UAV patrol strategy generation method and system for transmission line early warning

Abstract

The present invention discloses a method for generating a drone inspection strategy for power transmission line early warning. The method relates to the field of power transmission line monitoring technology. GIS information is used to construct a three-dimensional inspection map containing terrain, buildings, and vegetation, and to delineate a power transmission line model. Historical line anomaly data is analyzed to screen out sections with an abnormality probability higher than a preset value, mark them on the three-dimensional map, and further determine the abnormal sections based on environmental and line characteristics. A drone model is constructed on the three-dimensional map to generate several initial inspection routes. The routes are corrected through simulated movement and scanning to ensure that abnormal sections are preferentially covered. Route quality is evaluated based on three dimensions: energy consumption, inspection efficiency, and abnormal missed detection ratio, with reference inspection routes being preferred. The above-mentioned technical solution of the present invention improves the accuracy of abnormal section identification, optimizes drone inspection efficiency, reduces energy consumption and missed detection risks, and significantly improves the reliability and safety of power transmission line early warning.

Description

Technical Field

[0001] The present invention relates to the technical field of transmission line monitoring, and in particular to a method and system for generating a UAV patrol strategy for transmission line early warning. Background Art

[0002] With the rapid development of power systems, transmission lines, as a core component of power transmission, have a direct impact on the reliability of the power grid and the normal operation of the social economy. However, transmission lines are often distributed across vast geographical areas, facing multiple environmental challenges such as complex terrain, vegetation cover, and building interference. They are also subject to factors such as natural aging and meteorological disasters, making them prone to abnormalities such as disconnections, short circuits, and tilting. To ensure line safety, traditional inspection methods rely primarily on manual on-site inspections or fixed-route drone inspections, but these methods have numerous shortcomings. Manual inspections are inefficient, costly, and difficult to implement in remote or hazardous areas. Fixed-route drone inspections lack flexibility and cannot dynamically adjust inspection priorities based on the actual abnormality risk of the line. This can lead to high-risk sections being missed and low-risk sections being over-inspected. Furthermore, existing inspection strategies struggle to effectively address the impact of obstructions in complex environments and lack comprehensive quality evaluation criteria for inspection routes, making it difficult to achieve an optimal balance between energy consumption, inspection efficiency, and anomaly detection rates. Summary of the Invention

[0003] The purpose of the present invention is to provide a method and system for generating a UAV inspection strategy that can improve the inspection efficiency of transmission line inspections and reduce inspection costs.

[0004] The present invention discloses a method for generating a UAV patrol strategy for power transmission line early warning, comprising:

[0005] Step S100, constructing a three-dimensional patrol map using GIS information, and delineating a transmission line model on the three-dimensional patrol map;

[0006] Step S200: Analyze historical line anomaly data to determine the probability of anomalies occurring in different transmission line sections, screen out transmission line sections with anomaly probabilities greater than or equal to a preset value, delineate the transmission line model on the patrol 3D map to obtain several abnormal line model sections, perform feature analysis on the abnormal line model sections, and further determine several abnormal line model sections.

[0007] Step S300: Constructing a drone model on the patrol 3D map and constructing several initial patrol routes for the drone model. The drone model is driven to move along the initial patrol routes and scans a preset range of the drone model. If an abnormal route model section appears within the preset range, the initial patrol route is modified so that the initial patrol route preferentially covers the abnormal route model section.

[0008] Step S400 , evaluating the line quality of the initial inspection route based on three dimensions: energy consumption, inspection efficiency, and abnormal missed detection ratio, and selecting a reference inspection route based on the line quality evaluation.

[0009] In some embodiments disclosed herein, a method for constructing a patrol three-dimensional map and delineating a transmission line model on the patrol three-dimensional map includes:

[0010] Step S101: Obtain GIS information, determine a plan map of the area near the transmission line, and the terrain, buildings, and vegetation near the transmission line. Build obstruction models for the terrain, buildings, and vegetation, respectively, and place the obstruction models on the plan map.

[0011] In step S102 , the latitude, longitude and altitude of the wiring nodes in the transmission line are determined, each wiring node is arranged on a plane map, and each wiring node is connected in the manner of a transmission line to obtain a transmission line model.

[0012] In some embodiments disclosed herein, a method for performing feature analysis on abnormal line model sections and re-identifying a plurality of abnormal line model sections includes:

[0013] Step S201: determining the terrain, buildings, and vegetation within a preset range of the abnormal line model section, and performing feature analysis on each of them to obtain environmental characteristics. The inclination of the abnormal line model section and the length of the entire line to which it belongs are determined to obtain line characteristics of the abnormal line model section.

[0014] Step S202 : The environmental features and the line features are associated and combined to obtain an abnormal feature group, and the abnormal feature group is used to perform a traversal feature comparison on the transmission line model to determine a number of new abnormal line model sections that meet the abnormal feature group.

[0015] In some embodiments disclosed herein, a method for performing feature analysis on terrain, buildings, and vegetation within a preset range of an abnormal line model section includes:

[0016] Step S2011: Using the transmission line model as a baseline, construct parallel feature analysis boundary lines on both sides, and record the area between the baseline and the boundary line as a feature analysis area;

[0017] Step S2012: A plurality of height collection points are evenly set in the feature analysis area to form a height collection point array. The highest point of the terrain, building, or vegetation environment corresponding to each height collection point is determined, and the height difference between each highest point of the environment and the power transmission line model is calculated. The height difference is matched one-to-one with the height collection point array to obtain a height difference array.

[0018] In step S2013, the height difference of the parallel lines of the transmission line model mapped in the height difference array is defined as a height difference series, the continuous performance characteristics in each height difference series are parameterized to obtain a number of continuous performance parameter groups, and the several continuous performance parameter groups are matched with the height difference series to form a continuous performance parameter group array, and the continuous performance parameter group array is identified as an environmental feature.

[0019] In some embodiments disclosed herein, a method for parameterizing the continuous performance features in each height difference series includes:

[0020] Step S20131, intercepting a continuous height difference segment in the height difference sequence, wherein the continuous height difference segment includes a plurality of height differences, and the quadratic difference between adjacent height differences is less than or equal to a preset value;

[0021] Step S20132, determine the difference segment length and the average height difference of the continuous height difference segment, identify the difference segment length and the average height difference as parameters of the continuous height difference segment, and configure the parameters of the continuous height difference segment in situ in the height difference series to form a continuous performance parameter group array.

[0022] In some embodiments disclosed herein, a method for evaluating the line quality of an initial inspection route includes:

[0023] Step S401: randomly select several initial inspection routes, drive the drone model to perform simulated inspections along the randomly selected initial inspection routes, and calculate the energy consumption in each simulated inspection. The total length of the transmission line model sections inspected by the drone model within a preset inspection period is calculated, and the ratio of missed inspections of the randomly generated abnormal line model sections to all abnormal line model sections is calculated. The ratio of the total section length to the preset inspection period is recorded as the inspection efficiency.

[0024] Step S402: averaging the energy consumption, inspection efficiency, and abnormal missed detection ratio corresponding to several randomly selected initial inspection routes to obtain the energy consumption for comparison, the inspection efficiency for comparison, and the abnormal missed detection ratio for comparison;

[0025] Step S403: Analyze the energy consumption, inspection efficiency, and abnormal missed detection ratio of each initial inspection route, and calculate the energy consumption difference vector between the energy consumption and the comparison energy consumption, the inspection efficiency difference vector between the inspection efficiency and the comparison inspection efficiency, and the missed detection difference vector between the abnormal missed detection ratio and the comparison abnormal missed detection ratio.

[0026] Step S404 : determining a line quality evaluation of the initial inspection route based on the energy consumption difference vector, the inspection efficiency difference vector, and the missed inspection difference vector.

[0027] In some embodiments disclosed herein, a method for determining a line quality evaluation of an initial inspection route based on an energy consumption difference vector, an inspection efficiency difference vector, and a missed inspection difference vector includes:

[0028] Step S4041: a plurality of energy consumption difference vector intervals are set for the energy consumption difference vector, a plurality of inspection difference vector intervals are set for the inspection difference vector, a plurality of missed detection difference vector intervals are set for the missed detection difference vector, an energy consumption influencing parameter is set for each energy consumption difference vector interval, an inspection efficiency influencing parameter is set for each inspection difference vector interval, and a missed detection influencing parameter is set for each missed detection difference vector interval;

[0029] Step S4042: Construct a first line quality evaluation operator based on the energy consumption influencing parameter, construct a second line quality evaluation operator based on the inspection efficiency influencing parameter, construct a third line quality evaluation operator based on the missed detection influencing parameter, and determine the line quality evaluation based on the first line quality evaluation operator, the second line quality evaluation operator, and the third line quality evaluation operator.

[0030] In some embodiments disclosed herein, a method for constructing several initial inspection routes for a drone model includes:

[0031] Step S301: Divide the patrol 3D map into patrol blocks and mark the starting point and ending point of the UAV model in each patrol block;

[0032] Step S302: determine the main extension direction of the transmission line model in each patrol block, set the longest patrol journey, randomly select several abnormal line model sections, and randomly select abnormal line model sections in sequence according to the main extension direction and connect them to form an initial patrol route.

[0033] In some embodiments disclosed in the present invention, a UAV inspection strategy generation system for transmission line early warning is also disclosed, including:

[0034] The first module is used to construct a three-dimensional patrol map using GIS information and delineate the transmission line model on the three-dimensional patrol map;

[0035] The second module is used to analyze historical line anomaly data, determine the anomaly probability of abnormalities in different transmission line sections, screen out transmission line sections with an anomaly probability greater than or equal to a preset value, and delineate the transmission line model on the patrol 3D map to obtain several abnormal line model sections. The abnormal line model sections are then analyzed for their characteristics and further determined.

[0036] The third module is used to build a drone model on the patrol 3D map and construct several initial patrol routes for the drone model. The drone model is driven to move on the initial patrol route and scan a preset range for the drone model. If an abnormal line model section appears within the preset range, the initial patrol route is corrected so that the initial patrol route preferentially covers the abnormal line model section.

[0037] The fourth module is used to evaluate the line quality of the initial inspection route based on three dimensions: energy consumption, inspection efficiency, and abnormal missed inspection ratio, and select a reference inspection route based on the line quality evaluation.

[0038] The present invention discloses a method for generating a drone inspection strategy for power transmission line early warning. The method relates to the field of power transmission line monitoring technology. GIS information is used to construct a three-dimensional inspection map containing terrain, buildings, and vegetation, and to delineate a power transmission line model. Historical line anomaly data is analyzed to screen out sections with an abnormality probability higher than a preset value, mark them on the three-dimensional map, and further determine the abnormal sections based on environmental and line characteristics. A drone model is constructed on the three-dimensional map to generate several initial inspection routes. The routes are corrected through simulated movement and scanning to ensure that abnormal sections are preferentially covered. Route quality is evaluated based on three dimensions: energy consumption, inspection efficiency, and abnormal missed detection ratio, with reference inspection routes being preferred. The above-mentioned technical solution of the present invention improves the accuracy of abnormal section identification, optimizes drone inspection efficiency, reduces energy consumption and missed detection risks, and significantly improves the reliability and safety of power transmission line early warning.

[0039] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a method step diagram for a method for generating a drone patrol strategy for transmission line early warning disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0042] The following will be combined with the accompanying drawings and specific embodiments to clearly and completely describe the technical solutions of the present invention. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention and cannot be understood as limiting the scope of protection of the present invention. Those skilled in the art in this field can make some non-essential improvements and adjustments based on the content of the present invention described below. In the present invention, unless otherwise clearly specified and limited, the technical terms used in the present invention should have the common meanings understood by those skilled in the art of the present invention.

[0043] Example:

[0044] The present invention discloses a method for generating a UAV patrol strategy for early warning of power transmission lines. Figure 1 ,include:

[0045] Step S100: constructing a patrol three-dimensional map using GIS information, and delineating a transmission line model on the patrol three-dimensional map.

[0046] The principle behind using GIS information to construct a 3D inspection map and delineate the transmission line model in step S100 is to integrate multi-source spatial data using Geographic Information System (GIS) technology, transforming the geographic features of the transmission line and its surroundings into a manageable 3D digital model, providing an accurate spatial foundation for subsequent drone inspections. The core of this process is the combination of 2D planar data (such as latitude and longitude coordinates) with 3D environmental information (such as elevation and terrain relief) to create a virtual scene that reflects the real world. First, the GIS system obtains a planar map of the area surrounding the transmission line, including information on features such as roads and rivers, from satellite imagery, surveying and mapping data, or publicly available databases. Next, terrain data (such as a digital elevation model (DEM)) is overlaid to construct the terrain's relief. Simultaneously, building outline data and vegetation height information are used to generate obstruction models, such as a 20-meter-tall building or a forest with an average height of 15 meters. These obstruction models are then applied to the planar map to create a 3D inspection map. For example, suppose a transmission line passes through the junction of mountains and plains. GIS data may show that the mountainous part has a slope 100 meters high, while the plain part has 5-meter-high trees. The 3D map will present these features as geometric bodies of different heights.

[0047] Based on this, the specific location of the transmission line requires precise determination of the longitude, latitude, and altitude of each node. Nodes are typically the locations of transmission towers or poles, obtained through field measurements or design drawings. For example, a node located at 113.5° east longitude, 23.1° north latitude, and 50 meters above sea level is considered. These nodes are individually marked on a 3D map and connected according to the actual alignment of the line (e.g., straight or curved), forming a transmission line model. For example, a 5-kilometer line might contain 10 nodes. The GIS system adjusts the line path based on inter-node distances and terrain to avoid overlap with obstructions (such as mountains). This approach relies on the spatial analysis capabilities of GIS. Through coordinate transformation and vector overlay, it digitizes complex geographic environments and line layouts, providing a reliable basis for subsequent anomaly analysis and route planning. This approach is further refined in (S101-S102), emphasizing the implementation details of the obstruction model and node connections, demonstrating the complete logic from data collection to model construction.

[0048] In step S200, the historical line abnormality data is analyzed to determine the abnormality probability of abnormalities in different transmission line sections, and the transmission line sections with abnormality probabilities greater than or equal to a preset value are screened out. The transmission line model on the patrol three-dimensional map is delineated to obtain several abnormal line model sections, and the characteristics of the abnormal line model sections are analyzed to determine several abnormal line model sections again.

[0049] The principle behind determining abnormal line sections by analyzing historical line anomaly data and combining it with feature analysis in step S200 is to utilize statistics and feature extraction techniques to uncover patterns in anomaly occurrences from historical data. This, combined with environmental and line characteristics, predicts potential risk sections, providing key targets for drone inspections. First, historical anomaly data for the transmission line is collected, such as records of line breaks, short circuits, or tilts over the past five years. This data may be sourced from inspection reports or sensor monitoring. For example, a 100-kilometer line is divided into 50 2-kilometer sections. Statistics reveal that sections 10-15 experience two anomalies per year due to complex terrain, while sections 40-45 experience three anomalies due to vegetation cover. Based on this information, the anomaly probability is calculated for each section (e.g., 2% for section 10). A preset value (e.g., 1%) is set to screen out sections with an anomaly probability greater than or equal to 1%. These sections are marked on the 3D map as preliminary abnormal line model sections, such as sections 10-15 and 40-45.

[0050] However, relying solely on historical data may overlook potentially high-risk sections that haven't yet occurred, necessitating further refinement of feature analysis. (S201-S202) describe this process in detail: Environmental features are extracted by analyzing the terrain (e.g., a 20° slope), buildings (e.g., a 10-meter-tall building), and vegetation (e.g., a 15-meter-tall tree) within the pre-set range of the abnormal section. The line's inclination (e.g., 5°) and the entire length of the line (e.g., 100 kilometers) are also measured to form line features. These features are combined into an abnormal feature group, such as "steep slope + tall vegetation + 5° inclination." The entire transmission line model is then traversed to identify other sections matching this feature group. For example, if section 20 also has steep slopes and tall vegetation, it will be identified as a potential abnormal section, even though there are no historical anomalies. The principle is to use historical data to drive probabilistic analysis and combine it with environmental and physical characteristics for pattern matching, enabling the expansion of risk from known to unknown risks. (S2011-S20132) further quantified feature analysis, such as evaluating terrain impact through height difference arrays, to ensure the scientific and comprehensive nature of anomaly identification.

[0051] In step S300, a drone model is constructed on the patrol three-dimensional map, and several initial patrol routes are constructed for the drone model. The drone model is driven to move on the initial patrol route, and a preset range is scanned for the drone model. If an abnormal line model section appears within the preset range, the initial patrol route is corrected so that the initial patrol route preferentially covers the abnormal line model section.

[0052] The principle behind constructing a drone model and optimizing the inspection route on a 3D map in step S300 is to upgrade the drone inspection route from static planning to intelligent optimization through virtual simulation and dynamic adjustment technology, ensuring efficient coverage of abnormal sections. First, a drone model is constructed on the 3D map generated in step S100, and its basic parameters are set, such as a flight altitude of 50 meters, a speed of 10 meters per second, and a scanning radius of 200 meters. Based on the transmission line model, several initial inspection routes are generated, such as the method mentioned in (S301-S302): the map is divided into inspection blocks, the starting and ending points of each block (e.g., the location of transmission towers) are determined, and the abnormal sections are connected along the main extension direction of the line (e.g., north-south) to generate a route. Assuming a line has three abnormal sections (sections 10, 20, and 40), the initial route might be "starting point - section 10 - section 20 - section 40 - end point." Multiple candidate routes are generated through random selection and connection.

[0053] The drone model then simulates movement along the initial route while monitoring its surroundings within a pre-set scanning range. If an abnormal segment is detected (e.g., segment 10 falls within the 200-meter range), a route correction mechanism is triggered. For example, the waypoint sequence is adjusted to "starting point - 10th - 40th - 20th - end point" to shorten the flight distance between abnormal segments or to add local detours to cover more abnormal points. For example, if a mountain obstructs segment 20, the corrected route may bypass the mountain to ensure scanning effectiveness. This principle combines the drone's spatial perception capabilities with the spatial distribution of abnormal segments, dynamically optimizing the route through real-time feedback. (S301-S302) The details of the block division and random connection embody the logic of moving from global planning to local optimization. This approach not only relies on abnormal segment data from the S200 but also utilizes the S100's 3D map to avoid obstructions, ultimately enabling intelligent inspection route generation and prioritizing coverage of high-risk areas.

[0054] Step S400 , evaluating the line quality of the initial inspection route based on three dimensions: energy consumption, inspection efficiency, and abnormal missed detection ratio, and selecting a reference inspection route based on the line quality evaluation.

[0055] The principle of evaluating and selecting an inspection route based on energy consumption, inspection efficiency, and abnormal missed detection ratio in step S400 is to select a reference route with the best overall performance from multiple initial routes through multi-dimensional quantitative evaluation and mathematical modeling, thereby providing a scientific basis for actual inspections.

[0056] In some embodiments disclosed herein, a method for constructing a patrol three-dimensional map and delineating a transmission line model on the patrol three-dimensional map includes:

[0057] Step S101: Obtain GIS information, determine a plan map of the area near the transmission line, and the terrain, buildings, and vegetation near the transmission line. Build obstruction models for the terrain, buildings, and vegetation, respectively, and place the obstruction models on the plan map.

[0058] The principle of step S101 is to use GIS technology to obtain geospatial data surrounding the transmission line and convert it into a three-dimensional inspection map containing environmental elements, providing spatial context for subsequent line model delineation and drone inspections. Specifically, planar map data near the transmission line is first extracted from the GIS database. This typically includes basic terrain information in a latitude and longitude coordinate system, such as roads, rivers, and administrative boundaries. For example, a transmission line may pass through a suburban area, and the GIS data will provide a two-dimensional outline map of the area. Next, terrain information (such as a digital elevation model (DEM)) is obtained to reflect the surface undulations, e.g., a hillside along the line is 100 meters high or a plain is 10 meters above sea level. Simultaneously, building data (such as building height and outline) and vegetation data (such as the average height and density of trees) are collected to construct three-dimensional obstruction models. For example, a 10-meter-tall factory building is modeled as a rectangular volume, while a 15-meter-tall pine forest is modeled as a group of cylinders or cones. These obstruction models are assigned to the planar map through coordinate matching, and terrain height information is superimposed to form a three-dimensional inspection map.

[0059] For example, consider a 20-kilometer transmission line running from point A (113.4°E, 23.0°N) to point B (113.6°E, 23.1°N). GIS data shows that the line's midsection passes through a hill at an altitude of 150 meters and a forest with an average height of 12 meters. Near the hill is a 20-meter-tall water tower. Step S101 first creates a planar map of the 20-kilometer area. The hill is then modeled as a 150-meter-high curved surface, the water tower as a 20-meter-tall cylinder, and the forest as a set of 12-meter-tall cones. These are then superimposed on the map to create a three-dimensional scene. This approach relies on GIS's spatial overlay and 3D visualization technologies, integrating multidimensional data into a unified geographic model, which provides a basis for subsequent line positioning and drone obstacle avoidance.

[0060] In step S102 , the latitude, longitude and altitude of the wiring nodes in the transmission line are determined, each wiring node is arranged on a plane map, and each wiring node is connected in the manner of a transmission line to obtain a transmission line model.

[0061] In some embodiments disclosed herein, a method for performing feature analysis on abnormal line model sections and re-identifying a plurality of abnormal line model sections includes:

[0062] Step S201 determines the terrain, buildings, and vegetation within a preset range of the abnormal line model segment and performs feature analysis to obtain environmental characteristics. The slope of the abnormal line model segment and the length of the entire line to which it belongs are then determined to obtain the line characteristics of the abnormal line model segment.

[0063] Step S202 : The environmental features and the line features are associated and combined to obtain an abnormal feature group, and the abnormal feature group is used to perform a traversal feature comparison on the transmission line model to determine a number of new abnormal line model sections that meet the abnormal feature group.

[0064] In some embodiments disclosed herein, a method for performing feature analysis on terrain, buildings, and vegetation within a preset range of an abnormal line model section includes:

[0065] In step S2011, the transmission line model is used as a baseline, and parallel feature analysis boundary lines are constructed on both sides. The area between the baseline and the boundary line is recorded as a feature analysis area.

[0066] The principle of step S2011 is to use the transmission line model as the center line, draw parallel feature analysis boundary lines on both sides of it, and define a clear analysis area, thereby limiting the scope of environmental feature extraction to the space directly related to the line. Specifically, the transmission line model is a three-dimensional path composed of wire nodes (as generated in step S102). This path is used as the baseline, and parallel lines of a certain distance (such as 50 meters or 100 meters) are set on both sides of it as boundaries. The area between these two boundary lines and the baseline is recorded as the feature analysis area, with the purpose of focusing on the near-field environment that may affect the safety of the line. For example, suppose a transmission line passes through a forest. The baseline is the line path, and the boundary lines may be set to 50 meters on each side of the line, with a total width of 100 meters. The rectangular or curved area formed is the feature analysis area.

[0067] For example, if a section of a railway line is 1 kilometer long and runs north-south, and GIS data indicates a hillside to the west and a forest to the east, step S2011 will draw a boundary line parallel to the line on both sides. The western boundary line, 50 meters from the line, will cover the hillside, and the eastern boundary line will cover the forest. This delineation of the area is based on spatial geometry principles, ensuring that the analysis scope is neither too broad (to avoid interference from irrelevant data) nor too narrow (to omit key environmental factors). Its significance lies in providing a standardized spatial framework for subsequent height data collection, ensuring consistent and comparable extraction of environmental features.

[0068] In step S2012, a number of height collection points are evenly set in the feature analysis area to form a height collection point array. The highest point of the terrain, building or vegetation environment corresponding to each height collection point is determined, and the height difference between each highest point of the environment and the transmission line model is calculated. The height difference is matched one-to-one with the height collection point array to obtain a height difference array.

[0069] The principle of step S2012 is to evenly arrange height collection points within the feature analysis area to form a two-dimensional point array. By measuring the difference between the highest environmental point and the line height at each point, an array of height differences is generated to quantify the potential impact of the environment on the line. Specifically, within the feature analysis area defined in step S2011, collection points are set at regular intervals (e.g., every 10 or 20 meters) to form a grid-like array of height collection points. For each collection point, the corresponding highest point of terrain (e.g., hillside height), building (e.g., roof height), or vegetation (e.g., treetop height) is determined using GIS data or 3D map information. For example, if a collection point is located in a forest with a 15-meter tree height, the highest environmental point is 15 meters; if it is located on a hillside with an 80-meter terrain height, the highest point is 80 meters. The height difference between this highest point and the transmission line model at that location is then calculated to obtain the height difference. The height difference values ​​of all collection points form the height difference array.

[0070] For example, assuming the feature analysis area is 100 meters wide (50 meters on each side) and 1 kilometer long, with a point set every 20 meters along the route and every 10 meters perpendicularly, there are a total of 50 × 10 = 500 collection points. Assuming the route height is 20 meters, if the hillside height to the west of a point is 80 meters, the difference is 80 - 20 = 60 meters; if the tree height to the east is 15 meters, the difference is 15 - 20 = -5 meters. The differences from these 500 points form a 500-element array. This principle is based on spatial sampling and height comparison. This point array discretizes continuous environmental changes into a computable dataset, providing a quantitative basis for subsequent feature extraction. This approach leverages the spatial resolution capabilities of GIS to ensure comprehensive data coverage and uniform distribution.

[0071] In step S2013, the height difference of the parallel lines of the transmission line model mapped in the height difference array is defined as a height difference series, the continuous performance characteristics in each height difference series are parameterized to obtain a number of continuous performance parameter groups, and the several continuous performance parameter groups are matched with the height difference series to form a continuous performance parameter group array, and the continuous performance parameter group array is identified as an environmental feature.

[0072] The principle of step S2013 is to perform structured processing on the height difference array, extracting continuous variation characteristics parallel to the transmission line and quantifying them into a continuous performance parameter group, forming the final expression of the environmental characteristics. Specifically, the height difference values ​​parallel to the transmission line (i.e., along the baseline) are extracted from the height difference array to form an ordered set of height difference value series. For example, if the feature analysis area has 50 transverse acquisition points, each corresponding to a longitudinal position on the line, 50 series may be generated, each containing 10 differences (e.g., 5 on the west side and 5 on the east side). Each series is analyzed for its continuous performance characteristics, namely, the trend or pattern of variation in the height difference values, and these characteristics are quantified using a parameterization method. The continuous performance parameter group may include the length of the continuous segment and the average difference value (as described in steps S20131-S20132). Ultimately, these parameter groups are associated with the series to form a continuous performance parameter group array, which is identified as the environmental characteristics.

[0073] For example, consider the sequence [-5, -5, 0, 10, 60, 60, 50, 20, 0, -5], representing the elevation difference from east to west along a section of the route. Analysis reveals that [-5, -5, 0] represents a continuous low-variance segment (length 3, average -3.3), while [60, 60, 50] represents a continuous high-variance segment (length 3, average 56.7). These parameter groups reflect the characteristics of forests (low-variance) and hillsides (high-variance), and are combined into a parameter group array. This approach, based on data sequence analysis and pattern recognition, quantifies continuity features and converts elevation differences into indicators of environmental impact. This approach integrates with subsequent steps (e.g., S202) to match abnormal feature groups and identify potential risk sections.

[0074] In some embodiments disclosed herein, a method for parameterizing the continuous performance features in each height difference series includes:

[0075] Step S20131 , intercepting a continuously changing height difference segment in the height difference sequence, wherein the continuous height difference segment includes a plurality of height differences, and the quadratic difference between adjacent height differences is less than or equal to a preset value.

[0076] Step S20132, determine the difference segment length and the average height difference of the continuous height difference segment, identify the difference segment length and the average height difference as parameters of the continuous height difference segment, and configure the parameters of the continuous height difference segment in situ in the height difference series to form a continuous performance parameter group array.

[0077] In some embodiments disclosed herein, a method for evaluating the line quality of an initial inspection route includes:

[0078] In step S401, several initial inspection routes are randomly selected, and the UAV model is driven to perform simulated inspections according to the randomly selected initial inspection routes. The energy consumption in each simulated inspection is counted respectively, and the total section length of the transmission line model section inspected by the UAV model under the preset inspection period is counted. The abnormal missed inspection ratio of the abnormal line model sections that are missed in the randomly generated abnormal line model sections to all abnormal line model sections is counted, wherein the ratio of the total section length to the preset inspection period is recorded as the inspection efficiency.

[0079] Step S402: averaging the energy consumption, inspection efficiency, and abnormal missed detection ratio corresponding to several randomly selected initial inspection routes to obtain the energy consumption for comparison, the inspection efficiency for comparison, and the abnormal missed detection ratio for comparison;

[0080] In step S403, the energy consumption, inspection efficiency, and abnormal missed detection ratio of each initial inspection route are analyzed, and the energy consumption difference vector between the energy consumption and the energy consumption for comparison is calculated, the inspection efficiency difference vector between the inspection efficiency and the inspection efficiency for comparison is calculated, and the missed detection difference vector between the abnormal missed detection ratio and the abnormal missed detection ratio for comparison is calculated.

[0081] Step S404 : determining a line quality evaluation of the initial inspection route based on the energy consumption difference vector, the inspection efficiency difference vector, and the missed inspection difference vector.

[0082] In some embodiments disclosed herein, a method for determining a line quality evaluation of an initial inspection route based on an energy consumption difference vector, an inspection efficiency difference vector, and a missed inspection difference vector includes:

[0083] In step S4041, several energy consumption difference vector intervals are set for the energy consumption difference vector, several inspection difference vector intervals are set for the inspection difference vector, several missed detection difference vector intervals are set for the missed detection difference vector, energy consumption influencing parameters are set for each energy consumption difference vector interval, inspection efficiency influencing parameters are set for each inspection difference vector interval, and missed detection influencing parameters are set for each missed detection difference vector interval.

[0084] Step S4042: Construct a first line quality evaluation operator based on the energy consumption influencing parameter, construct a second line quality evaluation operator based on the inspection efficiency influencing parameter, construct a third line quality evaluation operator based on the missed detection influencing parameter, and determine the line quality evaluation based on the first line quality evaluation operator, the second line quality evaluation operator, and the third line quality evaluation operator.

[0085] The expression for calculating line quality evaluation is:

[0086] .

[0087] in, For line quality evaluation, is the preset energy consumption weight coefficient, is the preset inspection efficiency weight coefficient, is the preset missed detection weight coefficient, is the adjustment coefficient of the parameters affecting energy consumption, is the adjustment coefficient of the parameters affecting inspection efficiency, is the adjustment coefficient of missed detection influencing parameters, Adjust constants for preset energy consumption effects, The adjustment constants for the parameters affecting the preset inspection efficiency are: Adjust the constant for the preset missed detection influencing parameter. is the parameter affecting energy consumption, Parameters that affect inspection efficiency: It is the influencing parameter of missed detection.

[0088] In some embodiments disclosed herein, a method for constructing several initial inspection routes for a drone model includes:

[0089] In step S301 , the patrol 3D map is divided into patrol blocks, and the initial point and the end point of the UAV model in each patrol block are marked.

[0090] Step S302: determine the main extension direction of the transmission line model in each patrol block, set the longest patrol journey, randomly select several abnormal line model sections, and randomly select abnormal line model sections in sequence according to the main extension direction and connect them to form an initial patrol route.

[0091] In some embodiments disclosed in the present invention, a UAV inspection strategy generation system for transmission line early warning is also disclosed, including:

[0092] The first module is used to construct a three-dimensional patrol map using GIS information and to delineate the transmission line model on the three-dimensional patrol map.

[0093] The second module is used to analyze historical line anomaly data, determine the abnormal probability of abnormalities in different transmission line sections, screen out transmission line sections with abnormal probabilities greater than or equal to preset values, and delineate the transmission line model on the patrol three-dimensional map to obtain several abnormal line model sections. The abnormal line model sections are then subjected to feature analysis to determine several abnormal line model sections again.

[0094] The third module is used to build a drone model on the patrol three-dimensional map, and to build several initial patrol routes for the drone model, drive the drone model to move on the initial patrol route, and scan the drone model within a preset range. If an abnormal line model section appears within the preset range, the initial patrol route is corrected so that the initial patrol route preferentially covers the abnormal line model section.

[0095] The fourth module is used to evaluate the line quality of the initial inspection route based on three dimensions: energy consumption, inspection efficiency, and abnormal missed inspection ratio, and select a reference inspection route based on the line quality evaluation.

[0096] The present invention discloses a method for generating a drone inspection strategy for power transmission line early warning. The method relates to the field of power transmission line monitoring technology. GIS information is used to construct a three-dimensional inspection map containing terrain, buildings, and vegetation, and to delineate a power transmission line model. Historical line anomaly data is analyzed to screen out sections with an abnormality probability higher than a preset value, mark them on the three-dimensional map, and further determine the abnormal sections based on environmental and line characteristics. A drone model is constructed on the three-dimensional map to generate several initial inspection routes. The routes are corrected through simulated movement and scanning to ensure that abnormal sections are preferentially covered. Route quality is evaluated based on three dimensions: energy consumption, inspection efficiency, and abnormal missed detection ratio, with reference inspection routes being preferred. The above-mentioned technical solution of the present invention improves the accuracy of abnormal section identification, optimizes drone inspection efficiency, reduces energy consumption and missed detection risks, and significantly improves the reliability and safety of power transmission line early warning.

[0097] Through the above description of the embodiments, those skilled in the art will clearly understand that the present invention can be implemented via hardware or via software combined with a necessary general-purpose hardware platform. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product. This software product can be stored on a non-volatile storage medium (such as a CD-ROM, USB flash drive, or external hard drive) and includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute the methods described in various implementation scenarios of the present invention.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for generating a UAV patrol strategy for power transmission line early warning, characterized by: include: Step S100, constructing a three-dimensional patrol map using GIS information, and delineating a transmission line model on the three-dimensional patrol map; Step S200: Analyze historical line anomaly data to determine the probability of anomalies occurring in different transmission line sections, select transmission line sections with anomaly probabilities greater than or equal to a preset value, delineate the transmission line model on the patrol 3D map to obtain several abnormal line model sections, perform feature analysis on the abnormal line model sections, and further determine several abnormal line model sections; Step S300: Constructing a drone model on the patrol 3D map and constructing several initial patrol routes for the drone model. The drone model is driven to move along the initial patrol routes and scans a preset range of the drone model. If an abnormal route model section appears within the preset range, the initial patrol route is modified so that the initial patrol route preferentially covers the abnormal route model section. Step S400 , evaluating the line quality of the initial inspection route based on three dimensions: energy consumption, inspection efficiency, and abnormal missed detection ratio, and selecting a reference inspection route based on the line quality evaluation.

2. The method for generating a UAV patrol strategy for power transmission line early warning according to claim 1 is characterized in that: The method of constructing a patrol three-dimensional map and delineating a transmission line model on the patrol three-dimensional map includes: Step S101: Obtain GIS information, determine a plan map of the area near the transmission line, and the terrain, buildings, and vegetation near the transmission line. Build obstruction models for the terrain, buildings, and vegetation, respectively, and place the obstruction models on the plan map. In step S102 , the latitude, longitude and altitude of the wiring nodes in the transmission line are determined, each wiring node is arranged on a plane map, and each wiring node is connected in the manner of a transmission line to obtain a transmission line model.

3. The method for generating a UAV patrol strategy for power transmission line early warning according to claim 2 is characterized in that: The method of performing feature analysis on abnormal line model sections and re-identifying several abnormal line model sections includes: Step S201: determining the terrain, buildings, and vegetation within a preset range of the abnormal line model section, and performing feature analysis on each of them to obtain environmental characteristics. The inclination of the abnormal line model section and the length of the entire line to which it belongs are determined to obtain line characteristics of the abnormal line model section. Step S202 : The environmental features and the line features are associated and combined to obtain an abnormal feature group, and the abnormal feature group is used to perform a traversal feature comparison on the transmission line model to determine a number of new abnormal line model sections that meet the abnormal feature group.

4. The method for generating a UAV patrol strategy for power transmission line early warning according to claim 3 is characterized in that: Methods for analyzing the characteristics of terrain, buildings, and vegetation within a preset range of an abnormal line model section include: Step S2011: Using the transmission line model as a baseline, construct parallel feature analysis boundary lines on both sides, and record the area between the baseline and the boundary line as a feature analysis area; Step S2012: A plurality of height collection points are evenly set in the feature analysis area to form a height collection point array. The highest point of the terrain, building, or vegetation environment corresponding to each height collection point is determined, and the height difference between each highest point of the environment and the power transmission line model is calculated. The height difference is matched one-to-one with the height collection point array to obtain a height difference array. In step S2013, the height difference of the parallel lines of the transmission line model mapped in the height difference array is defined as a height difference series, the continuous performance characteristics in each height difference series are parameterized to obtain a number of continuous performance parameter groups, and the several continuous performance parameter groups are matched with the height difference series to form a continuous performance parameter group array, and the continuous performance parameter group array is identified as an environmental feature.

5. The method for generating a UAV patrol strategy for power transmission line early warning according to claim 4 is characterized in that: The method for parameterizing the continuous performance characteristics in each height difference series includes: Step S20131, intercepting a continuous height difference segment in the height difference sequence, wherein the continuous height difference segment includes a plurality of height differences, and the quadratic difference between adjacent height differences is less than or equal to a preset value; Step S20132, determine the difference segment length and the average height difference of the continuous height difference segment, identify the difference segment length and the average height difference as parameters of the continuous height difference segment, and configure the parameters of the continuous height difference segment in situ in the height difference series to form a continuous performance parameter group array.

6. The method for generating a UAV patrol strategy for power transmission line early warning according to claim 1 is characterized in that: Methods for evaluating the line quality of the initial inspection route include: Step S401: randomly select several initial inspection routes, drive the drone model to perform simulated inspections along the randomly selected initial inspection routes, and calculate the energy consumption in each simulated inspection. The total length of the transmission line model sections inspected by the drone model within a preset inspection period is calculated, and the ratio of missed inspections of the randomly generated abnormal line model sections to all abnormal line model sections is calculated. The ratio of the total section length to the preset inspection period is recorded as the inspection efficiency. Step S402: averaging the energy consumption, inspection efficiency, and abnormal missed detection ratio corresponding to several randomly selected initial inspection routes to obtain the energy consumption for comparison, the inspection efficiency for comparison, and the abnormal missed detection ratio for comparison; Step S403: Analyze the energy consumption, inspection efficiency, and abnormal missed detection ratio of each initial inspection route, and calculate the energy consumption difference vector between the energy consumption and the comparison energy consumption, the inspection efficiency difference vector between the inspection efficiency and the comparison inspection efficiency, and the missed detection difference vector between the abnormal missed detection ratio and the comparison abnormal missed detection ratio. Step S404 : determining a line quality evaluation of the initial inspection route based on the energy consumption difference vector, the inspection efficiency difference vector, and the missed inspection difference vector.

7. The method for generating a UAV patrol strategy for power transmission line early warning according to claim 6 is characterized in that: The method for determining the line quality evaluation of the initial inspection route based on the energy consumption difference vector, the inspection efficiency difference vector, and the missed inspection difference vector includes: Step S4041: a plurality of energy consumption difference vector intervals are set for the energy consumption difference vector, a plurality of inspection difference vector intervals are set for the inspection difference vector, a plurality of missed detection difference vector intervals are set for the missed detection difference vector, an energy consumption influencing parameter is set for each energy consumption difference vector interval, an inspection efficiency influencing parameter is set for each inspection difference vector interval, and a missed detection influencing parameter is set for each missed detection difference vector interval; Step S4042: Construct a first line quality evaluation operator based on the energy consumption influencing parameter, construct a second line quality evaluation operator based on the inspection efficiency influencing parameter, construct a third line quality evaluation operator based on the missed detection influencing parameter, and determine the line quality evaluation based on the first line quality evaluation operator, the second line quality evaluation operator, and the third line quality evaluation operator.

8. The method for generating a UAV patrol strategy for power transmission line early warning according to claim 1, characterized in that: Methods for constructing several initial inspection routes for the drone model include: Step S301: Divide the patrol 3D map into patrol blocks and mark the starting point and ending point of the UAV model in each patrol block; Step S302: determine the main extension direction of the transmission line model in each patrol block, set the longest patrol journey, randomly select several abnormal line model sections, and randomly select abnormal line model sections in sequence according to the main extension direction and connect them to form an initial patrol route.

9. A UAV inspection strategy generation system for power transmission line early warning, characterized by: A method for generating a drone patrol strategy for executing any one of claims 1 to 8, comprising: The first module is used to construct a three-dimensional patrol map using GIS information and delineate the transmission line model on the three-dimensional patrol map; The second module is used to analyze historical line anomaly data, determine the anomaly probability of abnormalities in different transmission line sections, screen out transmission line sections with an anomaly probability greater than or equal to a preset value, and delineate the transmission line model on the patrol 3D map to obtain several abnormal line model sections. The abnormal line model sections are then analyzed for their characteristics and further determined. The third module is used to build a drone model on the patrol 3D map and construct several initial patrol routes for the drone model. The drone model is driven to move on the initial patrol route and scan a preset range for the drone model. If an abnormal line model section appears within the preset range, the initial patrol route is corrected so that the initial patrol route preferentially covers the abnormal line model section. The fourth module is used to evaluate the line quality of the initial inspection route based on three dimensions: energy consumption, inspection efficiency, and abnormal missed inspection ratio, and select a reference inspection route based on the line quality evaluation.

Images