An intelligent monitoring system for lightning stroke fracture risk of a lightning rod

Abstract

The present application relates to lightning arrester intelligent monitoring technical field, specifically to a kind of lightning arrester lightning fracture risk intelligent monitoring system, system includes collecting lightning arrester tension electric field displacement disturbance data and rearranging section according to time, screening multi-source disturbance node and locating paragraph position, according to node number matching unmanned aerial vehicle continuous navigation section, extract corresponding image analysis brightness gradient and texture change, combine image and paragraph number cross relationship output lightning fracture risk section number.The present application is through the time series reorganization of tension, electric field, displacement disturbance data and paragraph number, constructs continuous disturbance structure, selects trigger node by the relationship of multiple physical quantity change, matches planning path according to node paragraph, compresses the range of inspection, identifies fracture suspect point by image gradient and texture change, combines the cross relationship of coordinate and outputs risk number section, enhances the coverage and traceability of risk identification.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring technology for lightning protection wires, and in particular to an intelligent monitoring system for the risk of lightning protection wire breakage after a lightning strike. Background Technology

[0002] The field of intelligent monitoring technology for lightning protection conductors involves the real-time perception and assessment of the operational status of lightning protection conductors in overhead transmission lines. This primarily includes core aspects such as lightning strike location, current measurement, structural stress monitoring, material aging diagnosis, and environmental information collection. Utilizing sensor networks, communication technologies, and data analysis methods, it systematically achieves continuous tracking and intelligent judgment of the health status of lightning protection conductors, thereby enhancing the safety and reliability of the power system under extreme weather conditions such as lightning strikes. It comprehensively utilizes multiple technologies, including sensing and measurement, wireless transmission, embedded processing, and data fusion, to construct a status monitoring system with early warning capabilities, serving the operation, maintenance, and risk management of transmission lines.

[0003] Among them, the traditional intelligent monitoring system for the risk of lightning conductor breakage after lightning strike refers to a technical solution used to monitor the potential breakage risk of lightning conductors after a lightning strike. It mainly targets safety hazards such as partial melting, internal damage, or micro-cracks in lightning conductors caused by lightning strikes. Traditional monitoring of the risk of lightning conductor breakage after lightning strikes usually uses regular manual inspections, on-site contact measurements, or infrared thermal imaging to obtain signs of damage on the surface of the lightning conductor, and then makes judgments based on past operating experience. Some solutions also use lightning strike counters to record the occurrence of lightning strikes and combine them with lightning current waveform analysis to make inferences. However, the overall approach still has certain limitations in terms of the continuity of information collection, the timeliness of data, and the accuracy of monitoring.

[0004] Existing technologies rely on manual inspections and contact-based equipment, resulting in limited information coverage and monitoring delays and omissions. Lightning strike counting only provides event records and cannot capture the structural response details of lightning protection wires. Thermal imaging methods are significantly affected by lighting interference, lack the ability to identify minute surface cracks, and lack a fusion path between image and structural data. Path planning does not consider numbered segments and spatial continuity, leading to repeated inspections or omissions. Fracture determination lacks a cross-processing mechanism for disturbance and image information, making it difficult to form a complete risk identification chain. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and propose an intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike, the system comprising, The stress disturbance extraction module collects tension, electric field, and displacement data of the lightning conductor tension arm, equipotential bonding ring, and clamp joint. It rearranges the data columns by time, removes interrupted data segments, assigns the numbers to segments, performs segmentation operations on the segments, and obtains multi-point disturbances of the lightning conductor to form segment groups. The high-sensitivity node screening module, based on the tension slope, electric field disturbance range and displacement change interval in the segment group composed of multi-point disturbances of the lightning protection line, screens out nodes with three types of disturbances, aligns the node numbers with the segment positions, extracts the number information, and obtains a list of multi-source triggering nodes of the lightning protection line. The flight path planning module searches for the corresponding segments in the UAV path based on the number and segment coordinates in the list of multi-source triggering nodes of the lightning protection line, filters out paths that do not cross numbered segments and have short spatial intervals, extracts continuous segments with more stable path directions as flight segments, and obtains the UAV lightning strike inspection path number group. The image fracture extraction module calls the image number in the UAV lightning strike inspection path number group, compares the brightness gradient, extracts the edge direction, identifies the texture change direction, extracts the image number corresponding to the fracture area, and obtains the image annotation information set of the lightning protection wire fracture location. The breakage risk assessment module compares the image annotation information set of the breakage location of the lightning conductor with the numbered positions in the list of multi-source triggering nodes of the lightning conductor, checks the intersection range of the image coordinates and the segment coordinates, extracts the adjacent numbered intervals, and outputs the lightning conductor breakage risk monitoring results.

[0007] As a further aspect of the present invention, the multi-point disturbance segment group of the lightning protection wire includes the temporal distribution of disturbance occurrence, the multi-node association structure within the segment, the data continuity mapping relationship, the disturbance category combination label, and the segment boundary index. The multi-source triggering node list of the lightning protection wire includes the tension change node number, the electric field interference critical point, the displacement change corresponding number, the segment number to which the node belongs, and the node triggering source type. The UAV lightning strike inspection path number group includes the path number list, the effective continuous path segment, the spatially adjacent path segment, the turning optimized path segment, and the mapping relationship between the path number and the segment number. The lightning protection wire fracture location image annotation information set includes the fracture image number, brightness edge features, texture density change area, fracture contour selection result, and suspected fracture location label. The lightning protection wire fracture risk monitoring result includes the high-risk segment number, the image and coordinate cross-comparison interval, the fracture trend intensity level, the risk segment boundary range, and the comprehensive judgment label.

[0008] As a further aspect of the present invention, the stress disturbance extraction module includes: The tension disturbance identification submodule collects tension disturbance operation data of the lightning conductor tension arm, equipotential bonding ring, and clamp joint. It adjusts the tension data order according to the time field, removes interrupted numbered segments, determines the tension change location based on the order change trend, selects the fluctuation direction and compares it with the peak interval to obtain the tension disturbance change amplitude. The electric field displacement synchronization submodule collects synchronous electric field and displacement disturbance operation data according to the amplitude of the tension disturbance change, aligns the electric field and displacement data according to the time field of the tension data, extracts the corresponding content at a uniform interval, compares the difference trends between tension, electric field and displacement, and obtains the range of multi-source disturbance differences. The segment interval processing submodule extracts the disturbance content within the assigned number segment based on the range of multi-source disturbance differences, filters out continuous parts with continuously changing difference trends, groups and processes them according to the numbering information, locates multiple segments with disturbance continuity on the lightning protection line, and obtains multi-point disturbance interval groups for the lightning protection line.

[0009] As a further aspect of the present invention, the high-sensitivity node screening module includes: The disturbance behavior extraction submodule extracts the tension change trend, electric field disturbance length and displacement change time interval before and after each segment based on the multi-point disturbance interval group of the lightning protection wire. It performs judgment operations on the three types of disturbance content in sequence, filters out nodes that simultaneously exhibit tension change, electric field extension and displacement change, and obtains a multi-disturbance joint identification node sequence. The node positioning and association submodule calls the multi-perturbation joint identifier node sequence, matches the start and end time range of the corresponding paragraph according to the time information of each node, obtains the node occurrence position within the paragraph, sorts the node numbers within the same interval, aligns the positions according to the paragraph order, and obtains the paragraph position sequence corresponding to the number. The trigger point aggregation submodule assigns a unified number to the disturbance nodes appearing in the corresponding paragraph under each number according to the paragraph position sequence of the number, filters out the numbers whose node distribution within the paragraph is greater than the node concentration benchmark, obtains the node set with prominent lightning protection line triggering characteristics, and obtains the list of multi-source triggering nodes of the lightning protection line.

[0010] As a further aspect of the present invention, the flight path planning module includes: The node path mapping submodule, based on the list of multi-source triggering nodes of the lightning protection line, obtains the node number and the corresponding paragraph position, retrieves the corresponding coordinate segment in the UAV flight path, performs a one-to-one correspondence judgment on the node number and path coordinate according to the start and end order of the paragraph, excludes path content that crosses different number segments, and obtains the set of path segments corresponding to the node. The continuous flight segment filtering submodule calls the path segment set corresponding to the node, detects the spatial interval between adjacent path segments, judges the turning situation according to the path point order, filters the turning frequency and path point continuity, retains the path content with fewer turning occurrences and continuous point order, and obtains a continuous flight path segment set. The patrol path confirmation submodule performs an alignment operation on the numbering of each path segment according to the set of continuous flight path segments, extracts the path number order in the corresponding lightning protection line segment, completes the order verification between path numbers, and obtains the UAV lightning strike patrol path number group.

[0011] As a further aspect of the present invention, the image fracture extraction module includes: The image brightness comparison submodule calls each numbered image in the UAV lightning strike patrol path numbering group, performs brightness gradient difference comparison on adjacent pixel blocks in the image, detects the number of boundary change directions, filters out image numbers with concentrated edge directions, and obtains the image edge direction distribution trend. The texture orientation recognition submodule identifies regions in the image where the texture distribution density changes with orientation based on the image edge orientation distribution trend, records the orientation information corresponding to density abrupt changes, performs overlap judgment on the orientation change line and the image edge trend line, and obtains the distribution of overlapping texture abrupt change segments. The suspected location extraction submodule extracts the corresponding numbered regions in the image based on the distribution of overlapping texture mutation segments, filters out locations with concentrated edges and abrupt changes in texture direction, judges the contour aggregation in the numbered image, records the location numbers with continuous boundary trends in the marked area, and obtains the image annotation information set of lightning protection wire breakage locations.

[0012] As a further aspect of the present invention, the fracture risk determination module includes: The coordinate range verification submodule, based on the image annotation information set of the lightning protection wire fracture location, compares the node number and segment position in the list of multi-source triggering nodes of the lightning protection wire to obtain the intersection range of the image coordinates and segment coordinates, performs a comparison operation, screens the overlapping area of ​​the image coordinates and segment coordinates, and obtains the coordinate intersection matching area. The adjacent interval extraction submodule extracts the coordinate intervals within adjacent numbered segments based on the coordinate intersection matching area, divides the coordinate range of adjacent areas according to the segment number information, filters the key risk locations within adjacent numbered intervals, and obtains the adjacent risk interval set. The risk monitoring result output submodule extracts all suspected fracture risk nodes based on the adjacent risk interval set, monitors the area range, marks the existing fracture risk points, outputs fracture risk assessment information, and obtains the lightning protection wire fracture risk monitoring results.

[0013] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by reorganizing and classifying the temporal sequence of lightning conductor tension, electric field, and displacement disturbance data into segments, a continuous disturbance structure is formed. Trigger nodes with associated characteristics are screened through the temporal relationship of multiple physical quantity changes. The inspection path is constrained by the correspondence between node numbers and spatial segments to reduce irrelevant segments. Suspected fracture areas are identified by pixel gradient and texture direction changes. The risk segment number is given by combining the intersection of image coordinates and disturbance segments, thereby improving the completeness and traceability of fracture risk identification after lightning strike. Attached Figure Description

[0014] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a flowchart illustrating the acquisition process of the stress disturbance extraction module of the present invention. Figure 3 This is a flowchart illustrating the acquisition process of the high-sensitivity node screening module of the present invention. Figure 4 This is a flowchart illustrating the acquisition process of the flight path planning module of the present invention. Figure 5 This is a flowchart illustrating the acquisition process of the image fracture extraction module of the present invention. Figure 6 This is a flowchart illustrating the acquisition process of the fracture risk assessment module of the present invention. Detailed Implementation

[0015] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0016] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0017] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0018] Please see Figure 1 This invention provides a technical solution: an intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike, the system comprising: The stress disturbance extraction module collects tension, electric field, and displacement disturbance data of the lightning conductor tension arm, equipotential bonding ring, and clamp joint. It rearranges the data columns according to time, removes interrupted data segments, assigns the numbers to segment intervals, performs partitioning processing on the segments, and obtains multi-point disturbances of the lightning conductor to form segment groups. The high-sensitivity node screening module, based on the slope of tension change, electric field interference length and displacement abrupt change before and after the occurrence of each segment in the segment group of the multi-point disturbance of the lightning protection line, sequentially screens out nodes with three types of disturbance phenomena, locates the segment interval where the node is located, aligns the node number with the segment position, and obtains the list of multi-source triggering nodes of the lightning protection line. The flight path planning module retrieves the corresponding coordinate segments in the UAV path based on the node number and corresponding segment position in the list of multi-source triggering nodes of the lightning protection line, filters out paths that do not cross number segments and have short spatial intervals, extracts the path segments with the fewest turning times and continuous points and classifies them as valid flight segments, and obtains the UAV lightning strike inspection path number group. The image fracture extraction module calls up each numbered image in the UAV lightning strike patrol path numbering group, compares the pixel block brightness gradient, extracts the edge direction quantity, identifies the texture density change direction and superimposes the contour range, extracts the suspected fracture image range number, and obtains the lightning protection wire fracture location image annotation information set. The fracture risk assessment module compares the image annotation information set of the lightning conductor fracture location with the numbered positions in the list of multi-source triggering nodes of the lightning conductor, checks the intersection range of image coordinates and segment coordinates, extracts adjacent numbered intervals, and outputs the lightning conductor fracture risk monitoring results.

[0019] The lightning protection wire multi-point disturbance segment group includes the temporal distribution of disturbance occurrence, the multi-node association structure within the segment, the data continuity mapping relationship, disturbance category combination labels, and segment boundary index. The lightning protection wire multi-source trigger node list includes tension change node number, electric field interference critical point, displacement change corresponding number, node to which segment number belongs, and node trigger source type. The UAV lightning strike inspection path number group includes a path number list, effective continuous path segments, spatially adjacent path segments, turning optimized path segments, and the mapping relationship between path number and segment number. The lightning protection wire fracture location image annotation information set includes fracture image number, brightness edge features, texture density change area, fracture contour selection result, and suspected fracture location label. The lightning protection wire fracture risk monitoring results include high-risk segment number, image and coordinate cross-comparison interval, fracture trend intensity level, risk segment boundary range, and comprehensive judgment label.

[0020] Please see Figure 2 The stress disturbance extraction module includes: The tension disturbance identification submodule collects tension disturbance operation data of the lightning conductor tension arm, equipotential bonding ring, and clamp joint. It adjusts the tension data order according to the time field, removes interrupted numbered segments, determines the tension change location based on the order change trend, selects the fluctuation direction and compares it with the peak interval to obtain the tension disturbance change amplitude. A stable communication link was established with the lightning protection wire monitoring terminal. The system was configured to poll and collect data from the field sensors at a high frequency of 50Hz, ensuring data acquisition every 20 milliseconds, sufficient to capture the details of the mechanical response at the moment of lightning strike. The data collected included strain gauge sensors deployed on the tension arm, pressure sensors at the equipotential bonding ring, and displacement sensors at the clamp joints. All acquired raw data streams were timestamped to milliseconds and merged to generate a raw tension time series. The system performed a rigorous time continuity check on this series, calculating the time difference between adjacent data points one by one. A 200ms time continuity threshold was applied to remove invalid data packets caused by network congestion, and fragmented data segments shorter than 5 seconds were automatically deleted, retaining only long-term continuous data with analytical value. Within the defined continuous data segments, the system used a sliding window difference technique with a window length of 10 to monitor the dynamic changes in tension values. The real-time calculated window mean change was compared with a preset 50N tension mutation threshold, which was set based on the experimentally measured differences between wind vibration and lightning strike loads. When the change in the monitored value exceeds this limit, the system locks that moment as the location of the tension mutation and searches for wave peaks before and after the mutation point. By comparing the wave peak value with the stable baseline value before the mutation, the amplitude of the tension disturbance mutation is directly obtained.

[0021] The electric field displacement synchronization submodule collects synchronous electric field and displacement disturbance operation data based on the amplitude of tension disturbance change, aligns the electric field and displacement data according to the time field of tension data, extracts the corresponding content at a uniform interval, compares the difference trends between tension, electric field and displacement, and obtains the range of multi-source disturbance differences. After acquiring the amplitude of the tension disturbance abrupt change and its exact occurrence time, the electric field-displacement synchronization submodule immediately initiates the spatiotemporal alignment procedure for multi-source data. The system constructs a 500ms time synchronization window centered on the moment of the tension abrupt change. This window width fully considers clock synchronization errors between different types of sensors and the inherent latency of data processing, ensuring complete capture of related events. Based on this time window, the system accurately retrieves the electric field intensity data and displacement monitoring data from a massive, parallel-stored historical database. Addressing the issue of non-uniform sampling frequencies from different sensors, it performs linear interpolation resampling based on the tension timestamp, uniformly mapping the electric field and displacement data onto the same time axis. After data alignment, the system extracts the three sets of data at uniform 10ms intervals and performs normalization to eliminate dimensional influences, ensuring comparability of different physical quantities. The rates of change of tension, electric field, and displacement within the synchronization window are calculated separately, with a focus on comparing the degree of difference between the tension rate of change and the rates of change of electric field and displacement. If the calculated difference value exceeds the heterogeneous disturbance characteristic boundary of 0.3 set based on historical lightning strike data statistics, the system determines that there is a significant inconsistency in the trend between the two, which usually means that a compound disturbance has occurred. The system will record in detail the specific time period during which this difference trend continues, thereby accurately obtaining the range of multi-source disturbance differences.

[0022] The segment interval processing submodule extracts the disturbance content within the assigned number segment based on the range of differences in multi-source disturbances, filters out continuous parts with continuously changing difference trends, groups and processes them according to the assigned number information, locates multiple segments with continuous disturbances on the lightning protection line, and obtains multi-point disturbance interval groups for the lightning protection line.

[0023] Upon receiving time information on the range of differences in multi-source disturbances, the system immediately loads the entire line's tower number database and span time index table for the lightning protection line. Through high-precision timestamp comparison, the system accurately maps the time of each disturbance occurrence to a specific assigned number segment, initially pinpointing the physical span where the disturbance occurred. Based on this, the system further analyzes the disturbance details within the assigned segment, screening for continuous portions with continuously changing difference trends. A 50ms continuity interruption threshold is used to handle minor fluctuations, treating records with intervals exceeding this threshold as independent event segments for segmentation. These independent segments are logically grouped according to their assigned number information, with a focus on checking the spatial logical continuity of segments within the same group. Based on the stress wave propagation speed of approximately 5000m / s in the steel strand and a certain buffer margin, the system verifies the spatiotemporal relationship of each segment within a group. Multiple segments that conform to the wave speed propagation logic in both time and space are considered responses to the same disturbance event at different locations and are merged. The specific start and end positions of the disturbance event on the lightning protection line are clearly defined, thus establishing the multi-point disturbance interval group for the lightning protection line. Please see Figure 3The high-sensitivity node screening module includes: The disturbance behavior extraction submodule extracts the tension change trend, electric field disturbance length and displacement change time interval before and after each segment based on the multi-point disturbance interval group of the lightning protection wire. It performs judgment operations on the three types of disturbance content in sequence, screens out the nodes that simultaneously exhibit tension change, electric field extension and displacement change, and obtains the multi-disturbance joint identification node sequence. The system acquires multiple disturbance interval groups for the lightning protection wire. For each interval unit, three key characteristic parameters are extracted: tension change trend, electric field disturbance length, and displacement abrupt change time interval. To ensure the accuracy of the screening results, the system executes a strict joint screening logic: First, the tension change trend must exhibit a specific single-pulse pattern, and its rising edge slope must be greater than 500 N / s, a significant characteristic distinguishing lightning strikes from ordinary mechanical vibrations. Second, the extracted electric field disturbance length must exceed a threshold of 1.5 m, a value derived from experimental measurements based on the physical characteristics of the lightning discharge channel and the radius of ground electric field distortion. The time interval before and after the displacement abrupt change must be less than 50 ms, an indicator set based on cable modal analysis to ensure that the initial impact response is captured rather than subsequent aftershocks. The system performs a logical AND operation on the above three conditions; that is, only when a node simultaneously meets the three conditions of high tension steepness, wide electric field coverage, and fast displacement response will it be determined as a valid disturbance node by the system. Once the determination is successful, a unique joint identifier ID will be assigned to the node, and all nodes that meet the conditions will be arranged in order to obtain the multi-perturbation joint identifier node sequence.

[0024] The node location association submodule calls the multi-perturbation joint identifier node sequence, matches the start and end time range of the corresponding paragraph according to the time information of each node, obtains the node occurrence position within the paragraph, sorts the node numbers within the same interval, aligns the positions according to the paragraph order, and obtains the paragraph position sequence corresponding to the number. The system retrieves the node occurrence times recorded in the multi-disturbance joint identifier node sequence and performs a detailed match with the operation schedule of each section of the lightning protection line. The operation schedule records the precise start and end times of monitoring for each section of the line. By searching for section records that meet the time range, the system can unambiguously determine the section ID to which each disturbance node belongs. After clarifying the section affiliation, the system further obtains the relative position coordinates of all nodes within that section. To facilitate subsequent path planning, the system uses the starting tower of each section as the spatial origin and sorts all node positions within the section in ascending order, constructing an ordered node distribution map within the section. Based on this, the system cascades and aligns the sorting results scattered within each section according to the physical connection order represented by the section ID. This process connects the originally isolated section information into a complete line topology structure, establishing a complete mapping relationship containing section identity information and node physical position sorting. A sequence of section positions corresponding to numbered nodes is generated. The trigger point aggregation submodule assigns a unified number to the disturbance nodes appearing in the corresponding paragraph under each number, based on the paragraph position sequence corresponding to the number. It then filters out the numbers whose node distribution within the paragraph is greater than the node concentration benchmark, thus obtaining a set of nodes with prominent lightning protection line triggering characteristics and acquiring a list of multi-source triggering nodes for the lightning protection line.

[0025] Based on the sequence of segment positions corresponding to the number, the system automatically counts the number of valid nodes under each number, i.e., each physical segment. To distinguish between random wind-induced interference and concentrated lightning strikes, the system introduces a node concentration benchmark of 3. This benchmark is derived from in-depth statistical analysis of historical lightning fault databases and non-lightning interference events, reflecting the characteristic of multiple densely triggered points generated by lightning strikes within the strike span. The system compares the number of nodes counted for each segment with this benchmark. When the number of nodes in a segment is greater than 3, the system determines that the node distribution in that segment has significant lightning strike characteristics and belongs to a high-risk area with prominent lightning strike features. All nodes in these high-risk segments are extracted, and their detailed attributes, including precise location, occurrence time, and intensity information, are packaged and processed uniformly. After this collection and filtering process, a list of multi-source triggered nodes for the lightning protection line is obtained. Please see Figure 4 The flight path planning module includes: The node path mapping submodule, based on the list of multi-source trigger nodes of the lightning protection line, obtains the node number and the corresponding paragraph position, retrieves the corresponding coordinate segment in the UAV flight path, performs a one-to-one correspondence judgment on the node number and path coordinate according to the start and end order of the paragraph, excludes path content that crosses different number segments, and obtains the set of path segments corresponding to the node. The system retrieves the precise geographic coordinates of all nodes in the multi-source trigger node list of the lightning protection line, including longitude, latitude, and altitude. It then iterates through a pre-defined UAV full-line inspection path library, which contains discrete path points covering the entire line. The system calculates the Euclidean distance between the coordinates of each node in the list and each point in the path library, applying a 30m mapping radius as a filtering criterion; any path point with a distance less than this radius is considered a valid corresponding point for that node. Simultaneously, to prevent path confusion, the system delineates polygonal electronic fences for each segment of the lightning protection line in three-dimensional space based on the coordinates of the starting and ending towers. The system performs rigorous fence verification on the filtered path points, checking whether they are within the fence range of the same numbered segment. Path content that spans different numbered segments, causing confusion in the inspection logic, is discarded. Only path point sequences whose starting and ending points are strictly within the fence range of the same segment are retained, thus obtaining the path segment set corresponding to each node.

[0026] The continuous flight segment filtering submodule calls the path segment set corresponding to the node, detects the spatial interval between adjacent path segments, judges the turning situation based on the path point order, filters the turning frequency and path point continuity, retains the path content with fewer turning occurrences and continuous point order, and obtains the continuous flight path segment set. The system retrieves the path segment set corresponding to each node and performs detailed detection on the spatial intervals between adjacent path points within each segment. If the single-step distance between adjacent points exceeds the UAV's maximum flight step length of 10m, the system determines that the path has a break or discontinuity and removes it to ensure flight safety. Based on the arrangement order of the path points, the system calculates the trajectory vector and derives the turning angle between adjacent vectors. The system sets a turning threshold of 30 degrees, which is determined based on the maximum angular velocity limit of the UAV gimbal when maintaining stable shooting. Any position with a turning angle exceeding 30 degrees is marked as a sharp turn point. The system counts the frequency of sharp turns within each path segment and executes filtering logic, retaining only those paths with no more than two sharp turns and whose point order is strictly continuous in space. This strategy effectively filters out inferior paths that, while covering the target, have overly tortuous flight trajectories and are not conducive to high-quality imaging. After this series of filtering and optimization, a continuous flight path segment set is obtained.

[0027] The patrol path confirmation submodule performs an alignment operation on the numbering of each path segment based on the set of continuous flight path segments, extracts the path number order in the corresponding lightning protection line segment, completes the order verification between path numbers, and obtains the UAV lightning strike patrol path number group.

[0028] The system extracts the lightning protection line segment number corresponding to each segment in the continuous flight path segment set, and aggregates multiple scattered path segments belonging to the same physical segment. These segments are then sorted and aligned according to their actual mileage positions on the lightning protection line. During this process, the system performs rigorous sequential logic checks, focusing on areas with overlapping mileage. For example, if some segments are detected to have overlapping coverage areas, the system merges these overlapping parts to generate a single continuous segment with wider coverage, while maintaining the original reasonable spacing between individual segments. After merging and sorting, the system generates a list of patrol waypoints in ascending order of mileage. To facilitate task management and scheduling, the system assigns a unique task number to each generated waypoint list, which includes the segment ID and sequence information. The optimized UAV lightning strike patrol path number set is then output.

[0029] Please see Figure 5 The image fracture extraction module includes: The image brightness comparison submodule calls each numbered image in the UAV lightning strike patrol path numbering group, performs brightness gradient difference comparison on adjacent pixel blocks in the image, detects the number of boundary change directions, filters out image numbers with concentrated edge directions, and obtains the image edge direction distribution trend. The system retrieves grayscale inspection images collected by the drone according to the patrol path numbering group and performs brightness gradient calculation on each pixel in the image. Using edge detection algorithms such as the Sobel operator, the system calculates the gradient magnitude and direction of each pixel. To extract the cable outline from the complex background, the system sets a gradient magnitude threshold of 50, selecting only pixels with magnitudes exceeding this threshold as edge candidates. Statistical analysis is performed on the gradient directions of these candidate points to construct a direction distribution histogram. The system pays particular attention to the concentration trend of gradient directions. If it finds that the cumulative number of pixels within a 10-degree directional interval exceeds 40% of the total edge pixels, the system determines that there is a clear edge direction concentration trend in the image. This trend usually corresponds to the main direction of the lightning protection cable, distinguishing it from the chaotic background texture. Once this trend is confirmed, the system records the image number and the corresponding edge direction distribution data, thus obtaining the image edge direction distribution trend.

[0030] The texture orientation recognition submodule identifies regions in the image where the texture distribution density changes with direction based on the distribution trend of the image edge orientation, records the orientation information corresponding to the density abrupt change, performs overlap judgment on the orientation change line and the image edge trend line, and obtains the distribution of overlapping fragments of texture abrupt change. Based on the distribution trend of image edges, a dedicated texture scanning window is established, and a sliding scan is performed along a path perpendicular to the main direction of the cable. At each scanning position, the system uses gray-level co-occurrence matrix technology to calculate the contrast and energy indices of the texture, thereby deriving the texture density parameters. The system monitors the changes in this texture density with spatial location in real time, looking for abnormal abrupt changes. A texture density abrupt change threshold of 2.0 is set, a value based on experimental observations that the metallic luster reflection at the broken strand causes a doubling of texture clutter. When the density ratio of adjacent windows exceeds this threshold, the system records the exact orientation information of that location and connects these abrupt change points to form an orientation change line. This orientation change line is then overlapped with the image edge trend line, i.e., the outline of the main cable body. Only when the area of ​​the overlapping region exceeds 80% in space is the system confirmed that this texture abrupt change occurs on the cable body itself, rather than background interference. The coordinate information of the overlapping part is extracted to obtain the distribution of overlapping segments of texture abrupt changes. The suspected location extraction submodule extracts the corresponding numbered regions in the image based on the distribution of overlapping fragments with texture abrupt changes, filters out locations with concentrated edges and abrupt changes in texture direction, judges the contour aggregation in the numbered images, records the location numbers with continuous boundary trends in the marked areas, and obtains the image annotation information set of lightning protection wire breakage locations.

[0031] Based on the distribution of overlapping texture fragments, regions of interest (ROIs) are delineated on the original high-resolution image. Within these specific regions, the system performs fine Canny edge detection to extract minute contour fragments. The system analyzes the aggregation of these contours, focusing on the number and directional dispersion of the fragments. A fracture characteristic criterion is set as follows: if the number of contour fragments in a region exceeds 10 and the variance of directional dispersion is greater than 0.5, the region is considered to exhibit typical characteristics of fracture or strand breakage, as intact cable surface contours are usually continuous and oriented in a consistent manner. To eliminate false alarms, the system further tracks the boundary trends on both sides of the marked region. If continuous and smooth boundaries exist on both sides of the fracture characteristic region, representing an unbroken section of the cable, the region is confirmed as the true fracture point. The pixel center coordinates and corresponding image ID at this location are recorded, generating a detailed record containing the image ID, pixel coordinates, and defect type, thereby obtaining an image annotation information set for the lightning protection wire fracture location.

[0032] Please see Figure 6 The fracture risk assessment module includes: The coordinate range verification submodule, based on the image annotation information set of the lightning protection wire fracture location, compares the node number and segment position in the list of multi-source triggering nodes of the lightning protection wire to obtain the intersection range of the image coordinates and segment coordinates, performs a comparison operation, screens the overlapping area of ​​the image coordinates and segment coordinates, and obtains the coordinate intersection matching area. Based on the image annotation information set of lightning protection wire fracture locations, and utilizing the intrinsic parameters of the UAV camera, GPS coordinates at the time of shooting, and the regional digital elevation model, the system uses collinearity equation inversion calculations to restore the pixel coordinates of the fracture points in the images to their true geospatial coordinates. Simultaneously, the system accesses the list of multi-source triggering nodes of the lightning protection wire to obtain the segment coordinate range of the lightning strike node. The system performs a rigorous spatial comparison operation, calculating the intersection of the inverted defect geographic coordinates and the node segment coordinate range. Considering the errors in civilian-grade GPS positioning and the deviations in the inversion calculation, a matching tolerance of 5 meters is set. Only when the image inversion coordinates fall within the segment coordinate range and its tolerance zone is the system considered to have substantial overlap between the image coordinates and the segment coordinates. All areas meeting the overlap condition are filtered out to obtain the coordinate intersection matching area.

[0033] The adjacent interval extraction submodule extracts the coordinate intervals within adjacent numbered segments based on the coordinate intersection matching area, divides the coordinate range of adjacent areas according to the segment number information, filters the key risk locations within adjacent numbered intervals, and obtains the adjacent risk interval set. Based on the coordinate cross-matching region, the system first identifies the central segment where the risk lies, and then automatically extracts the preceding and following segments according to the topological connection structure of the lightning protection wire. Given that tension imbalance propagates to adjacent tension segments when the steel strand breaks or is damaged, the system, referencing mechanical experimental data, defines the significant impact area of ​​tension redistribution as approximately 20% of the span. Accordingly, the system formulates a rule for dividing risk-related intervals: the entire central risk segment, along with the latter 20% of the adjacent preceding segment and the former 20% of the adjacent following segment, are collectively designated as key risk locations. Based on this rule, the system filters all relevant segments, extracting nodes located within these key areas to obtain a set of adjacent risk intervals.

[0034] The risk monitoring result output submodule extracts all suspected fracture risk nodes based on adjacent risk interval sets, monitors the area range, marks existing fracture risk points, outputs fracture risk assessment information, and obtains the lightning protection wire fracture risk monitoring results.

[0035] Based on a set of adjacent risk intervals, all suspected breakage risk nodes within that range are extracted and comprehensively monitored in conjunction with the regional scope. The system calculates a risk coefficient, which is a weighted average of image recognition confidence and sensor triggering intensity. The weighting is set based on statistical analysis of historical cases, with image anomalies weighted at 0.4 and sensor anomalies weighted at 0.6, reflecting the higher confidence of sensor data in breakage confirmation. The system compares the calculated comprehensive risk coefficient with a preset risk threshold of 0.75. When the risk coefficient of a node exceeds this threshold, the system marks it as a "high-risk breakage point," meaning that there is a very high probability of physical breakage at that location. The system packages all marked points and their risk levels, outputting breakage risk assessment information to obtain the lightning conductor breakage risk monitoring results.

[0036] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike, characterized in that, The system includes: The stress disturbance extraction module collects tension, electric field, and displacement data of the lightning conductor tension arm, equipotential bonding ring, and clamp joint. It rearranges the data columns by time, removes interrupted data segments, assigns the numbers to segments, performs segmentation operations on the segments, and obtains multi-point disturbances of the lightning conductor to form segment groups. The high-sensitivity node screening module, based on the tension slope, electric field disturbance range and displacement change interval in the segment group composed of multi-point disturbances of the lightning protection line, screens out nodes with three types of disturbances, aligns the node numbers with the segment positions, extracts the number information, and obtains a list of multi-source triggering nodes of the lightning protection line. The flight path planning module searches for the corresponding segments in the UAV path based on the number and segment coordinates in the list of multi-source triggering nodes of the lightning protection line, filters out paths that do not cross numbered segments and have short spatial intervals, extracts continuous segments with more stable path directions as flight segments, and obtains the UAV lightning strike inspection path number group. The image fracture extraction module calls the image number in the UAV lightning strike inspection path number group, compares the brightness gradient, extracts the edge direction, identifies the texture change direction, extracts the image number corresponding to the fracture area, and obtains the image annotation information set of the lightning protection wire fracture location.

2. The intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike as described in claim 1, characterized in that: The lightning protection wire multi-point disturbance segment group includes the disturbance occurrence time distribution, multi-node association structure within the segment, data continuity mapping relationship, disturbance category combination label, and segment boundary index. The lightning protection wire multi-source trigger node list includes tension change node number, electric field interference critical point, displacement change corresponding number, segment number to which the node belongs, and node trigger source type. The UAV lightning strike inspection path number group includes a path number list, effective continuous path segments, spatially adjacent path segments, turning optimized path segments, and the mapping relationship between path number and segment number. The lightning protection wire fracture location image annotation information set includes fracture image number, brightness edge features, texture density change area, fracture contour selection result, and suspected fracture location label.

3. The intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike as described in claim 1, characterized in that, The stress disturbance extraction module includes: The tension disturbance identification submodule collects tension disturbance operation data of the lightning conductor tension arm, equipotential bonding ring, and clamp joint. It adjusts the tension data order according to the time field, removes interrupted numbered segments, determines the tension change location based on the order change trend, selects the fluctuation direction and compares it with the peak interval to obtain the tension disturbance change amplitude. The electric field displacement synchronization submodule collects synchronous electric field and displacement disturbance operation data according to the amplitude of the tension disturbance change, aligns the electric field and displacement data according to the time field of the tension data, extracts the corresponding content at a uniform interval, compares the difference trends between tension, electric field and displacement, and obtains the range of multi-source disturbance differences. The segment interval processing submodule extracts the disturbance content within the assigned number segment based on the range of multi-source disturbance differences, filters out continuous parts with continuously changing difference trends, groups and processes them according to the numbering information, locates multiple segments with disturbance continuity on the lightning protection line, and obtains multi-point disturbance interval groups for the lightning protection line.

4. The intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike as described in claim 1, characterized in that, The high-sensitivity node screening module includes: The disturbance behavior extraction submodule extracts the tension change trend, electric field disturbance length and displacement change time interval before and after each segment based on the multi-point disturbance interval group of the lightning protection wire. It performs judgment operations on the three types of disturbance content in sequence, filters out nodes that simultaneously exhibit tension change, electric field extension and displacement change, and obtains a multi-disturbance joint identification node sequence. The node positioning and association submodule calls the multi-perturbation joint identifier node sequence, matches the start and end time range of the corresponding paragraph according to the time information of each node, obtains the node occurrence position within the paragraph, sorts the node numbers within the same interval, aligns the positions according to the paragraph order, and obtains the paragraph position sequence corresponding to the number. The trigger point aggregation submodule assigns a unified number to the disturbance nodes appearing in the corresponding paragraph under each number according to the paragraph position sequence of the number, filters out the numbers whose node distribution within the paragraph is greater than the node concentration benchmark, obtains the node set with prominent lightning protection line triggering characteristics, and obtains the list of multi-source triggering nodes of the lightning protection line.

5. The intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike as described in claim 1, characterized in that, The flight path planning module includes: The node path mapping submodule, based on the list of multi-source triggering nodes of the lightning protection line, obtains the node number and the corresponding paragraph position, retrieves the corresponding coordinate segment in the UAV flight path, performs a one-to-one correspondence judgment on the node number and path coordinate according to the start and end order of the paragraph, excludes path content that crosses different number segments, and obtains the set of path segments corresponding to the node. The continuous flight segment filtering submodule calls the path segment set corresponding to the node, detects the spatial interval between adjacent path segments, judges the turning situation according to the path point order, filters the turning frequency and path point continuity, retains the path content with fewer turning occurrences and continuous point order, and obtains a continuous flight path segment set. The patrol path confirmation submodule performs an alignment operation on the numbering of each path segment according to the set of continuous flight path segments, extracts the path number order in the corresponding lightning protection line segment, completes the order verification between path numbers, and obtains the UAV lightning strike patrol path number group.

6. The intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike as described in claim 1, characterized in that, The image fracture extraction module includes: The image brightness comparison submodule calls each numbered image in the UAV lightning strike patrol path numbering group, performs brightness gradient difference comparison on adjacent pixel blocks in the image, detects the number of boundary change directions, filters out image numbers with concentrated edge directions, and obtains the image edge direction distribution trend. The texture orientation recognition submodule identifies regions in the image where the texture distribution density changes with orientation based on the image edge orientation distribution trend, records the orientation information corresponding to density abrupt changes, performs overlap judgment on the orientation change line and the image edge trend line, and obtains the distribution of overlapping texture abrupt change segments. The suspected location extraction submodule extracts the corresponding numbered regions in the image based on the distribution of overlapping texture mutation segments, filters out locations with concentrated edges and abrupt changes in texture direction, judges the contour aggregation in the numbered image, records the location numbers with continuous boundary trends in the marked area, and obtains the image annotation information set of lightning protection wire breakage locations.

7. The intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike as described in claim 1, characterized in that, The system also includes: The breakage risk assessment module compares the image annotation information set of the breakage location of the lightning conductor with the number position in the list of multi-source triggering nodes of the lightning conductor, checks the intersection range of the image coordinates and the segment coordinates, extracts the adjacent number intervals, and outputs the breakage risk monitoring results of the lightning conductor. The monitoring results of lightning protection wire breakage risk include high-risk section number, image and coordinate cross-comparison interval, breakage trend intensity level, risk section boundary range, and comprehensive judgment label.

8. The intelligent monitoring system for the risk of lightning conductor breakage after a lightning strike as described in claim 7, characterized in that, The fracture risk assessment module includes: The coordinate range verification submodule, based on the image annotation information set of the lightning protection wire fracture location, compares the node number and segment position in the list of multi-source triggering nodes of the lightning protection wire to obtain the intersection range of the image coordinates and segment coordinates, performs a comparison operation, screens the overlapping area of ​​the image coordinates and segment coordinates, and obtains the coordinate intersection matching area. The adjacent interval extraction submodule extracts the coordinate intervals within adjacent numbered segments based on the coordinate intersection matching area, divides the coordinate range of adjacent areas according to the segment number information, filters the key risk locations within adjacent numbered intervals, and obtains the adjacent risk interval set. The risk monitoring result output submodule extracts all suspected fracture risk nodes based on the adjacent risk interval set, monitors the area range, marks the existing fracture risk points, outputs fracture risk assessment information, and obtains the lightning protection wire fracture risk monitoring results.

Images