, the target parameter, the inclination of the tower, and the electromagnetic field change, includes the following steps:

[0110] S310, obtaining the line height and line position of the power transmission line in the occurrence area;

[0111] S320, determining the movement path of the current fault source according to the target parameters, and determining the influence factor and influence range of the current fault source on the transmission line according to the movement path, the target parameters, the inclination of the tower, the line height and the line position, wherein the target parameters include the height data, the length data and the speed data of the current fault source;

[0112] S330, determining at least one type of fault source to be tested based on the influencing factor and the influencing range;

[0113] S340: Determine a fault source category of a current fault source in at least one fault source category to be detected according to the change in the electromagnetic field.

[0114] In this embodiment, the line height refers to the vertical distance between the line and the ground, and the line position includes the spatial distribution of the transmission line in the geographic coordinate system. Specifically, after the occurrence area is determined, the line height and line position of the transmission line can be called in time through the database, and the line height and line position of the transmission line with different positioning are stored in the database.

[0115] Determine the movement path of the current fault source according to the target parameters. Specifically, first, construct a movement model of the current fault source according to the target parameters (including the height data, length data and speed data of the fault source). Next, compare the movement path of the current fault source with the spatial position of the transmission line, and calculate the change of the shortest distance between the two over time, which can be achieved by calculating the shortest distance from a point to a line segment. At the same time, consider the inclination of the tower, because the inclination of the tower will change the actual position and tension distribution of the line. The influencing factors include contact time and force direction, where the force direction is determined by the current fault source movement direction and the inclination of the tower; the contact time refers to the time when the current fault source contacts the transmission line in the occurrence area. The determination of the impact range can be achieved by setting a critical distance. If the distance between the current fault source and the line is less than the critical value, it is determined that the current fault source enters the impact range.

[0116] After determining the influencing factors and the scope of influence, the category of the fault source to be tested can be determined based on the influencing factors and the scope of influence. In one embodiment, the category of the fault source to be tested can be obtained by inputting the relevant data of the influencing factors and the scope of influence into a pre-trained machine learning model. The machine learning model can be a random forest model. The specific construction steps of the random forest model are as follows: 1. Combine the influencing factors (such as the direction of force and contact time) and the scope of influence (the area of ​​the affected area) into a feature vector. 2. Use historical data to train the random forest model. Each decision tree randomly selects a feature subset when the node splits to improve the generalization ability of the model. 3. Input the feature vector of the current fault source into the trained model to obtain the probability of each category. 4. The category above the preset threshold is used as the category of the fault source to be tested.

[0117] Based on the category of the fault source to be tested determined in S330, combined with the electromagnetic field change information, the specific category of the current fault source can be determined. First, the electromagnetic field change data is obtained by installing an electromagnetic field sensor array around the transmission line, and the characteristics of the electromagnetic field change are extracted. Next, these electromagnetic field characteristics are compared with the category of the fault source to be tested determined in step S330 to determine the fault source category of the current fault source. Among them, the characteristics of the electromagnetic field change include frequency domain characteristics, time domain characteristics, and time-frequency characteristics. Time domain characteristics include characteristics such as spectrum peaks and harmonic ratios obtained by fast Fourier transform (FFT). Time domain characteristics include statistical quantities such as the mean, variance, peak factor, skewness, and kurtosis of the signal. Time-frequency characteristics include time-frequency distribution characteristics obtained by wavelet transform or Hilbert-Huang transform.

[0118] This implementation method greatly improves the accuracy of identification by gradually narrowing the scope of possible fault sources and comprehensively considering physical characteristics and electromagnetic characteristics. It can not only determine the category of the fault source, but also evaluate its potential impact on the transmission line, providing strong support for preventive maintenance and emergency response, and helping to improve the safety and reliability of transmission lines.

[0119] Figure 2 The flowchart of the steps of determining the impact factor and impact range of the current fault source on the transmission line provided by the embodiment of the present application is shown as follows: Figure 2 As shown, in one implementation of this embodiment, the movement path of the current fault source is determined according to the target parameter, and the influence factor and influence range of the current fault source on the transmission line are determined according to the movement path, the target parameter, the inclination of the tower, the line height and the line position, including the following steps:

[0120] S410: construct a motion model of the current fault source according to the height data, length data and speed data of the current fault source:

[0121] S420, using a motion model to simulate the position change of the current fault source within a preset time period;

[0122] S430, combining the height data and the length data, determining the space occupied area at each time point;

[0123] S440, determining a movement path of the current fault source according to a position change of the current fault source within a preset time period and a spatial occupied area at each time point;

[0124] S450, simulating the movement path, line height and line position of the current fault source to obtain the intersection and overlapping area between the current fault source and the transmission line, the length of the overlapping area is the contact length, and the area of ​​the overlapping area is used to characterize the impact range;

[0125] S460, using a contact time calculation formula, according to the contact length and speed data, calculate the contact time between the current fault source and the transmission line;

[0126] S470. Determine the direction of the force of the current fault source on the transmission line based on the inclination and speed data of the tower, wherein the contact time and the direction of the force are used to characterize the influencing factors.

[0127] First, the physical characteristics of the fault source need to be determined, including its shape and aerodynamic characteristics. The height data provides the initial position information of the fault source, the length data is used to determine the geometric size of the fault source, and the speed data provides the initial motion state.

[0128] When constructing the motion model, this embodiment uses a three-dimensional coordinate system to describe the current position of the fault source. Considering the influence of gravity, a parabola model can be used to describe the motion trajectory of the fault source. In the horizontal direction, assuming that the air resistance can be ignored, the fault source maintains a uniform motion; in the vertical direction, the fault source is subjected to a uniform acceleration motion due to the action of gravity. The specific motion equation can be expressed as:

[0129] x(t)=x 0 +υ x ×t;

[0130] y(t)=y 0 +υ y ×t-0.5×g×t 2 ;

[0131] Among them, (x 0 ,y 0 ) is the initial position, v x and v y are the initial velocity components in the horizontal and vertical directions, respectively, and g is the acceleration due to gravity (about 9.8 m / s 2 ).

[0132] Next, the motion model constructed above is time-deduced to simulate the position change of the current fault source within a preset time period. Based on the numerical integration method, the position coordinates of the current fault source at different time points can be obtained by discretizing and solving the motion equation.

[0133] When implementing it, you first need to determine the time range and time step of the simulation. The time range depends on the expected impact time of the fault source and can be set to a few seconds to a few minutes. The time step can be selected in the order of 0.01 seconds to 0.1 seconds.

[0134] Specifically, for each time step, the corresponding x and y coordinates are calculated to obtain the position of the current fault source at each time point.

[0135] Combined with the height data and length data of the current fault source, the spatial occupied area of ​​the current fault source at each time point is determined, which is mainly used to expand the point-like location information into a spatial area with actual size.

[0136] In the specific implementation, firstly, it is necessary to select a suitable geometric model according to the shape characteristics of the current fault source. For slender current fault sources (such as tree branches), a line segment model can be used; for larger current fault sources (such as overheight vehicles), a rectangular or elliptical model can be used. Taking the line segment model as an example, the implementation steps are as follows:

[0137] 1. For each time point t, obtain the position coordinates (x(t), y(t)) calculated in S420.

[0138] 2. According to the length L of the current fault source and the current movement direction θ(t), calculate the coordinates of the two endpoints of the line segment:

[0139] Endpoint 1: (x 1 (t), y 1 (t))=(x(t)-0.5L*cos(θ(t)), y(t)-0.5L*sin(θ(t)));

[0140] Endpoint 2: (x2(t), y2(t))=(x(t)+0.5L*cos(θ(t)), y(t)+0.5L*sin(θ(t)));

[0141] 3. Connect the two endpoints to form a line segment representing the current fault source.

[0142] Similarly, for the rectangular model, the implementation steps are as follows:

[0143] 1. Get the position coordinates (x(t), y(t)) as the center of the rectangle.

[0144] 2. According to the length L and width W of the current fault source, as well as the current direction θ(t), calculate the coordinates of the four vertices of the rectangle:

[0145] Vertex 1: (x(t)-0.5L*cos(θ(t))+0.5W*sin(θ(t)), y(t)-0.5L*sin(θ(t))-0.5W*cos(θ(t)));

[0146] Vertex 2: (x(t)+0.5L*cos(θ(t))+0.5W*sin(θ(t)), y(t)+0.5L*sin(θ(t))-0.5W*cos(θ(t)));

[0147] Vertex 3: (x(t)+0.5L*cos(θ(t))-0.5W*sin(θ(t)), y(t)+0.5L*sin(θ(t))+0.5W*cos(θ(t)));

[0148] Vertex 4: (x(t)-0.5L*cos(θ(t))-0.5W*sin(θ(t)), y(t)-0.5L*sin(θ(t))+0.5W*cos(θ(t))).

[0149] 3. Connect the above four vertices to form a rectangle representing the current fault source.

[0150] Through the above implementation, the specific space occupied by the current fault source at each time point can be obtained.

[0151] After that, the position change of the current fault source within the preset time period and the space occupied area at each time point are comprehensively considered to finally determine the complete motion path of the current fault source. In the specific implementation, first, the position coordinate sequence obtained in S420 and the space occupied area information calculated in S430 are integrated. For each time point t, the center point coordinates (x(t), y(t)) and the geometric description of the space occupied area, such as the coordinates of the line segment endpoints or the rectangle vertex coordinates, are known.

[0152] For the center point trajectory, a spline interpolation method (such as cubic spline interpolation) can be used to connect the center coordinates of all time points to generate a smooth curve. The interpolation function can be expressed as:

[0153] x(t)=a i +b i (tt i )+c i (tt i ) 2 +d i (tt i ) 3 ;

[0154] y(t)=e i +f i (tt i )+g i (tt i ) 2 +h i (tt i ) 3 ;

[0155] Among them, i represents the interval between adjacent time points, a i to h i is the interpolation coefficient.

[0156] For the space occupied area, the linear interpolation method is used to connect the boundary points of the space occupied area at adjacent time points to form a continuous envelope surface, and the preset smoothing algorithm is applied to smooth the generated trajectory to eliminate possible noise and discontinuities. The envelope surface represents all the spaces that the current fault source may occupy during the movement. Finally, the parameterized equation is used to represent the final motion path:

[0157] R(t)=(x(t), y(t), S(t));

[0158] Among them, (x(t), y(t)) are the coordinates of the center point, and S(t) is the spatial occupied area at time t.

[0159] Through the above implementation, a continuous, smooth motion path that takes actual size into consideration can be generated, which can better reflect the actual motion situation of the current fault source than a point sequence.

[0160] Afterwards, the motion path, line height and line position of the current fault source can be simulated to determine potential intersections and overlapping areas, and accurately quantify the interaction between the current fault source and the transmission line. Specifically, for each time step, check whether the spatial occupied area of ​​the current fault source intersects with the transmission line. Specifically, a preset line segment intersection algorithm (such as a line segment intersection algorithm) can be used to detect the intersection, and for the time period when the intersection is detected, calculate the overlapping part of the spatial occupied area of ​​the current fault source and the transmission line. The polygon clipping algorithm can be used to calculate the area of ​​the overlapping area. The total contact length is obtained by adding up the projection lengths of all overlapping areas on the transmission line. For each overlapping area, the contact length can be obtained by calculating the maximum span of the overlapping polygon in the line direction.

[0161] The contact time is calculated based on the contact length determined in S450 and the speed data of the current fault source, so as to obtain the contact time by dividing the contact length by the relative speed of the current fault source. First, determine the effective speed of the current fault source relative to the transmission line. Assume that the speed vector of the current fault source is v = (v x , v y ), the direction vector of the transmission line is I=(I x , I y ), then the relative velocity can be calculated by projection:

[0162] υ=|υ.l| / |l|=|υ x *l x +υ y *l y | / √(lx 2 +ly 2 );

[0163] Next, use the contact time calculation formula:

[0164] T=L / v ; Where T is the contact time and L is the contact length.

[0165] However, the above basic formula assumes that the speed of the current fault source remains constant throughout the contact process, which may not be true in actual situations. In order to improve the calculation accuracy, a segmented calculation method can be used. The entire contact process is divided into several small segments, and each segment is calculated using the average speed within the time period, and then the total contact time is accumulated. The formula is as follows:

[0166] T t =∑L i / υ i ;

[0167] Among them, L i is the contact length of the ith segment, v i is the average relative velocity of the segment.

[0168] Based on the inclination and speed data of the tower, the direction of the force acting on the transmission line by the current fault source and the influence of the inclination of the tower on the force of the transmission line can be determined. The inclination of the tower will change the tension distribution and geometry of the transmission line, thereby affecting the response of the line to external forces. Assuming that the inclination angle of the tower is θ, the force of the transmission line can be decomposed into two components: horizontal and vertical. Among them, the horizontal component is F h , the vertical component is F v , the formula is as follows:

[0169] F h =F*cos(θ);

[0170] F υ =F*sin(θ);

[0171] Where F is the total force applied by the current fault source.

[0172] Next, consider the velocity data of the current fault source. The velocity not only determines the magnitude of the impact force, but also determines the direction of the force. Assume that the velocity vector of the current fault source is v = (v x , v y ), then its direction of movement can be calculated:

[0173] α=atan2(υ y , x );

[0174] Among them, atan2 is a function used to calculate the azimuth angle α.

[0175] Combining the tower inclination and the current direction of movement of the fault source, the direction of the force in the transmission line coordinate system can be obtained:

[0176] β = α - θ;

[0177] The direction vector of the force can be expressed as: d = (cos(β), sin(β)).

[0178] This implementation method can more accurately assess the potential collision risk between the current fault source and the transmission line by combining the spatial occupancy area information, thereby effectively improving the safety and operation efficiency of the transmission line.

[0179] In one implementation of this embodiment, combining the influencing factor and the influencing range to determine at least one type of fault source to be detected includes the following steps:

[0180] S510, determining the nature of the current fault source according to the contact time and the direction of the force, where the nature of the current fault source includes instantaneous impact and continuous interference;

[0181] S520: When the nature of the current fault source is instantaneous impact, determine at least one type of fault source to be tested in combination with the impact range; or

[0182] S530: When the nature of the current fault source is continuous interference, determine at least one type of fault source to be tested in combination with the impact range;

[0183] Among them, the categories of fault sources to be tested include over-high vehicles, cranes, tree obstacles and foreign objects.

[0184] This embodiment can determine the nature of the current fault source based on the contact time and the direction of the force, and classify it as a transient impact or a continuous interference. Specifically, if the contact time T is less than the critical contact time threshold, it tends to be judged as a transient impact; otherwise, it may be a continuous interference. The critical contact time threshold can be 0.5 seconds. Next, a direction factor D can be defined to represent the cosine value of the angle between the direction of the force and the normal direction of the transmission line. When D is close to 1, it means that the direction of the force is nearly perpendicular to the line, which is more likely to cause a transient impact; when D is close to 0, it means that the direction of the force is nearly parallel to the line, which is more likely to cause a continuous interference. D can be calculated using the following formula:

[0185] D = |F·n| / (|F|*|n|);

[0186] Where F is the force vector and n is the normal vector of the line at the contact point.

[0187] By analyzing the characteristics and impact range of the instantaneous impact, the most likely candidate can be selected from the preset fault source categories (including over-height vehicles, cranes, tree obstacles and foreign objects). For instantaneous impact, the impact range is usually manifested as a relatively concentrated area. This area can be described by the area A of the impact range and the shape factor S. The shape factor can be defined as the ratio of the square of the perimeter of the impact area to the area, that is, S = P 2 / A, where P is the perimeter. Different types of fault sources will have different characteristic values. For example, an over-high vehicle may cause a larger but more regular (smaller S value) impact range, while a foreign object such as a tree branch may cause a smaller but irregular (larger S value) impact range.

[0188] A similarity function can be constructed to compare the observed impact range with the characteristics of various fault sources in the database:

[0189] S(i)=w 1 *exp(-|AA i | / σA)+w 2 *exp(-|SS i | / σS);

[0190] The function exp(x) represents the natural exponential function, that is, the exponential function with the real number e (e≈2.71828) as the base, A and S are the observed values, A i and S i is the characteristic value of the i-th type fault source in the database, σA and σS are standardized parameters, and w 1 and w 2 is the weight coefficient.

[0191] For fault sources that are determined to be of a continuous interference nature, the specific category of the fault source to be tested can be determined in combination with the impact range. Continuous interference usually manifests itself as a long-term effect on the transmission line, and its impact range may change over time, so time series data needs to be considered.

[0192] First, a time series feature database is constructed, which contains typical influence range change patterns and other related parameters of various potential continuous interference fault sources. For continuous interference, the following features can be used to describe the change of influence range: 1. Time function of influence range area A(t); 2. Time function of influence range shape factor S(t); 3. Moving trajectory of influence center position P(t) = (x(t), y(t)); 4. Time function of interference force F(t).

[0193] Time series similarity measures, such as the Dynamic Time Warping algorithm, can be used to compare the observed impact range time series with the characteristic series of various fault sources in the database. The Dynamic Time Warping algorithm can handle time series of different lengths and different rates to calculate the similarity between them.

[0194] The distance of the dynamic time warping algorithm can be expressed as:

[0195] DTW(X, Y) = min(∑d(x i ,y j ));

[0196] Among them, DTW(X, Y) represents the distance of the dynamic time warping algorithm, X and Y are two time series, and d(x i ,y j ) is a single point distance metric.

[0197] In addition to time series similarity, some statistical features can also be extracted, such as: the mean and standard deviation of the impact area; the rate of change of the shape factor; the total distance and average speed of the impact center movement; the maximum, minimum and average values ​​of the interference force.

[0198] Then, machine learning algorithms, such as long short-term memory networks (LSTM) or temporal convolutional networks (TCN), can be used to process time series data and statistical features. Machine learning algorithms can capture long-term dependencies in time series and determine the final list of fault source categories to be tested. The specific implementation steps are as follows: 1. Collect time series data of the impact range. 2. Use the DTW algorithm to calculate the similarity between the observed sequence and the characteristic sequences of various types of fault sources in the database. 3. Extract statistical features and construct feature vector V. 4. Input the time series data and feature vector into the trained LSTM or TCN model. 5. The model outputs the probability distribution of each fault source category. 6. Combine the DTW similarity and the model output probability to determine the final list of fault source categories to be tested.

[0199] This implementation effectively distinguishes between two basic fault types, instantaneous impact and continuous interference, by analyzing the contact time and force direction. For the instantaneous impact type, it combines the impact range characteristics and machine learning technology to accurately identify fast-moving fault sources such as super-high vehicles and cranes. For the continuous interference type, it uses time series analysis and deep learning methods to effectively handle complex situations with long-term effects such as tree barriers. Not only can it adapt to different types of fault sources, but it can also improve recognition accuracy through continuous learning and updating, providing more comprehensive and accurate fault source identification results for transmission line safety management.

[0200] In one implementation manner of this embodiment, determining a fault source category in at least one fault source category to be detected according to the change of the electromagnetic field includes the following steps:

[0201] S610, determining the nature of the electromagnetic field change, where the nature of the electromagnetic field change includes instantaneous nature, slow nature, and continuous nature;

[0202] S620: When the property of the electromagnetic field change is instantaneous, determine that the current fault source is related to metal, and determine a fault source category related to metal in at least one fault source category to be tested;

[0203] S630: When the electromagnetic field changes slowly and / or continuously, determine that the current fault source is related to non-metal, and determine a fault source category related to non-metal in at least one fault source category to be tested.

[0204] This embodiment first uses short-time Fourier transform (STFT) to perform time-frequency analysis on the processed signal. Short-time Fourier transform can provide spectrum information of the signal in different time windows. By analyzing the STFT results, the following features can be obtained: 1. The instantaneous rate of change of the spectrum can be obtained by calculating the difference between the spectra of adjacent time windows. 2. The duration of the main frequency component can be obtained by tracking the existence duration of a specific frequency component. 3. The change pattern of energy distribution can be obtained by the change of energy in different frequency bands over time.

[0205] Based on the above characteristics, the judgment criteria can be defined: Instantaneous property: the instantaneous change rate of the spectrum is high, and the duration of the main frequency component is short (usually less than 1 second). Slow property: the spectrum change rate is medium, and the duration of the main frequency component is between 1 second and several minutes. Sustained property: the spectrum change rate is low, and the duration of the main frequency component is long (greater than several minutes).

[0206] When metal objects interact with electromagnetic fields, eddy currents and electromagnetic induction effects are generated. Specifically, the effects of metal objects on electromagnetic field changes are: metal objects usually cause large instantaneous field strength changes; metal objects often produce significant harmonics in specific frequency bands. Therefore, for electromagnetic field changes that are determined to be instantaneous, the current fault source is determined to be related to metal, and the specific metal-related fault source category is identified in the fault source category to be tested.

[0207] In practical applications, the specific type of metal object also needs to be considered. For example, for an over-height vehicle, it may appear as a larger peak amplitude and faster rise / fall time. For a crane, it may appear as a strong signal with a slightly longer duration and accompanied by periodic changes. For a metal foreign body, it may appear as a smaller peak amplitude but with unique spectral characteristics.

[0208] Since the interaction between non-metallic objects and electromagnetic fields usually manifests as slow or continuous changes, this is because non-metallic objects do not produce strong eddy currents and electromagnetic induction effects like metals. On the contrary, non-metallic objects may affect the field intensity distribution by changing the electromagnetic properties of the surrounding medium (such as dielectric constant) or physically blocking the propagation of electromagnetic waves. Therefore, for electromagnetic field changes that are determined to be slow and / or continuous, it is determined that the current fault source is related to non-metals, and the specific non-metal-related fault source category is identified in the fault source category to be tested.

[0209] In practical applications, the characteristics of different types of non-metallic fault sources also need to be considered: Tree obstacles may show seasonal changes (such as leaf growth cycles) and random fluctuations under the influence of wind. Plastic foreign objects (such as kites): may show intermittent, weak field strength changes. Meteorological factors (such as fog, rain): may cause slow changes in the overall field strength and are correlated with meteorological data.

[0210] This implementation can distinguish different types of interference sources by accurately capturing the temporal characteristics of electromagnetic field changes, and fully utilizes the physical characteristics of the interaction between metal objects and electromagnetic fields to achieve high-accuracy fault source classification, effectively ensuring the timeliness and accuracy of early warnings, and further effectively improving the safety of transmission lines.

[0211] In one implementation of this embodiment, the meteorological data includes wind direction data and wind speed data. According to the meteorological data of the future time period, the dancing amplitude and the dancing frequency of the power transmission line in the future time period are predicted, including the following steps:

[0212] S710: Input the wind direction data and wind speed data of the future time period into a preset dynamic equation, and solve the dynamic equation to obtain the dancing amplitude and dancing frequency of the conductor of the transmission line at each time point in the future time period.

[0213] The wind direction data and wind speed data of the future time period are input into a preset dynamic equation, and the equation is solved to predict the dancing amplitude and dancing frequency of the transmission line conductor at each time point in the future time period. The dynamic equation in this embodiment is as follows:

[0214]

[0215] Where y is the vertical displacement of the wire at position x and time t, ρ is the linear density of the wire, A is the cross-sectional area of ​​the wire, E is Young's modulus, I is the moment of inertia, T is the tension of the wire, and F(x, t) is the wind force acting on the wire.

[0216] The wind force F(x, t) can be expressed as:

[0217] F(x,t)=0.5×Cd×ρa×d×V(t)2 ×sin(θ(t));

[0218] Among them, Cd is the drag coefficient, ρa is the air density, d is the wire diameter, V(t) is the wind speed, and θ(t) is the angle between the wind direction and the perpendicular direction of the wire.

[0219] Substitute the wind direction data θ(t) and wind speed data V(t) of the future time period into the above equation. The finite element method is used to solve the above partial differential equations to discretize the wire into multiple units and transform the original problem into a set of ordinary differential equations. The time integration method can be used to iteratively solve the set of ordinary differential equations to obtain the displacement q(t) of the wire at each time point. From the displacement data, the dancing amplitude (maximum displacement amplitude) and the dancing frequency (main frequency component of the displacement) can be extracted.

[0220] This implementation method combines weather forecast data with an accurate physical model to accurately predict the swing amplitude and frequency of the conductor in the future time period, not only taking into account the changes in wind direction and wind speed, but also simulating the complex dynamic behavior of the conductor. This makes it easier for power grid managers to identify potential risks in advance and formulate targeted preventive measures, thereby significantly improving the safety and reliability of transmission lines and improving the overall operating efficiency of the power grid.

[0221] In one implementation of this embodiment, combining the dancing amplitude, dancing frequency and motion path, determining the target probability of the target fault source causing a secondary fault on the transmission line, and determining the high-risk section of the transmission line, includes the following steps:

[0222] S810, determining a stress concentration area on the transmission line within a target area according to the dancing swing amplitude and the dancing frequency, the target area including the occurrence area, and the stress concentration area is used to characterize a high-risk section of the transmission line;

[0223] S820, determining the number of overlaps between the stress concentration area on the transmission line and the movement path of the target fault source;

[0224] S830: Divide the number of overlaps by the length of the power transmission lines in the target area to obtain a value as a target probability of a target fault source.

[0225] The stress concentration area in the target area is determined based on the dancing swing amplitude and dancing frequency of the transmission line. The stress concentration area is regarded as a high-risk section of the transmission line. The dancing swing amplitude reflects the maximum distance that the conductor deviates from the equilibrium position, while the dancing frequency indicates the speed at which the conductor vibrates. The dancing swing amplitude and the dancing frequency together determine the dynamic stress borne by the conductor. Generally speaking, larger swing amplitudes and higher frequencies result in greater stress. In order to determine the stress concentration area, the stress distribution can be calculated along the length of the conductor. The specific implementation steps can be 1. Discretize the conductor into multiple units. 2. For each unit, calculate the local dancing swing amplitude and frequency according to its position. 3. Use the preset stress calculation formula to calculate the dynamic stress of each unit. 4. Set a stress threshold and mark the area exceeding this threshold as a stress concentration area.

[0226] Afterwards, the number of overlaps between the stress concentration areas on the transmission line and the movement path of the target fault source is determined to quantify the possibility of secondary faults.

[0227] First, the stress concentration area determined in S810 and the target fault source movement path need to be represented in the same spatial coordinate system. The stress concentration area can be represented as a series of line segments or intervals on the transmission line, and the movement path of the fault source can be represented as a continuous curve or a discrete point sequence. Next, a spatial analysis is required to determine the overlap. The specific implementation steps may be: 1. Represent the stress concentration area as a series of discrete points, each point represents a small area unit. If the movement path of the fault source is a continuous curve, it is also discretized into a point sequence. 2. Use spatial indexing technology, such as R-tree or quadtree, to organize the spatial data of the stress concentration area. 3. For each point on the fault source movement path, search for stress concentration area points nearby. If the distance between a point on the fault source path and a stress concentration area point is less than d, it is considered that an overlap occurs. 4. Count the number of times the overlap condition is met.

[0228] The target probability is calculated as follows:

[0229] P = N / L;

[0230] Among them, P is the target probability, N is the number of overlaps, and L is the total length of the transmission line in the target area. In other words, the more overlaps there are, or the higher the overlap ratio relative to the total length of the line, the greater the probability of a secondary failure.

[0231] This implementation method can significantly improve the safety and reliability of transmission lines, reduce the possibility of accidents, and ultimately ensure the stable operation of the power system by determining the target probability of a target fault source causing a secondary fault on the transmission line and determining the high-risk sections of the transmission line.

[0232] In one implementation of this embodiment, determining the stress concentration area on the transmission line in the target area according to the dancing swing amplitude and the dancing frequency includes the following steps:

[0233] S910, using a preset stress calculation formula to calculate stress values ​​of the conductor of the transmission line at different dancing swing amplitudes and dancing frequencies;

[0234] S920. The position with the maximum stress value is regarded as the stress concentration area.

[0235] This embodiment first uses a preset stress calculation formula to calculate the stress value of the transmission line conductor under different dancing swing amplitudes and dancing frequencies, so as to quantify the mechanical stress of the conductor under dynamic conditions through a physical model. Specifically, first, a dynamic stress model suitable for the transmission line conductor is constructed. The dynamic stress model in this implementation is constructed based on the catenary theory, and the stress calculation formula is as follows:

[0236] σ t =σ s +σ d ;

[0237] Among them, σ t is the total stress, σ s is the static stress, σ d is the dynamic stress.

[0238] Static stress σ s It can be calculated by the catenary equation, which is as follows:

[0239] σ s=(w×L 2 ) / (8×d);

[0240] Where w is the weight per unit length of the conductor, L is the span length, and d is the conductor sag. Dynamic stress σ d Then we need to consider the influence of dancing amplitude and dancing frequency. The following formula can be used:

[0241] σ d =E×y×(2π×f) 2 ×A / g;

[0242] Among them, E is the Young's modulus of the wire, y is the dancing swing amplitude, f is the dancing frequency, A is the cross-sectional area of ​​the wire, and g is the acceleration due to gravity.

[0243] In practical applications, the transmission line needs to be discretized and divided into multiple calculation units. For each unit, the dancing swing, frequency and environmental parameters of different sub-areas of the target area are used for calculation. This can be achieved by the following steps: 1. Divide the transmission line into N equal-length units. 2. For each unit i (i = 1, 2, ..., N): a. Obtain the local dancing swing y of the unit i and frequency f i b. Obtain the environmental parameters (temperature, wind speed, etc.) of each sub-area. c. Apply the above stress calculation formula to calculate σ f 3. Store the stress value of each unit to form a stress distribution map of the entire transmission line in the sub-area.

[0244] Next, through the stress distribution diagram, we can identify the key areas that are most prone to failure or damage, that is, the stress concentration areas. Specifically, first traverse all calculation units to find the location with the largest stress value. You can use the comparison algorithm:

[0245] σ max =max(σ f ), i=1, 2, ..., N;

[0246] Where N is the total number of units.

[0247] Secondly, a stress threshold is set based on the yield strength or fatigue limit of the wire material. For example, 80% of the material yield strength can be taken as the threshold. The continuous area where the stress value exceeds the stress threshold is defined as the stress concentration area. The steps are as follows: a. Initialize an empty stress concentration area list. b. Traverse all calculation units, and when the stress value exceeds the stress threshold, start a new stress concentration area. c. Traverse adjacent units until the stress value drops below the stress threshold, at which time the definition of the current stress concentration area is terminated. d. Repeat steps b and c until all units are traversed and all stress concentration areas are obtained.

[0248] For each identified stress concentration area, the line length covered by the stress concentration area, the maximum stress value in the stress concentration area, and the rate of stress change in the stress concentration area are determined, and weights are assigned to the line length covered by the stress concentration area, the maximum stress value in the stress concentration area, and the rate of stress change in the stress concentration area, respectively. The stress values ​​used to characterize the stress concentration area can be obtained, the stress values ​​are sorted, and the stress concentration area corresponding to the largest stress value is taken as the key stress concentration area, indicating that the key stress concentration area has the greatest risk.

[0249] It should be noted that for special structural locations, such as insulator connections and special spans across obstacles, special attention still needs to be paid even if the calculated stress value is not the highest.

[0250] This refined analysis can provide precise guidance for the maintenance and overhaul of transmission lines. For example, targeted inspection plans can be developed to prioritize high-weight stress concentration areas or install additional monitoring equipment in these areas. This method can also be used to evaluate the effectiveness of improvement measures and quantify safety improvements by comparing stress distribution before and after improvements.

[0251] By determining the stress concentration area, this implementation method can accurately locate the high-risk areas on the transmission line, which helps to improve the safety and reliability of the transmission line, reduce the risk of failure, optimize resource allocation, and ultimately achieve the goal of improving the stability of the entire power system.

[0252] The present application also provides a transmission line fault identification and early warning system, including:

[0253] a memory configured to store instructions; and

[0254] The processor is configured to call instructions from the memory and implement the above-mentioned fault identification and early warning method for the power transmission line when executing the instructions.

[0255] An embodiment of the present application also provides a machine-readable storage medium, on which instructions are stored, and the instructions are used to enable a machine to execute the above-mentioned transmission line fault identification and early warning method.

[0256] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may 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.

[0257] 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 generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0258] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate 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 A function specified in one or more boxes.

[0259] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions 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.

[0260] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0261] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0262] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0263] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0264] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for fault identification and early warning of a power transmission line, characterized in that: include: In response to receiving a fault flashover signal, determining an occurrence area of ​​the fault flashover signal, and acquiring a target image of the occurrence area; Determining whether there is abnormal discharge in the power transmission line in the occurrence area; In the case where abnormal discharge exists in the power transmission line in the occurrence area, image recognition is performed on the target image to obtain a first image recognition result, wherein the first image recognition result includes at least one fault source to be detected; In response to receiving the target parameters sent by the laser radar, obtaining the inclination of the tower of the power transmission line in the occurrence area and the electromagnetic field change of the power transmission line at the current moment; Determine a movement path of a current fault source according to the target parameter, and determine a fault source category of the current fault source according to the movement path, the target parameter, the inclination of the tower, and the electromagnetic field change; According to the fault source category, determining a target fault source among the fault sources to be tested, wherein the target fault source is used to characterize the current fault source; Acquire meteorological data for a future time period, and predict the dancing amplitude and dancing frequency of the power transmission line for the future time period based on the meteorological data for the future time period; Determine the target probability of the target fault source causing a secondary fault on the transmission line in combination with the dancing swing amplitude, the dancing frequency and the motion path, and determine the high-risk section of the transmission line; The target probability and the high-risk section of the transmission line are output.

2. The method according to claim 1, characterized in that The determining whether there is abnormal discharge in the power transmission line in the occurrence area comprises: In response to receiving an electromagnetic field signal sent by an electromagnetic field sensor, performing a fast Fourier transform on the electromagnetic field signal to obtain a spectrum diagram; Extracting spectrum features of the spectrum graph, wherein the spectrum features include amplitude distribution, spectrum graph symmetry, phase distribution, spectrum energy distribution, bandwidth, fundamental frequency harmonic components, and spectrum tilt; In the discharge type feature library, a target spectrum feature matching the amplitude distribution, the spectrum graph symmetry and the fundamental frequency harmonic component is determined according to a preset similarity algorithm, and a target discharge type corresponding to the target spectrum feature is determined; Acquiring a reference feature of the target discharge type, and determining whether the target discharge type is an abnormal discharge according to the reference feature; Among them, the benchmark features include benchmark phase distribution, benchmark spectral energy distribution, benchmark frequency bandwidth and benchmark spectral inclination. When there are non-fundamental harmonic components and at least two of the benchmark features are inconsistent with the corresponding spectral features, the target discharge type is determined to be abnormal discharge.

3. The method according to claim 1, characterized in that The step of determining a movement path of a current fault source according to the target parameter, and determining a fault source category of the current fault source according to the movement path, the target parameter, the inclination of the tower, and the change in the electromagnetic field, includes: Obtain the line height and line location of the transmission line in the occurrence area; Determine a movement path of the current fault source according to the target parameter, and determine an influence factor and influence range of the current fault source on the transmission line according to the movement path, the target parameter, the inclination of the tower, the line height and the line position, wherein the target parameter includes height data, length data and speed data of the current fault source; Determine at least one category of a fault source to be tested by combining the influencing factor and the influencing range; According to the change in the electromagnetic field, a fault source category of a current fault source is determined in at least one of the fault source categories to be detected.

4. The method according to claim 3, characterized in that Determining the movement path of the current fault source according to the target parameter, and determining the influence factor and influence range of the current fault source on the transmission line according to the movement path, the target parameter, the inclination of the tower, the line height and the line position, includes: Constructing a motion model of the current fault source according to the height data, length data and speed data of the current fault source; Using the motion model, simulating the position change of the current fault source within a preset time period; Determine the spatial occupied area at each time point by combining the height data and the length data; Determine the movement path of the current fault source according to the position change of the current fault source within a preset time period and the spatial occupied area at each time point; Simulating the movement path of the current fault source, the line height and the line position to obtain the intersection and overlapping area between the current fault source and the transmission line, the length of the overlapping area being the contact length, and the area of ​​the overlapping area being used to characterize the impact range; Using a contact time calculation formula, according to the contact length and the speed data, the contact time between the current fault source and the transmission line is calculated; The direction of the force applied by the current fault source to the transmission line is determined according to the inclination of the tower and the speed data, wherein the contact time and the direction of the force are used to characterize the influencing factor.

5. The method according to claim 4, characterized in that The determining at least one type of fault source to be detected by combining the influencing factor and the influencing range includes: Determine the nature of the current fault source according to the contact time and the direction of the force, wherein the nature of the current fault source includes instantaneous impact and continuous interference; In the case where the nature of the current fault source is an instantaneous impact, determining at least one type of fault source to be tested in combination with the impact range; or In the case where the nature of the current fault source is continuous interference, determining at least one category of the fault source to be tested in combination with the impact range; The categories of fault sources to be tested include over-height vehicles, cranes, tree obstacles and foreign objects.

6. The method according to claim 5, characterized in that Determining the fault source category in at least one of the fault source categories to be detected according to the change in the electromagnetic field includes: Determining the properties of the electromagnetic field change, the properties of the electromagnetic field change including instantaneous properties, slow properties and continuous properties; In the case where the property of the electromagnetic field change is instantaneous, determining that the current fault source is related to metal, and determining a fault source category related to metal in at least one of the fault source categories to be tested; In the case that the property of the electromagnetic field change is slow and / or continuous, it is determined that the current fault source is related to non-metal, and a fault source category related to non-metal is determined in at least one of the fault source categories to be tested.

7. The method according to claim 1, characterized in that The meteorological data includes wind direction data and wind speed data. The prediction of the swing amplitude and the swing frequency of the power transmission line in the future time period based on the meteorological data in the future time period includes: The wind direction data and the wind speed data of the future time period are input into a preset dynamic equation, and the dynamic equation is solved to obtain the dancing amplitude and dancing frequency of the conductor of the transmission line at each time point in the future time period.

8. The method according to claim 1, characterized in that: The step of combining the dancing amplitude, the dancing frequency and the motion path to determine the target probability of the target fault source causing a secondary fault on the transmission line and determining a high-risk section of the transmission line includes: Determine a stress concentration area on the transmission line within a target area according to the dancing swing amplitude and the dancing frequency, wherein the target area includes the occurrence area, and the stress concentration area is used to characterize a high-risk section of the transmission line; Determining the amount of overlap between the stress concentration area on the transmission line and the movement path of the target fault source; A value obtained by dividing the overlap number by the length of the power transmission line in the target area is used as the target probability of the target fault source.

9. The method according to claim 8, characterized in that The step of determining the stress concentration area on the transmission line in the target area according to the dancing swing amplitude and the dancing frequency includes: Using a preset stress calculation formula to calculate the stress value of the conductor of the power transmission line at different dancing swing amplitudes and dancing frequencies; The position where the stress value is the largest is regarded as the stress concentration area.

10. A fault identification and early warning system for a power transmission line, characterized in that: include: a memory configured to store instructions; as well as A processor is configured to call the instruction from the memory and implement the fault identification and early warning method for a power transmission line according to any one of claims 1 to 9 when executing the instruction.

Images