Lightning protection method and system for high-altitude power transmission line

Abstract

The invention relates to the technical field of power system safety protection, discloses a lightning protection method and system for a high-altitude power transmission line, and aims to solve the problem that a data acquisition method of existing terminal equipment is poor in accuracy and safety. The scheme mainly comprises the steps of collecting image data of a power transmission line and current data and voltage data in the power transmission line according to a preset period; judging whether lightning stroke occurs or not according to the image data, and if yes, judging the lightning type according to the energy characteristics of different components in the current data; obtaining a lightning stroke risk coefficient according to the image data, the current data, the voltage data and the lightning type and based on a pre-trained lightning stroke risk identification model; and determining whether to open the power transmission line according to the lightning stroke risk coefficient. Risk quantification is achieved, the maloperation rate is reduced, and the method is particularly suitable for high-altitude areas where line patrol is difficult.

Description

Technical Field

[0001] This invention relates to the field of power system safety protection technology, specifically to a method and system for lightning protection of high-altitude transmission lines. Background Technology

[0002] With the development of hydropower, mountain wind power, and photovoltaic power, high-altitude overhead transmission lines are widely used. Lightning strikes pose a significant threat to power equipment in high-altitude areas, potentially causing damage, power outages, and even casualties. Overhead transmission lines, due to their long routes and the complex terrain they traverse, are particularly vulnerable to lightning strikes.

[0003] Traditional lightning protection typically employs a combination of lightning conductors, surge arresters, and overcurrent protection devices. However, these methods have significant drawbacks: First, traditional protection devices usually operate based on current amplitude thresholds, failing to effectively distinguish between direct lightning strikes, induced lightning strikes, and other overvoltage faults, which can easily lead to false trips or failures to operate, reducing power supply reliability. Second, the lack of multi-dimensional information fusion makes it impossible to accurately assess lightning strike risks or achieve adaptive protection control based on risk assessment. Summary of the Invention

[0004] This invention aims to address the problems of high false alarm rates and lack of intelligent risk assessment mechanisms in existing lightning protection methods, and proposes a lightning protection method and system for high-altitude transmission lines.

[0005] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0006] In a first aspect, the present invention provides a method for lightning protection of high-altitude transmission lines, the method comprising:

[0007] Image data of the transmission line, as well as current and voltage data in the transmission line, are collected according to a preset cycle;

[0008] The image data is used to determine whether a lightning strike has occurred. If so, the lightning type is determined based on the energy characteristics of different components in the current data. The lightning type includes induced lightning and direct lightning strike.

[0009] Based on the image data, current data, voltage data, and lightning type, and using a pre-trained lightning risk identification model, a lightning risk coefficient is obtained.

[0010] The decision to disconnect the power transmission line is based on the lightning strike risk coefficient.

[0011] Further, determining whether a lightning strike has occurred based on the image data includes:

[0012] When current or voltage data changes abruptly, an image recognition algorithm is used to detect whether there is an electric arc, smoke, or line damage in the image. If so, it is determined that a lightning strike has occurred.

[0013] Furthermore, the lightning type is determined based on the energy characteristics of different components in the current data, including:

[0014] The current data is decomposed into multiple characteristic quantities of different frequencies. The ratio of non-oscillating decay current to high-frequency oscillating decay current is calculated based on the decomposition results. If the ratio is greater than 1, it is determined to be an induced lightning strike. If the ratio is less than 1, it is determined to be a direct lightning strike.

[0015] Furthermore, the decomposition formula for the current data is as follows:

[0016] ;

[0017] in, express Total current at time 10:00 This represents the total number of frequency components. Indicates the first The current amplitude of each frequency component, Represents an empirical constant. Indicates the first The propagation speed of each frequency component during the main discharge process. Indicates the length of the current path. Represents the natural constant. Represents the imaginary unit. Indicates the first The frequency of each frequency component Indicates the first The initial phase angle of each frequency component, Indicates the first The attenuation coefficient of each frequency component, Indicates time.

[0018] Furthermore, the formula for calculating the ratio is as follows:

[0019] ;

[0020] in, This represents the ratio of the non-oscillating decaying current to the high-frequency oscillating decaying current. This represents the amplitude of the non-oscillating decaying current. The attenuation coefficient represents the attenuation coefficient of the non-oscillating decaying current. This represents the amplitude of the high-frequency oscillation decay current. This indicates the frequency of the high-frequency oscillating decay current. This represents the attenuation coefficient of the high-frequency oscillation decay current.

[0021] Furthermore, the training method for the lightning strike risk identification model includes:

[0022] Collect historical lightning strike sample data, which includes historical image data, historical current data, historical voltage data, historical lightning type labels, and corresponding historical lightning strike consequence labels;

[0023] The historical lightning strike sample data is preprocessed and features are extracted to construct a sample feature set, wherein the features include visual damage features extracted from the historical image data, electrical features extracted from the historical current data and historical voltage data, and historical lightning type labels;

[0024] Using the sample feature set as input and the historical lightning strike consequence labels as output targets, the preset machine learning model is trained and validated to obtain the lightning strike risk identification model.

[0025] Furthermore, determining whether to trip the transmission line based on the aforementioned lightning strike risk coefficient includes:

[0026] The lightning strike risk coefficient is compared with a preset risk threshold.

[0027] If the lightning strike risk coefficient is greater than or equal to the risk threshold, a tripping command is generated and sent to the circuit breaker actuator corresponding to the transmission line.

[0028] If the lightning strike risk coefficient is less than the risk threshold, the transmission line will remain closed and monitoring will continue.

[0029] Furthermore, the method also includes a model update step:

[0030] The image data, current data, voltage data, determined lightning type, and final lightning risk coefficient of this lightning strike will be used as a new training sample.

[0031] The new training samples are used to incrementally learn the pre-trained lightning risk identification model in order to update the parameters of the lightning risk identification model.

[0032] Furthermore, after determining whether to trip the transmission line based on the lightning strike risk coefficient, the method further includes:

[0033] Generate lightning protection maintenance recommendations, the content of which is determined based on the lightning type and lightning strike risk factor.

[0034] The lightning protection maintenance recommendations are associated with and stored along with the image data, location information, and time of the lightning strike, and then output.

[0035] In a second aspect, the present invention provides a lightning protection system for high-altitude transmission lines, used to implement the lightning protection method for high-altitude transmission lines as described in the first aspect, the system comprising:

[0036] The acquisition module is used to acquire image data of the transmission line, as well as current and voltage data in the transmission line, according to a preset cycle.

[0037] The judgment module is used to determine whether a lightning strike has occurred based on the image data. If so, it determines the lightning type based on the energy characteristics of different components in the current data. The lightning type includes induced lightning and direct lightning strike.

[0038] The identification module is used to obtain the lightning risk coefficient based on the image data, current data, voltage data and lightning type, and on a pre-trained lightning risk identification model.

[0039] The tripping module is used to determine whether to trip the transmission line based on the lightning strike risk coefficient.

[0040] The beneficial effects of this invention are as follows: The lightning protection method and system for high-altitude transmission lines provided by this invention greatly reduce the risk of false tripping caused by single electrical quantity detection through multi-source data fusion judgment. It only trips when necessary, ensuring continuous power supply to the power grid and avoiding losses to lines and equipment caused by unnecessary switching operations and short-circuit current impacts, thus extending the service life of the equipment. By introducing the Pony algorithm and lightning risk identification model, it can accurately distinguish the type of lightning strike and precisely quantify the risk, realizing intelligent risk assessment and making the protection strategy more scientific and reasonable. Attached Figure Description

[0041] Figure 1 A schematic flowchart of a lightning protection method for high-altitude transmission lines provided in an embodiment;

[0042] Figure 2 This is a schematic diagram of the lightning protection system for a high-altitude power transmission line provided in an embodiment. Detailed Implementation

[0043] Currently, lightning protection often employs a combination of lightning rods, surge arresters, and overcurrent protection devices. These methods typically operate based on current amplitude thresholds, resulting in a high false trip rate and poor power supply reliability and safety.

[0044] Based on this, the technical solution of this invention is proposed. In this invention, firstly, multi-dimensional data, including image data, current data, and voltage data, are collected to form a dual verification of visual and electrical signals, improving the accuracy of lightning strike identification. Then, the image data is used to determine whether a lightning strike event has occurred. After confirming that a lightning strike has occurred, the transient traveling wave component in the current data is extracted. The current signal is decomposed into sub-components of different frequencies through spectrum analysis, and parameters such as amplitude, frequency, and attenuation coefficient of each component are calculated. If non-oscillating low-frequency energy dominates, it is determined to be induced lightning; if high-frequency oscillating energy dominates, it is determined to be direct lightning. Then, the image data, current data, voltage data, and lightning type are input into a lightning strike risk identification model pre-trained with a large amount of historical data to obtain a comprehensive lightning strike risk coefficient. This coefficient is a value between 0 and 1, representing the probability that this lightning strike will cause a permanent fault in the line, leading to tripping or even system collapse. For a high-risk coefficient, the circuit breaker will be immediately tripped to protect the main power grid from impact and prevent the fault from spreading. For a low-risk coefficient, the line will remain closed and monitoring will continue. Because many lightning strikes only cause momentary faults, the electric arc can extinguish itself, and the line does not need to be shut down, thus ensuring the continuity of power supply.

[0045] The technical solutions in this embodiment will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0046] Figure 1 A schematic diagram of a lightning protection method for high-altitude power transmission lines is shown. Please refer to [link / reference]. Figure 1 The method includes the following steps:

[0047] Step 1: Collect image data of the transmission line, as well as current and voltage data of the transmission line, according to the preset cycle.

[0048] In practical applications, high-definition cameras or drones can be used for automatic inspections to periodically acquire image data of components such as the line appearance, insulators, and towers. Current and voltage data of the line can be collected in real time through current transformers and voltage transformers installed on the line.

[0049] Step 2: Determine whether a lightning strike has occurred based on the image data. If so, determine the lightning type based on the energy characteristics of different components in the current data. The lightning type includes induced lightning and direct lightning strike.

[0050] In this embodiment, determining whether a lightning strike has occurred based on the image data includes:

[0051] When current or voltage data changes abruptly, an image recognition algorithm is used to detect whether there is an electric arc, smoke, or line damage in the image. If so, it is determined that a lightning strike has occurred.

[0052] In practical applications, continuous monitoring of current / voltage data is performed. If a drastic change in voltage, measured in microseconds or even nanoseconds, is detected, image recognition algorithms (such as computer vision models) are invoked to analyze image data from the same time period. Criteria include: arc flash recognition (detecting extremely bright, momentary flashes in the image); smoke and dust recognition (detecting whether smoke or flying debris is produced at the lightning strike point); and line damage recognition (comparing images before and after the lightning strike to check for insulator bursts, conductor erosion points, etc.). If arc flash, smoke, or line damage is detected, a lightning strike is confirmed, effectively eliminating interference signals caused by internal equipment malfunctions, switching operations, or other reasons.

[0053] In this embodiment, determining the lightning type based on the energy characteristics of different components in the current data includes:

[0054] The current data is decomposed into multiple characteristic quantities of different frequencies. The ratio of non-oscillating decay current to high-frequency oscillating decay current is calculated based on the decomposition results. If the ratio is greater than 1, it is determined to be an induced lightning strike. If the ratio is less than 1, it is determined to be a direct lightning strike.

[0055] The decomposition formula for the current data is as follows:

[0056] ;

[0057] in, express Total current at time 10:00 This represents the total number of frequency components. Indicates the first The current amplitude of each frequency component, Represents an empirical constant. Indicates the first The propagation speed of each frequency component during the main discharge process. Indicates the length of the current path. Represents the natural constant. Represents the imaginary unit. Indicates the first The frequency of each frequency component Indicates the first The initial phase angle of each frequency component, Indicates the first The attenuation coefficient of each frequency component, Indicates time.

[0058] The formula for calculating the ratio is as follows:

[0059] ;

[0060] in, This represents the ratio of the non-oscillating decaying current to the high-frequency oscillating decaying current. This represents the amplitude of the non-oscillating decaying current. The attenuation coefficient represents the attenuation coefficient of the non-oscillating decaying current. This represents the amplitude of the high-frequency oscillation decay current. This indicates the frequency of the high-frequency oscillating decay current. This represents the attenuation coefficient of the high-frequency oscillation decay current.

[0061] When lightning strikes, the different proportions of electrostatic induction and electromagnetic induction energy caused by different types of lightning strikes can help determine the type of lightning. Specifically, after confirming a lightning strike, the transient traveling wave component in the current data is extracted. Spectral analysis (such as Fourier transform or wavelet transform) is used to decompose the current signal into sub-components of different frequencies, and the amplitude, frequency, attenuation coefficient, and other parameters of each component are calculated. Then, based on the decomposition results, the ratio is calculated using a ratio calculation formula. If the ratio is greater than 1, it indicates that non-oscillating low-frequency energy dominates, and it is identified as induced lightning (the amplitude of induced lightning current is relatively low, but the high-frequency component is abundant). If the ratio is less than 1, it indicates that high-frequency oscillating energy dominates, and it is identified as direct lightning (the amplitude of direct lightning current is extremely high, and its enormous energy is mainly manifested in the low-frequency part).

[0062] Step 3: Based on the image data, current data, voltage data, and lightning type, and using a pre-trained lightning risk identification model, obtain the lightning risk coefficient.

[0063] It is understandable that the lightning strike risk identification model is a machine learning model (such as a neural network or random forest) pre-trained with a large amount of historical data. The training methods include:

[0064] Collect historical lightning strike sample data, which includes historical image data, historical current data, historical voltage data, historical lightning type labels, and corresponding historical lightning strike consequence labels;

[0065] The historical lightning strike sample data is preprocessed and features are extracted to construct a sample feature set, wherein the features include visual damage features extracted from the historical image data, electrical features extracted from the historical current data and historical voltage data, and historical lightning type labels;

[0066] Using the sample feature set as input and the historical lightning strike consequence labels as output targets, the preset machine learning model is trained and validated to obtain the lightning strike risk identification model.

[0067] The trained lightning strike risk identification model can extract electrical features such as amplitude, energy, waveform steepness, and frequency components from current / voltage data, and visual features such as arc size, damage degree, and lightning strike location from image data. Using electrical features, visual features, and type features as input features, and employing complex mathematical algorithms, it finds hidden and complex nonlinear patterns from massive multidimensional features, and outputs a lightning strike risk coefficient. This coefficient is a value between 0 and 1, representing the probability that the lightning strike will cause a permanent fault in the line, leading to tripping or even system collapse.

[0068] Step 4: Determine whether to trip the power transmission line based on the lightning strike risk coefficient.

[0069] In this embodiment, the lightning strike risk coefficient is compared with a preset risk threshold; if the lightning strike risk coefficient is greater than or equal to the risk threshold, a tripping command is generated and sent to the circuit breaker actuator corresponding to the transmission line; if the lightning strike risk coefficient is less than the risk threshold, the transmission line remains closed and monitoring continues.

[0070] Specifically, for high-risk scenarios, the circuit breaker will be immediately tripped to protect the main power grid from impact and prevent the fault from escalating. For low-risk scenarios, the line will remain closed while monitoring continues. Because many lightning strikes only cause momentary faults and the arc can extinguish itself, the line does not need to be shut down, ensuring the continuity of power supply.

[0071] In this embodiment, the method further includes a model update step:

[0072] The image data, current data, voltage data, determined lightning type, and final lightning risk coefficient of this lightning strike are used as a new training sample. The pre-trained lightning risk identification model is incrementally learned using the new training sample to update the parameters of the lightning risk identification model.

[0073] It's understandable that lightning strike characteristics may vary slightly in different high-altitude areas. By continuously collecting local data and updating the model, the model can become increasingly adapted to the actual conditions of that specific route, solving the problem of performance degradation in machine learning models after deployment due to changes in data distribution.

[0074] In this embodiment, after determining whether to trip the transmission line based on the lightning strike risk coefficient, the method further includes:

[0075] A lightning protection maintenance recommendation is generated, the content of which is determined based on the lightning type and lightning strike risk coefficient. The lightning protection maintenance recommendation is then associated with and stored along with the image data, location information, and time of the lightning strike, and then output.

[0076] Through the above process, maintenance personnel no longer need to manually analyze the severity and type of faults from massive amounts of alarm information. The process directly provides diagnostic conclusions and preliminary repair directions, greatly shortening fault response and processing time. It is particularly suitable for areas where line patrol is difficult, such as high altitudes.

[0077] In summary, the lightning protection method for high-altitude transmission lines provided in this embodiment, through multi-source data fusion judgment, greatly reduces the risk of false tripping caused by single electrical quantity detection. It only trips when necessary, ensuring continuous power supply to the grid and avoiding losses to lines and equipment caused by unnecessary switching operations and short-circuit current surges, thus extending the service life of the equipment. By introducing the Pony algorithm and lightning risk identification model, it can accurately distinguish the type of lightning strike and precisely quantify the risk, realizing intelligent risk assessment and making the protection strategy more scientific and reasonable.

[0078] Based on the above technical solution, this embodiment also proposes a lightning protection system for high-altitude transmission lines, used to implement the lightning protection method for high-altitude transmission lines as described in the embodiment. Please refer to [link to relevant documentation]. Figure 2 The system includes:

[0079] The acquisition module is used to acquire image data of the transmission line, as well as current and voltage data in the transmission line, according to a preset cycle.

[0080] The judgment module is used to determine whether a lightning strike has occurred based on the image data. If so, it determines the lightning type based on the energy characteristics of different components in the current data. The lightning type includes induced lightning and direct lightning strike.

[0081] The identification module is used to obtain the lightning risk coefficient based on the image data, current data, voltage data and lightning type, and on a pre-trained lightning risk identification model.

[0082] The tripping module is used to determine whether to trip the transmission line based on the lightning strike risk coefficient.

[0083] It is understood that since the lightning protection system for high-altitude transmission lines described in this embodiment is a system for implementing the lightning protection method for high-altitude transmission lines described in the embodiment, the system disclosed in the embodiment is relatively simple to describe because it corresponds to the method disclosed in the embodiment. For relevant parts, please refer to the description of the method, and it will not be repeated here.

Claims

1. A method for lightning protection of high-altitude transmission lines, characterized in that, The method includes: Image data of the transmission line, as well as current and voltage data in the transmission line, are collected according to a preset cycle; The image data is used to determine whether a lightning strike has occurred. If so, the lightning type is determined based on the energy characteristics of different components in the current data. The lightning type includes induced lightning and direct lightning strike. Based on the image data, current data, voltage data, and lightning type, and using a pre-trained lightning risk identification model, a lightning risk coefficient is obtained. The decision to disconnect the power transmission line is based on the lightning strike risk coefficient.

2. The lightning protection method for high-altitude transmission lines according to claim 1, characterized in that, Determining whether a lightning strike has occurred based on the image data includes: When current or voltage data changes abruptly, an image recognition algorithm is used to detect whether there is an electric arc, smoke, or line damage in the image. If so, it is determined that a lightning strike has occurred.

3. The lightning protection method for high-altitude transmission lines according to claim 1, characterized in that, The lightning type is determined based on the energy characteristics of different components in the current data, including: The current data is decomposed into multiple characteristic quantities of different frequencies. The ratio of non-oscillating decay current to high-frequency oscillating decay current is calculated based on the decomposition results. If the ratio is greater than 1, it is determined to be an induced lightning strike. If the ratio is less than 1, it is determined to be a direct lightning strike.

4. The lightning protection method for high-altitude transmission lines according to claim 3, characterized in that, The decomposition formula for the current data is as follows: ； in, express Total current at time 10:00 This represents the total number of frequency components. Indicates the first The current amplitude of each frequency component, Represents an empirical constant. Indicates the first The propagation speed of each frequency component during the main discharge process. Indicates the length of the current path. Represents the natural constant. Represents the imaginary unit. Indicates the first The frequency of each frequency component Indicates the first The initial phase angle of each frequency component, Indicates the first The attenuation coefficient of each frequency component, Indicates time.

5. The lightning protection method for high-altitude transmission lines according to claim 4, characterized in that, The formula for calculating the ratio is as follows: ； in, This represents the ratio of the non-oscillating decaying current to the high-frequency oscillating decaying current. This represents the amplitude of the non-oscillating decaying current. The attenuation coefficient represents the attenuation coefficient of the non-oscillating decaying current. This represents the amplitude of the high-frequency oscillation decay current. This indicates the frequency of the high-frequency oscillating decay current. This represents the attenuation coefficient of the high-frequency oscillation decay current.

6. The lightning protection method for high-altitude transmission lines according to claim 1, characterized in that, The training method for the lightning strike risk identification model includes: Collect historical lightning strike sample data, which includes historical image data, historical current data, historical voltage data, historical lightning type labels, and corresponding historical lightning strike consequence labels; The historical lightning strike sample data is preprocessed and features are extracted to construct a sample feature set, wherein the features include visual damage features extracted from the historical image data, electrical features extracted from the historical current data and historical voltage data, and historical lightning type labels; Using the sample feature set as input and the historical lightning strike consequence labels as output targets, the preset machine learning model is trained and validated to obtain the lightning strike risk identification model.

7. The lightning protection method for high-altitude transmission lines according to claim 1, characterized in that, Determining whether to trip the transmission line based on the aforementioned lightning strike risk coefficient includes: The lightning strike risk coefficient is compared with a preset risk threshold. If the lightning strike risk coefficient is greater than or equal to the risk threshold, a tripping command is generated and sent to the circuit breaker actuator corresponding to the transmission line. If the lightning strike risk coefficient is less than the risk threshold, the transmission line will remain closed and monitoring will continue.

8. The lightning protection method for high-altitude transmission lines according to claim 1, characterized in that, The method also includes a model update step: The image data, current data, voltage data, determined lightning type, and final lightning risk coefficient of this lightning strike will be used as a new training sample. The new training samples are used to incrementally learn the pre-trained lightning risk identification model in order to update the parameters of the lightning risk identification model.

9. The lightning protection method for high-altitude transmission lines according to claim 1, characterized in that, After determining whether to trip the transmission line based on the lightning strike risk coefficient, the method further includes: Generate lightning protection maintenance recommendations, the content of which is determined based on the lightning type and lightning strike risk factor. The lightning protection maintenance recommendations are associated with and stored along with the image data, location information, and time of the lightning strike, and then output.

10. A lightning protection system for high-altitude power transmission lines, characterized in that, For implementing the lightning protection method for high-altitude transmission lines as described in any one of claims 1 to 9, the system comprises: The acquisition module is used to acquire image data of the transmission line, as well as current and voltage data in the transmission line, according to a preset cycle. The judgment module is used to determine whether a lightning strike has occurred based on the image data. If so, it determines the lightning type based on the energy characteristics of different components in the current data. The lightning type includes induced lightning and direct lightning strike. The identification module is used to obtain the lightning risk coefficient based on the image data, current data, voltage data and lightning type, and on a pre-trained lightning risk identification model. The tripping module is used to determine whether to trip the transmission line based on the lightning strike risk coefficient.

Images