A method and system for identifying lightning disaster and analyzing grounding fault of a wind farm
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUANENG JIANGXI CLEAN ENERGY GENERATION CO LTD
- Filing Date
- 2026-05-27
- Publication Date
- 2026-07-21
AI Technical Summary
Existing methods for analyzing lightning hazards in wind farms fail to effectively consider the differences in lightning discharge characteristics among different wind farm types, fail to dynamically reflect the impact of blade movement on the electric field, and lack the fusion analysis of multi-source detection data in grounding fault diagnosis, resulting in inaccurate risk assessment.
By collecting historical lightning and environmental parameters, the differential strike-concentration coefficient and leader path correction factor are obtained. The electric field distortion coefficient is calculated by combining the real-time motion parameters of the wind turbine blades. Feature fingerprints are extracted from multi-source detection data and weighted fusion is performed to generate a lightning strike risk index.
It improves the sensitivity and reliability of lightning risk early warning, and can dynamically reflect the physical process of lightning strikes under different wind farm types and environmental conditions, accurately identifying lightning disasters and grounding faults.
Smart Images
Figure CN122432938A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of wind farm technology and relates to a method and system for identifying lightning disasters and analyzing grounding faults in wind farms. Background Technology
[0002] Wind turbines are vulnerable to lightning strikes due to their tall structure, rotating blades, and location in open areas. Therefore, it is necessary to accurately identify lightning disaster risks and diagnose grounding faults in a timely manner.
[0003] Existing methods for analyzing lightning hazards in wind farms have the following shortcomings: First, they fail to consider the differences in lightning discharge characteristics across different wind farm types. Environmental parameters such as soil resistivity, topographic relief, salt spray concentration, and dust charge vary significantly between mountainous, coastal, offshore, and Gobi desert areas, leading to variations in lightning current amplitude distribution, leader development paths, and strike-and-concentration probabilities. Existing lightning parameter models cannot provide differentiated strike-and-concentration coefficients and leader path correction factors for specific scenarios, reducing the scenario adaptability of risk assessments. Second, existing simulation methods are mostly based on static or discrete attitude electrostatic field calculations, failing to consider the dynamic distortion effect of continuous blade rotation on the spatial electric field. They also fail to incorporate real-time blade motion parameters (angular velocity, pitch angle, azimuth angle) into the calculation of the electric field distortion coefficient, resulting in dynamic lightning induced risk values that cannot accurately reflect the real-time impact of blade motion on the upward leader induction probability. Third, the failure of grounding devices is often caused by the coupling of multiple factors such as soil corrosion, equipment vibration fatigue, and lightning current impulse overload. Existing technologies mostly rely on a single detection parameter (such as grounding resistance) for judgment, lacking the fusion analysis of multi-source detection data. It is difficult to extract characteristic fingerprints of soil corrosion, vibration fatigue, and impulse overload, thus failing to obtain confidence vectors that reflect the probability of each failure mode, resulting in ambiguous fault location and inaccurate early warning information. Summary of the Invention
[0004] To address the shortcomings of existing technologies, the present invention aims to provide a method and system for identifying lightning disasters and analyzing grounding faults in wind farms.
[0005] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a method for identifying lightning disasters and analyzing grounding faults in wind farms, comprising the following steps: collecting historical lightning parameters and environmental parameters for the target wind farm type; obtaining a differentiated lightning strike concentration coefficient and a leader path correction factor based on the historical lightning parameters and environmental parameters; acquiring the initial background electric field of the target wind farm; correcting the initial background electric field based on the leader path correction factor to obtain the corrected background electric field; calculating the electric field distortion coefficient based on the real-time motion parameters of the target wind turbine blades; calculating the dynamic induced lightning risk value based on the corrected background electric field and the electric field distortion coefficient; collecting multi-source detection data of the grounding system; preprocessing the multi-source detection data to obtain feature fingerprints; obtaining a confidence vector based on the feature fingerprints; the feature fingerprints include soil corrosion feature fingerprints, equipment vibration fatigue feature fingerprints, and lightning current impulse overload feature fingerprints; weighted fusing the differentiated lightning strike concentration coefficient, the dynamic induced lightning risk value, and the confidence vector to obtain a lightning strike risk index; and outputting a lightning strike risk warning signal when the lightning strike risk index is greater than or equal to a preset threshold.
[0006] Furthermore, the differentiated hit coefficient is :
[0007]
[0008]
[0009]
[0010] in, This is the lightning concentration coefficient. As a soil environmental correction factor, Altitude correction factor This is an air humidity correction factor. For soil correction functions, For soil resistivity, This is the altitude correction function. The altitude of the target wind farm. This is a humidity correction function. This refers to the relative humidity of the air. This represents soil saturation.
[0011] Furthermore, the leader path correction factor is :
[0012]
[0013]
[0014] in, The polarity influence coefficient. The polar influence coefficient, The polarity influences the weighting coefficient. It is a positive leading probability. The slope influence weighting coefficient, The slope angle of the site. The weighting coefficients for soil conductivity characteristics. The electrical conductivity characteristic value of the site soil. The baseline soil conductivity characteristic value.
[0015] Furthermore, the real-time motion parameters include blade rotational angular velocity, propeller pitch angle, and azimuth angle; The electric field distortion coefficient is :
[0016] in, The influence coefficient of rotational speed. Angular velocity of rotation This is the azimuth influence coefficient. This is the azimuth angle.
[0017] Further, the step of obtaining the confidence vector based on the feature fingerprint includes: The soil corrosion fingerprint, equipment vibration fatigue fingerprint, and lightning current impact overload fingerprint are respectively input into a pre-trained Bayesian classifier or support vector machine model to output the soil corrosion probability. Equipment vibration fatigue probability Probability of lightning current impulse The confidence vector is ,and + + =1.
[0018] Furthermore, the lightning strike risk index is :
[0019] in, Weights for site lightning strike concentration characteristics. For differentiated hit-and-pull coefficients, For dynamic lightning trap risk weighting, This represents the dynamic risk value for lightning strikes. For grounding system fault confidence weights, For soil corrosion probability, This represents the probability of equipment vibration fatigue. This represents the probability of lightning current impact. .
[0020] Furthermore, the multi-source detection data includes grounding resistance detection data, soil resistivity monitoring data, unit equipment vibration monitoring data, lightning current waveform acquisition data, and grounding loop temperature data; the preprocessing includes data cleaning, outlier removal, time sequence alignment, and noise reduction; the feature fingerprints include soil corrosion feature fingerprints, equipment vibration fatigue feature fingerprints, and lightning current impulse overload feature fingerprints; the preprocessing of the multi-source detection data to obtain feature fingerprints includes: based on the preprocessed multi-source detection data, extracting soil corrosion rate and grounding impedance attenuation value to form soil corrosion feature fingerprints, extracting vibration amplitude, vibration frequency, and fatigue loss value to form equipment vibration fatigue feature fingerprints, and extracting lightning current peak value, impulse duration, and overload count to form lightning current impulse overload feature fingerprints.
[0021] Further, the step of correcting the initial background electric field based on the leader path correction factor to obtain the corrected background electric field includes: linearly correcting the initial background electric field with the leader path correction factor to obtain the corrected background electric field. ;
[0022] in, Let be the initial background electric field of the target wind farm. This is a correction factor for the leading path; The corrected background electric field is used to quantify the interference effects of topography and lightning polarity on the background electric field distribution of wind farms.
[0023] Furthermore, the target wind farm types include mountain type, coastal type, offshore type and Gobi type.
[0024] This invention also provides a wind farm lightning disaster identification and grounding fault analysis system, including an acquisition module for collecting historical lightning parameters and environmental parameters of the target wind farm type, and obtaining a differentiated lightning convergence coefficient and a leader path correction factor based on the historical lightning parameters and environmental parameters; a calculation module for acquiring the initial background electric field of the target wind farm, correcting the initial background electric field based on the leader path correction factor, obtaining the corrected background electric field, calculating the electric field distortion coefficient based on the real-time motion parameters of the target wind turbine blades, and calculating the dynamic induced lightning risk value based on the corrected background electric field and the electric field distortion coefficient; a processing module for collecting multi-source detection data of the grounding system, preprocessing the multi-source detection data to obtain feature fingerprints, and obtaining a confidence vector based on the feature fingerprints; the feature fingerprints include soil corrosion feature fingerprints, equipment vibration fatigue feature fingerprints, and lightning current impulse overload feature fingerprints; and a fusion module for weighted fusion of the differentiated lightning convergence coefficient, the dynamic induced lightning risk value, and the confidence vector to obtain a lightning strike risk index, and outputting a lightning strike risk warning signal when the lightning strike risk index is greater than or equal to a preset threshold.
[0025] Compared with the prior art, the present invention has the following beneficial technical effects: This invention provides a method for identifying lightning disasters and analyzing grounding faults in wind farms. It combines differentiated lightning convergence coefficients and leader path correction factors with historical lightning parameters and environmental parameters to more realistically reflect the physical processes of lightning strikes under different wind farm types and environmental conditions. The method uses the leader path correction factor to correct the initial background electric field and calculates the electric field distortion coefficient by combining real-time motion parameters of the wind turbine blades, dynamically acquiring induced lightning risk values and capturing the real-time impact of the wind turbine's motion state on the local electric field. Based on multi-source detection data of the grounding system, it extracts characteristic fingerprints such as soil corrosion, equipment vibration fatigue, and lightning current impulse overload, and generates a confidence vector to comprehensively quantify the health status of the grounding system. By weighted fusion of differentiated lightning convergence coefficients, dynamic induced lightning risk values, and the confidence vector, a lightning risk index is obtained, triggering an early warning signal. This overcomes the limitations of single-parameter assessment and improves the sensitivity and reliability of lightning risk early warning.
[0026] The target wind farms include mountain, coastal, offshore, and Gobi types. Based on the unique topography, climate, and lightning activity patterns of each type, historical lightning and environmental parameters are collected to obtain differentiated strike-concentration coefficients and leader path correction factors, thus better reflecting the actual operating conditions of wind farms.
[0027] Data from multiple sources, including grounding resistance, soil resistivity, unit vibration, lightning current waveform, and grounding loop temperature, were collected and processed through data cleaning, outlier removal, time alignment, and noise reduction to ensure data consistency. Soil corrosion fingerprints, equipment vibration fatigue fingerprints, and lightning current impulse overload fingerprints were extracted from the preprocessed data to transform the original physical quantities into quantitative indicators. Through these three types of fingerprints (corrosion, vibration, and overload), the degradation state under different environmental and electrical stresses was quantified, thereby improving the objectivity and reliability of lightning strike risk assessment. Attached Figure Description
[0028] Figure 1 This is a flowchart of a method for identifying lightning disasters and analyzing grounding faults in wind farms according to the present invention. Detailed Implementation
[0029] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0030] Example 1 This invention discloses a method for identifying lightning disasters and analyzing grounding faults in wind farms, comprising the following steps: collecting historical lightning parameters and environmental parameters for the target wind farm type; obtaining a differentiated lightning convergence coefficient and a leader path correction factor based on the historical lightning parameters and environmental parameters; acquiring the initial background electric field of the target wind farm; correcting the initial background electric field based on the leader path correction factor to obtain the corrected background electric field; calculating the electric field distortion coefficient based on the real-time motion parameters of the target wind turbine blades; calculating the dynamic lightning induced risk value based on the corrected background electric field and the electric field distortion coefficient; collecting multi-source detection data of the grounding system; preprocessing the multi-source detection data to obtain feature fingerprints; obtaining a confidence vector based on the feature fingerprints; the feature fingerprints include soil corrosion feature fingerprints, equipment vibration fatigue feature fingerprints, and lightning current impulse overload feature fingerprints; weighted fusing the differentiated lightning convergence coefficient, the dynamic lightning induced risk value, and the confidence vector to obtain a lightning strike risk index; when the lightning strike risk index is greater than or equal to a preset threshold, outputting a lightning strike risk warning signal, such as... Figure 1 As shown.
[0031] Specifically, the target wind farm type includes at least one of the following: mountainous, coastal, offshore, and Gobi desert. The type of wind farm is determined through geographic information system data, wind farm design documents, or on-site survey results. After type confirmation, historical lightning parameters and real-time environmental parameters corresponding to this type of wind farm are collected. Historical lightning parameters include lightning frequency, leader propagation patterns, and positive and negative polarity lightning distribution characteristics. Environmental parameters include soil resistivity, altitude, relative humidity, soil saturation, site slope angle, and soil conductivity characteristics. Based on the collected historical lightning parameters and environmental parameters, the differential knock-concentration coefficient and leader path correction factor are calculated.
[0032] By using multi-dimensional environmental factors for coupling correction, the lightning accumulation capacity of different sites is quantified, and the differentiated lightning accumulation coefficient is calculated by coupling the basic lightning accumulation coefficient with multiple environmental correction factors.
[0033]
[0034] in, This is the lightning concentration coefficient. As a soil environmental correction factor, Altitude correction factor This is the air humidity correction factor.
[0035]
[0036] For soil correction functions, Soil resistivity is a key indicator. Different soil resistivities correspond to significant differences in soil conductivity. The lower the resistivity, the better the soil conductivity and the stronger its lightning discharge capability.
[0037]
[0038] This is the altitude correction function. The altitude of the target wind farm is considered. In high-altitude areas, the air is thin and the insulation performance is weak.
[0039]
[0040] This is a humidity correction function. This refers to the relative humidity of the air. Soil saturation and air humidity directly affect the near-surface air conductivity and surface charge accumulation. Charges tend to accumulate more easily in humid environments, which can increase the risk of lightning strikes on the site.
[0041]
[0042]
[0043] The polarity influences the weighting coefficient. The positive leader probability directly affects the development direction and propagation intensity of the lightning leader, and the inherent error of conventional lightning path assessment is offset by probability deviation correction.
[0044]
[0045] The slope influence weighting coefficient, The slope angle of the site. The weighting coefficients for soil conductivity characteristics. The electrical conductivity characteristic value of the site soil. The baseline soil conductivity characteristic value.
[0046] Mountain wind farms have significant slope variations, which alter the near-surface electric field distribution. Coastal and offshore wind farms, on the other hand, have soils with high salinity and high conductivity, while Gobi desert wind farms have dry soil with high resistivity. These differences in soil conductivity under different scenarios directly interfere with the grounding path of lightning leaders. The polarity influence coefficient... and the influence coefficient of the earth pole The fusion correction yields a leading path correction factor that can quantify the interference of various complex scenarios on the lightning propagation path.
[0047] Based on the acquired leader path correction factor, the initial background electric field of the target wind farm is first collected using electric field monitoring equipment. This initial background electric field is the original electric field value without wind turbine interference or site correction, reflecting only the basic electric field state. Through linear correction, the leader path correction factor is used to iterate the initial background electric field, eliminating interference from topography and lightning polarity, restoring the true spatial electric field distribution of the wind farm, and quantifying the interference effect of topography and lightning polarity on the background electric field distribution of the wind farm. During operation, the wind turbine blades continuously rotate, directly altering the surrounding spatial electric field distribution and causing electric field distortion. Therefore, real-time motion parameters of the target wind turbine blades are collected, including blade rotation angular velocity, pitch angle, and azimuth angle. The electric field distortion coefficient is calculated based on these collected real-time motion parameters.
[0048] electric field distortion coefficient is :
[0049] in, The influence coefficient of rotational speed. Angular velocity of rotation This is the azimuth influence coefficient. This is the azimuth angle.
[0050] Based on the corrected background electric field, coupled calculations are performed using the electric field distortion coefficient to obtain the dynamic lightning induced risk value of the target wind farm.
[0051] Multi-source detection data of the grounding system are collected, including grounding resistance detection data, soil resistivity monitoring data, unit equipment vibration monitoring data, lightning current waveform acquisition data, and grounding loop temperature data; preprocessing includes data cleaning, outlier removal, time sequence alignment, and noise reduction; the characteristic fingerprints include equipment vibration fatigue characteristic fingerprints and lightning current impulse overload characteristic fingerprints. Preprocessing of multi-source detection data to obtain feature fingerprints includes: extracting soil corrosion rate and grounding impedance attenuation value to form soil corrosion feature fingerprints based on the preprocessed multi-source detection data; extracting vibration amplitude, vibration frequency, and fatigue loss value to form equipment vibration fatigue feature fingerprints; and extracting lightning current peak value, impact duration, and overload number to form lightning current impact overload feature fingerprints.
[0052] Soil corrosion fingerprints, equipment vibration fatigue fingerprints, and lightning current impulse overload fingerprints correspond to soil corrosion faults, equipment vibration fatigue faults, and lightning current impulse overload faults in the grounding system, respectively. These fingerprints are then input into a pre-trained Bayesian classifier or support vector machine model to output the soil corrosion probability. Equipment vibration fatigue probability Probability of lightning current impulse probability of soil corrosion Equipment vibration fatigue probability Probability of lightning current impulse By applying constraints, we obtain the confidence vector.
[0053] The lightning strike risk index is obtained by weighted fusion of the differentiated strike-and-convergence coefficient, dynamic induced lightning risk value, and confidence vector. ;
[0054] in, Weights for site lightning strike concentration characteristics. For differentiated hit-and-pull coefficients, For dynamic lightning trap risk weighting, This represents the dynamic risk value for lightning strikes. For grounding system fault confidence weights, For soil corrosion probability, This represents the probability of equipment vibration fatigue. This represents the probability of lightning current impact. .
[0055] The lightning risk index is compared with a preset threshold. When the lightning risk index is greater than or equal to the preset threshold, the target wind farm is at risk of lightning disaster and grounding fault. A lightning risk warning signal is output, and maintenance and investigation are carried out. When the lightning risk index is less than the preset threshold, routine monitoring continues.
[0056] Example 2 This invention discloses a wind farm lightning disaster identification and grounding fault analysis system, comprising: an acquisition module for collecting historical lightning parameters and environmental parameters of the target wind farm type, and obtaining a differentiated lightning convergence coefficient and a leader path correction factor based on the historical lightning parameters and environmental parameters; a calculation module for acquiring the initial background electric field of the target wind farm, correcting the initial background electric field based on the leader path correction factor, obtaining the corrected background electric field, calculating the electric field distortion coefficient based on the real-time motion parameters of the target wind turbine blades, and calculating the dynamic induced lightning risk value based on the corrected background electric field and the electric field distortion coefficient; a processing module for collecting multi-source detection data of the grounding system, preprocessing the multi-source detection data to obtain feature fingerprints, and obtaining a confidence vector based on the feature fingerprints; the feature fingerprints include soil corrosion feature fingerprints, equipment vibration fatigue feature fingerprints, and lightning current impulse overload feature fingerprints; and a fusion module for weighted fusion of the differentiated lightning convergence coefficient, the dynamic induced lightning risk value, and the confidence vector to obtain a lightning strike risk index, and outputting a lightning strike risk warning signal when the lightning strike risk index reaches a preset threshold.
[0057] It should be noted that the wind farm lightning disaster identification and grounding fault analysis system provided by the present invention can implement the same method steps as the above method, so it will not be described again.
[0058] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
Claims
1. A method for identifying lightning hazards and analyzing grounding faults in wind farms, characterized in that, Includes the following steps: Historical lightning and environmental parameters for the target wind farm type are collected, and the differential knock-aggregation coefficient and leader path correction factor are obtained based on the historical lightning and environmental parameters. The initial background electric field of the target wind farm is obtained, and the initial background electric field is corrected based on the leader path correction factor to obtain the corrected background electric field. The electric field distortion coefficient is calculated based on the real-time motion parameters of the target wind turbine blades, and the dynamic lightning induced risk value is calculated based on the corrected background electric field and the electric field distortion coefficient. Multi-source detection data of the grounding system is collected, and the multi-source detection data is preprocessed to obtain feature fingerprints. A confidence vector is obtained based on the feature fingerprints. The feature fingerprints include soil corrosion feature fingerprints, equipment vibration fatigue feature fingerprints, and lightning current impulse overload feature fingerprints. The lightning strike risk index is obtained by weighted fusion of the differentiated lightning strike coefficient, dynamic lightning induced risk value and confidence vector. When the lightning strike risk index is greater than or equal to a preset threshold, a lightning strike risk warning signal is output.
2. The method for identifying lightning disasters and analyzing grounding faults in wind farms according to claim 1, characterized in that: The differential repulsion coefficient is : in, This is the lightning concentration coefficient. As a soil environmental correction factor, Altitude correction factor This is an air humidity correction factor. For soil correction functions, For soil resistivity, This is the altitude correction function. The altitude of the target wind farm. This is a humidity correction function. This refers to the relative humidity of the air. This represents soil saturation.
3. The method for identifying lightning disasters and analyzing grounding faults in wind farms according to claim 2, characterized in that: The leader path correction factor is : in, The polarity influence coefficient. The polar influence coefficient, The polarity influences the weighting coefficient. It is a positive leading probability. The slope influence weighting coefficient, The slope angle of the site. The weighting coefficients for soil conductivity characteristics. The electrical conductivity characteristic value of the site soil. The baseline soil conductivity characteristic value.
4. The method for identifying lightning disasters and analyzing grounding faults in wind farms according to claim 1, characterized in that: The real-time motion parameters include blade rotational angular velocity, propeller pitch angle, and azimuth angle; The electric field distortion coefficient is : in, The influence coefficient of rotational speed. Angular velocity of rotation This is the azimuth influence coefficient. It is the azimuth angle.
5. The method for identifying lightning disasters and analyzing grounding faults in wind farms according to claim 1, characterized in that: The process of obtaining the confidence vector based on feature fingerprints includes: The soil corrosion fingerprint, equipment vibration fatigue fingerprint, and lightning current impact overload fingerprint are respectively input into a pre-trained Bayesian classifier or support vector machine model to output the soil corrosion probability. Equipment vibration fatigue probability Probability of lightning current impulse ; The confidence vector is ,and + + =1.
6. The method for identifying lightning disasters and analyzing grounding faults in wind farms according to claim 5, characterized in that: The lightning strike risk index is: : in, Weights for site lightning strike concentration characteristics. For differentiated hit-and-pull coefficients, For dynamic lightning trap risk weighting, This represents the dynamic risk value for lightning strikes. For grounding system fault confidence weights, For soil corrosion probability, This represents the probability of equipment vibration fatigue. This represents the probability of lightning current impact. .
7. The method for identifying lightning disasters and analyzing grounding faults in wind farms according to claim 1, characterized in that: The multi-source detection data includes grounding resistance detection data, soil resistivity monitoring data, unit equipment vibration monitoring data, lightning current waveform acquisition data, and grounding circuit temperature data. The preprocessing includes data cleaning, outlier removal, time-series alignment, and noise reduction. The characteristic fingerprints include soil corrosion characteristic fingerprints, equipment vibration fatigue characteristic fingerprints, and lightning current impulse overload characteristic fingerprints. Preprocessing the multi-source detection data to obtain feature fingerprints includes: extracting soil corrosion rate and grounding impedance attenuation value to form soil corrosion feature fingerprints based on the preprocessed multi-source detection data; extracting vibration amplitude, vibration frequency, and fatigue loss value to form equipment vibration fatigue feature fingerprints; and extracting lightning current peak value, impact duration, and overload number to form lightning current impact overload feature fingerprints.
8. The method for identifying lightning disasters and analyzing grounding faults in wind farms according to claim 1, characterized in that: The step of correcting the initial background electric field based on the leader path correction factor to obtain the corrected background electric field includes: The initial background electric field is linearly corrected using the leader path correction factor to obtain the corrected background electric field. ; in, Let be the initial background electric field of the target wind farm. This is a correction factor for the leading path; The corrected background electric field is used to quantify the interference effects of topography and lightning polarity on the background electric field distribution of wind farms.
9. The method for identifying lightning disasters and analyzing grounding faults in wind farms according to claim 1, characterized in that: The target wind farms include mountain type, coastal type, offshore type and Gobi type.
10. A system for identifying lightning hazards and analyzing grounding faults in wind farms, characterized in that, include: Acquisition module: used to collect historical lightning parameters and environmental parameters for the target wind farm type, and obtain the differentiated clustering coefficient and leader path correction factor based on the historical lightning parameters and environmental parameters; Calculation module: used to obtain the initial background electric field of the target wind farm, correct the initial background electric field based on the leader path correction factor, obtain the corrected background electric field, calculate the electric field distortion coefficient according to the real-time motion parameters of the target wind turbine blades, and calculate the dynamic lightning induced risk value according to the corrected background electric field and the electric field distortion coefficient. Processing module: used to collect multi-source detection data of the grounding system, preprocess the multi-source detection data to obtain feature fingerprints, and obtain confidence vectors based on feature fingerprints; the feature fingerprints include soil corrosion feature fingerprints, equipment vibration fatigue feature fingerprints, and lightning current impulse overload feature fingerprints; The fusion module is used to perform weighted fusion of the differentiated lightning strike coefficient, dynamic lightning induced risk value and confidence vector to obtain a lightning strike risk index. When the lightning strike risk index is greater than or equal to a preset threshold, a lightning strike risk warning signal is output.