Unmanned aerial vehicle path planning method and system for transmission and distribution line inspection
By constructing a three-dimensional model of the transmission and distribution lines, using point cloud filtering and Bayesian estimation to update obstacle confidence, and combining this with a fast travel algorithm to plan the UAV inspection path, the problem of dynamic obstacle recognition was solved, improving the accuracy and safety of the inspection.
Patent Information
- Application Number
- CN202511600669.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-04
- Publication Date
- 2026-01-06
AI Technical Summary
Existing drone path planning schemes are unable to adapt to dynamically changing obstacle distributions and cannot accurately identify obstacles in complex environments, leading to safety hazards during inspections.
By constructing a three-dimensional model of the transmission and distribution line, the obstacle confidence is updated using point cloud filtering algorithm and Bayesian estimation, and the UAV inspection path is planned by combining the fast travel algorithm to achieve dynamic correction and multi-path fusion.
It significantly improves the adaptability of drones in complex terrain and dynamic environments, ensures that inspection paths match the actual environment, avoids collisions and missed inspections, achieves inspection accuracy and safety, and optimizes resource utilization.
Smart Images
Figure CN121277205A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to unmanned aerial vehicle path planning, in particular to an unmanned aerial vehicle path planning method and system for power transmission and distribution line inspection. BACKGROUND
[0002] With the continuous expansion of the power system, the inspection and maintenance of power transmission and distribution lines have become a key link to ensure the safe and stable operation of the power grid. The traditional manual inspection method is limited by complex geographical environment, low inspection efficiency, and high safety risks, and it is difficult to meet the requirements of modern power grid for efficiency, accuracy and safety. Especially in mountainous areas, rivers and high-altitude areas, manual inspection not only has high cost, but also has safety risks such as falling and electric shock. Therefore, using unmanned aerial vehicles for power transmission and distribution line inspection has become an important development direction of the industry.
[0003] At present, unmanned aerial vehicle inspection technology has been gradually applied to the power industry, but the existing unmanned aerial vehicle path planning scheme still has limitations. For example, path planning based on static environment modeling cannot adapt to the dynamic changes of obstacle distribution (such as tree growth, temporary buildings, etc.), resulting in the need for frequent manual intervention during inspection; and path planning based on a single data source (such as relying only on two-dimensional maps or low-precision three-dimensional models) cannot accurately identify small obstacles in complex environments, which is prone to collisions. In addition, traditional methods usually ignore the uncertainty of obstacle confidence during path optimization, resulting in a certain deviation between the planned path and the actual environment, affecting the inspection efficiency and safety.
[0004] In view of the above problems, there is an urgent need for an unmanned aerial vehicle path planning method and system for power transmission and distribution line inspection that can integrate multi-source data, dynamically update environmental information, and optimize path reliability, providing technical support for intelligent operation and maintenance of power transmission and distribution lines. SUMMARY
[0005] (I) Technical problems to be solved
[0006] In view of the above shortcomings of the prior art, the present application provides an unmanned aerial vehicle path planning method and system for power transmission and distribution line inspection, which can effectively overcome the defects that the prior art cannot adapt to the dynamic changes of obstacle distribution and cannot accurately identify obstacles in complex environments, resulting in a large safety risk of the planned path.
[0007] (II) Technical solutions
[0008] To achieve the above purpose, the present application is realized by the following technical solutions:
[0009] The unmanned aerial vehicle path planning method for power transmission and distribution line inspection comprises the following steps:
[0010] S1, modeling the tower and surrounding environment of the transmission and distribution line, determining the obstacles and detection points in the transmission and distribution line;
[0011] S2, planning a first unmanned aerial vehicle inspection path according to the obstacles and detection points in the transmission and distribution line;
[0012] S3, collecting three-dimensional point cloud data of the transmission and distribution line, and establishing a three-dimensional model;
[0013] S4, filtering the three-dimensional point cloud data by using a point cloud filtering algorithm to obtain the obstacles in the three-dimensional model;
[0014] S5, updating the confidence of the obstacles in the three-dimensional model based on Bayesian estimation to obtain an obstacle map;
[0015] S6, calculating the Euclidean distance between each obstacle in the obstacle map by using a fast marching algorithm, and planning a second unmanned aerial vehicle inspection path according to the obstacle map and the Euclidean distance;
[0016] S7, determining a final unmanned aerial vehicle inspection path by combining the first unmanned aerial vehicle inspection path and the second unmanned aerial vehicle inspection path.
[0017] Preferably, the modeling of the tower and surrounding environment of the transmission and distribution line in S1 to determine the obstacles and detection points in the transmission and distribution line comprises:
[0018] According to the inspection task range and the performance indicators of the unmanned aerial vehicle, the flight area of the unmanned aerial vehicle is divided, and the position data of various target objects existing on the digital map in the flight area is calibrated;
[0019] The shape of the target object is processed, and on the basis of simplifying the shape of the target object, a new closed curve, i.e. an obstacle avoidance boundary, is formed by moving outward by a corresponding distance;
[0020] The obstacle avoidance range is stereoscopic, and the obstacle avoidance range around the target object is expanded to multiple closed three-dimensional spaces, i.e. obstacle avoidance bodies.
[0021] Preferably, the stereoscopic expansion of the obstacle avoidance range around the target object to multiple closed three-dimensional spaces, i.e. obstacle avoidance bodies, comprises:
[0022] According to the shape of the target object, different methods are used to expand the obstacle avoidance range around the target object to multiple closed three-dimensional spaces, i.e. obstacle avoidance bodies;
[0023] The shape of the target object includes a shape containing only straight lines, a mixed shape containing straight lines and curves, and a step-like shape.
[0024] Preferably, the detection points in the transmission and distribution line include insulators, cross arms and lightning rods.
[0025] Preferably, in step S2, the first UAV inspection path is planned based on obstacles and detection points in the transmission and distribution line, including:
[0026] Based on the obstacle avoidance objects and detection points in the transmission and distribution line, the particle swarm optimization algorithm is used to plan the inspection path of the first UAV. By introducing the adaptive shrinkage factor to balance the global exploration and local development capabilities of the algorithm, and by introducing a dispersal strategy to perform dispersal operations in dense particle locations, the search capability of the algorithm is enhanced.
[0027] Preferably, the three-dimensional point cloud data of the transmission and distribution lines collected in S3 includes:
[0028] The two-dimensional space is meshed, and the mesh size is adjusted according to the type of power facility and the point cloud density;
[0029] The feature information in each grid is extracted and combined to obtain the three-dimensional point cloud data of the transmission and distribution line;
[0030] The feature information includes average height, point cloud density, and color information.
[0031] Preferably, the three-dimensional model is established in S3, including:
[0032] The position and shape of the tower in the 3D point cloud data are represented by geometric objects or solid models, and the height and area of the tower are marked to obtain the 3D model of the tower.
[0033] By representing the starting point, ending point, and line type of the transmission and distribution line in the 3D point cloud data with line segments or curves, a 3D model of the line is obtained.
[0034] Preferably, in step S4, a point cloud filtering algorithm is used to filter the 3D point cloud data to obtain obstacles in the 3D model, including:
[0035] The number of filter windows is determined based on the grid, so that the number of grids is equal to the number of filter windows;
[0036] Traverse each grid within each filtering window, sort the values of all grids in the neighborhood of the target grid, and select the value in the middle position as the median value to replace the value of the target grid;
[0037] The obstacles and their location information in the 3D model are determined based on the cruise altitude and grid values.
[0038] Preferably, in step S5, the confidence level of obstacles in the 3D model is updated based on Bayesian estimation to obtain an obstacle map, including:
[0039] The confidence level of the obstacle at the previous moment is used as the prior probability. The likelihood of the obstacle is calculated using the average height and point cloud density in the 3D point cloud data. The prior probability is multiplied by the likelihood and then normalized to obtain the posterior probability.
[0040] Set a confidence threshold, compare the posterior probability with the confidence threshold, update the confidence of obstacles in the 3D model based on the comparison result, and combine the updated obstacles to form an obstacle map.
[0041] The UAV path planning system for power transmission and distribution line inspection includes an environmental modeling module, a first inspection path planning module, a 3D model building module, an obstacle extraction module, an obstacle map generation module, a second inspection path planning module, and a multi-path fusion decision module.
[0042] The environmental modeling module models the towers and surrounding environment of the transmission and distribution lines, and identifies obstacles and detection points in the transmission and distribution lines.
[0043] The first inspection path planning module plans the first UAV inspection path based on obstacles and detection points in the power transmission and distribution line.
[0044] The 3D model building module collects 3D point cloud data of transmission and distribution lines and builds 3D models.
[0045] The obstacle extraction module uses a point cloud filtering algorithm to filter the 3D point cloud data to obtain the obstacles in the 3D model;
[0046] The obstacle map generation module updates the confidence of obstacles in the 3D model based on Bayesian estimation to obtain the obstacle map;
[0047] The second inspection path planning module uses a fast travel algorithm to calculate the Euclidean distance between each obstacle in the obstacle map, and plans the second UAV inspection path based on the obstacle map and the Euclidean distance.
[0048] The multi-path fusion decision module combines the first and second UAV inspection paths to determine the final UAV inspection path.
[0049] (III) Beneficial Effects
[0050] Compared with existing technologies, the UAV path planning method and system for power transmission and distribution line inspection provided by this invention have the following advantages:
[0051] 1) Enhance environmental perception and improve adaptability to complex scenarios
[0052] Step S1 involves constructing a model of the transmission and distribution line's towers and surrounding environment, combined with obstacle and detection point positioning, to comprehensively capture static and dynamic environmental elements around the line (such as trees, buildings, temporary facilities, etc.); Step S2 plans the first UAV inspection path based on the above environmental information, enabling the UAV to effectively avoid known obstacles during inspection and reduce path conflicts caused by missing environmental information. This closed-loop design of "modeling-identification-planning" significantly enhances the UAV's adaptability in complex terrain (such as mountainous areas, densely populated urban areas, etc.) and dynamically changing environments.
[0053] 2) Dynamically correct the path to ensure the accuracy and safety of the inspection.
[0054] Steps S3-S6 involve acquiring and filtering 3D point cloud data to construct a high-precision 3D model and extract obstacles. Based on the confidence update mechanism of Bayesian estimation, the uncertainty of obstacle position and shape (such as vegetation growth, temporary obstacle movement, etc.) can be dynamically corrected to generate a real-time obstacle map. Combined with the fast travel algorithm, the Euclidean distance between obstacles is calculated to further optimize the path and avoid risk areas. This achieves an upgrade from "static modeling" to "dynamic correction", ensuring that the UAV inspection path always matches the actual environment, avoiding collisions or missed inspections caused by environmental changes, and significantly improving the accuracy and safety of inspection.
[0055] 3) Integrate multiple source pathways to achieve a balance between resources and efficiency.
[0056] Step S7 integrates the first and second UAV inspection paths to form a final UAV inspection path that balances global coverage and local optimization. On the one hand, the first UAV inspection path ensures that no inspection points are missed, meeting the requirements for the integrity of the inspection task. On the other hand, the second UAV inspection path avoids real-time risks and reduces invalid flight distance. This "dual-path fusion" strategy avoids the lag of a single static path and prevents the waste of computing resources caused by over-reliance on real-time data, ultimately achieving the optimal balance between inspection efficiency, safety, and resource utilization. Attached Figure Description
[0057] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.
[0058] Figure 1 This is a schematic diagram of the process of the present invention;
[0059] Figure 2This is a flowchart illustrating the process of planning the first UAV inspection path in this invention.
[0060] Figure 3 This is a schematic diagram of the process for planning the second UAV inspection path in this invention. Detailed Implementation
[0061] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0062] The following describes the specific process of the UAV path planning method for power transmission and distribution line inspection provided by this invention, using a concrete example (e.g.) Figure 1 (as shown) and technical effects.
[0063] S1. Model the towers and surrounding environment of the transmission and distribution lines to identify obstacles and detection points within the lines, such as... Figure 2 As shown, it includes:
[0064] Based on the inspection task scope and the performance indicators of the UAV, the flight area of the UAV is divided, and the location data of various targets on the digital map within the flight area are marked.
[0065] The shape of the target object is processed. Based on the simplification of the target object's shape, it is moved outward by a corresponding distance to form a new closed curve, i.e., the obstacle avoidance boundary.
[0066] The obstacle avoidance range is made three-dimensional, expanding the obstacle avoidance range around the target object into multiple closed three-dimensional spaces, i.e., obstacle avoidance bodies.
[0067] Specifically, the obstacle avoidance range is made three-dimensional, expanding the obstacle avoidance range around the target object into multiple enclosed three-dimensional spaces, i.e., obstacle avoidance bodies, such as... Figure 2 As shown, it includes:
[0068] Based on the shape of the target object, different methods are used to expand the obstacle avoidance range around the target object into multiple closed three-dimensional spaces, i.e., obstacle avoidance bodies;
[0069] The shape of the target object includes shapes that contain only straight lines, mixed shapes that contain both straight lines and curves, and step-like shapes.
[0070] In the technical solution of this application, the detection points in the transmission and distribution lines include insulators, crossarms, and lightning rods.
[0071] S2. Plan the first UAV inspection path based on obstacles and detection points in the transmission and distribution line, such as... Figure 2 As shown, it includes:
[0072] Based on the obstacle avoidance objects and detection points in the transmission and distribution line, the particle swarm optimization algorithm is used to plan the inspection path of the first UAV. By introducing the adaptive shrinkage factor to balance the global exploration and local development capabilities of the algorithm, and by introducing a dispersal strategy to perform dispersal operations in dense particle locations, the search capability of the algorithm is enhanced.
[0073] The above technical solution, in step S1, by constructing a model of the transmission and distribution line towers and surrounding environment, combined with obstacle and detection point positioning, can comprehensively capture static and dynamic environmental elements (such as trees, buildings, temporary facilities, etc.) around the line; in step S2, based on the above environmental information, the first UAV inspection path is planned, so that the UAV can effectively avoid known obstacles during inspection and reduce path conflicts caused by missing environmental information. This closed-loop design of "modeling-identification-planning" significantly enhances the adaptability of UAVs in complex terrain (such as mountainous areas, densely populated urban areas, etc.) and dynamically changing environments.
[0074] S3. Collect three-dimensional point cloud data of transmission and distribution lines and establish a three-dimensional model.
[0075] Specifically, three-dimensional point cloud data of transmission and distribution lines are collected, such as... Figure 3 As shown, it includes:
[0076] The two-dimensional space is meshed, and the mesh size is adjusted according to the type of power facility and the point cloud density;
[0077] The feature information in each grid is extracted and combined to obtain the three-dimensional point cloud data of the transmission and distribution line;
[0078] The feature information includes average height, point cloud density, and color information.
[0079] Specifically, a three-dimensional model is created, such as Figure 3 As shown, it includes:
[0080] The position and shape of the tower in the 3D point cloud data are represented by geometric objects or solid models, and the height and area of the tower are marked to obtain the 3D model of the tower.
[0081] By representing the starting point, ending point, and line type of the transmission and distribution line in the 3D point cloud data with line segments or curves, a 3D model of the line is obtained.
[0082] S4. A point cloud filtering algorithm is used to filter the 3D point cloud data to obtain obstacles in the 3D model, such as... Figure 3 As shown, it includes:
[0083] The number of filter windows is determined based on the grid, so that the number of grids is equal to the number of filter windows;
[0084] Traverse each grid within each filtering window, sort the values of all grids in the neighborhood of the target grid, and select the value in the middle position as the median value to replace the value of the target grid;
[0085] The obstacles and their location information in the 3D model are determined based on the cruise altitude and grid values.
[0086] S5. Update the confidence scores of obstacles in the 3D model based on Bayesian estimation to obtain the obstacle map, such as... Figure 3 As shown, it includes:
[0087] The confidence level of the obstacle at the previous moment is used as the prior probability. The likelihood of the obstacle is calculated using the average height and point cloud density in the 3D point cloud data. The prior probability is multiplied by the likelihood and then normalized to obtain the posterior probability.
[0088] Set a confidence threshold, compare the posterior probability with the confidence threshold, update the confidence of obstacles in the 3D model based on the comparison result, and combine the updated obstacles to form an obstacle map.
[0089] S6. Use a fast travel algorithm to calculate the Euclidean distance between obstacles in the obstacle map, and plan the second UAV inspection path based on the obstacle map and the Euclidean distance.
[0090] In the above technical solution, steps S3 to S6 involve acquiring and filtering 3D point cloud data to construct a high-precision 3D model and extract obstacles. Based on a Bayesian estimation confidence update mechanism, the uncertainty of obstacle position and shape (such as vegetation growth, temporary obstacle movement, etc.) can be dynamically corrected to generate a real-time obstacle map. Combined with a fast travel algorithm to calculate the Euclidean distance between obstacles, the path is further optimized to avoid risk areas. This achieves an upgrade from "static modeling" to "dynamic correction," ensuring that the UAV inspection path always matches the actual environment, avoiding collisions or missed inspections caused by environmental changes, and significantly improving the accuracy and safety of inspections.
[0091] S7. Combine the first and second UAV inspection paths to determine the final UAV inspection path.
[0092] In the above technical solution, step S7 integrates the first UAV inspection path and the second UAV inspection path to form a final UAV inspection path that takes into account both global coverage and local optimization. On the one hand, the first UAV inspection path ensures that no inspection points are missed, meeting the requirements of the integrity of the inspection task; on the other hand, the second UAV inspection path avoids real-time risks and reduces invalid flight distance. This "dual-path fusion" strategy avoids the lag of a single static path and prevents the waste of computing resources caused by over-reliance on real-time data, ultimately achieving the optimal balance between inspection efficiency, safety and resource utilization.
[0093] Based on the above-disclosed method for UAV path planning for power transmission and distribution line inspection, this application also discloses a UAV path planning system for power transmission and distribution line inspection, including an environment modeling module, a first inspection path planning module, a 3D model building module, an obstacle extraction module, an obstacle map generation module, a second inspection path planning module, and a multi-path fusion decision module.
[0094] The environmental modeling module models the towers and surrounding environment of the transmission and distribution lines, and identifies obstacles and detection points in the transmission and distribution lines.
[0095] The first inspection path planning module plans the first UAV inspection path based on obstacles and detection points in the power transmission and distribution line.
[0096] The 3D model building module collects 3D point cloud data of transmission and distribution lines and builds 3D models.
[0097] The obstacle extraction module uses a point cloud filtering algorithm to filter the 3D point cloud data to obtain the obstacles in the 3D model;
[0098] The obstacle map generation module updates the confidence of obstacles in the 3D model based on Bayesian estimation to obtain the obstacle map;
[0099] The second inspection path planning module uses a fast travel algorithm to calculate the Euclidean distance between each obstacle in the obstacle map, and plans the second UAV inspection path based on the obstacle map and the Euclidean distance.
[0100] The multi-path fusion decision module combines the first and second UAV inspection paths to determine the final UAV inspection path.
[0101] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for unmanned aerial vehicle path planning for power transmission and distribution line inspection, characterized in that: The method comprises the following steps: S1, modeling the towers and the surrounding environment of the transmission and distribution line, and determining the obstacles and detection points in the transmission and distribution line; S2, planning a first unmanned aerial vehicle inspection path according to the obstacles and detection points in the transmission and distribution line; S3, collecting three-dimensional point cloud data of the transmission and distribution line, and establishing a three-dimensional model; S4, filtering the three-dimensional point cloud data by using a point cloud filtering algorithm to obtain the obstacles in the three-dimensional model; S5, updating the confidence of the obstacles in the three-dimensional model based on Bayesian estimation to obtain an obstacle map; S6, calculating the Euclidean distance between each obstacle in the obstacle map by using a fast marching algorithm, and planning a second unmanned aerial vehicle inspection path according to the obstacle map and the Euclidean distance; S7, determining a final unmanned aerial vehicle inspection path by combining the first unmanned aerial vehicle inspection path and the second unmanned aerial vehicle inspection path. 2.The method of claim 1, wherein: In S1, the towers and the surrounding environment of the transmission and distribution line are modeled, and the obstacles and detection points in the transmission and distribution line are determined, comprising: According to the inspection task range and the performance indicators of the unmanned aerial vehicle, the flight area of the unmanned aerial vehicle is divided, and the position data of various target objects existing on the digital map in the flight area is calibrated; The shape of the target object is processed, and on the basis of simplifying the shape of the target object, a new closed curve, i.e. an obstacle avoidance boundary, is formed by moving outward by a corresponding distance; The obstacle avoidance range is stereoscopic, and the obstacle avoidance range around the target object is expanded to multiple closed three-dimensional spaces, i.e. obstacle avoidance bodies. 3.The method of claim 2, wherein: The obstacle avoidance range around the target object is expanded to multiple closed three-dimensional spaces, i.e. obstacle avoidance bodies, by different methods according to the shape of the target object; The shape of the target object includes a shape containing only straight lines, a mixed shape containing straight lines and curves, and a step-like shape. The detection points in the transmission and distribution line include insulators, crossarms and lightning rods.
4. The unmanned aerial vehicle path planning method for pipeline inspection according to claim 2, characterized in that: In S2, the first unmanned aerial vehicle inspection path is planned according to the obstacles and detection points in the transmission and distribution line, comprising:
5. The unmanned aerial vehicle path planning method for pipeline inspection according to claim 2, characterized in that: According to the obstacle avoidance bodies and detection points in the transmission and distribution line, a particle swarm algorithm is used to plan the first unmanned aerial vehicle inspection path, the ability of global exploration and local development is balanced by introducing an adaptive contraction factor, and the search ability of the algorithm is enhanced by introducing a dispersion strategy for dispersion operation in the position where particles are dense. In S3, the three-dimensional point cloud data of the transmission and distribution line is collected, comprising: 6.The method for unmanned aerial vehicle path planning for power transmission and distribution line inspection according to claim 1, characterized in that: The two-dimensional space is gridded, and the grid size is adjusted according to the type of power facility and the point cloud density; Feature information in each grid is extracted and combined to obtain the three-dimensional point cloud data of the transmission and distribution line; The feature information includes average height, point cloud density and color information. In S3, the three-dimensional model is established, comprising:
7. The unmanned aerial vehicle path planning method for pipeline inspection according to claim 6, characterized in that: The position and shape of the tower in the three-dimensional point cloud data are represented by geometric bodies or solid models, and the height and footprint area of the tower are marked to obtain a tower three-dimensional model; The start point, end point and line type of the transmission and distribution line in the three-dimensional point cloud data are represented by line segments or curves to obtain a line three-dimensional model. In S4, the three-dimensional point cloud data is filtered by using a point cloud filtering algorithm to obtain the obstacles in the three-dimensional model, comprising:
8. The unmanned aerial vehicle path planning method for pipeline inspection according to claim 7, characterized in that: The number of filter windows is determined according to the grid, so that the number of grids is equal to the number of filter windows; Each grid in each filter window is traversed, and the values of all grids in the target grid neighborhood are sorted, and the value located in the middle position is selected as the median to replace the value of the target grid; The cruise altitude and the value of the grid are used to determine the obstacle and its position information in the three-dimensional model. 9.The method for unmanned aerial vehicle path planning for power transmission and distribution line inspection according to claim 8, characterized in that: In S5, the confidence of the obstacle in the three-dimensional model is updated based on Bayesian estimation to obtain an obstacle map, including: The confidence of the obstacle at the previous moment is taken as a prior probability, the likelihood of the obstacle is calculated through the average height and the point cloud density in the three-dimensional point cloud data, and the prior probability is multiplied by the likelihood and then normalized to obtain a posterior probability; A confidence threshold is set, the posterior probability is compared with the confidence threshold, and the confidence of the obstacle in the three-dimensional model is updated according to the comparison result, and the updated obstacles are combined to form an obstacle map.
10. The UAV path planning system for transmission and distribution line inspection, configured to perform the UAV path planning method for transmission and distribution line inspection according to claim 1, characterized in that: The system comprises an environment modeling module, a first inspection path planning module, a three-dimensional model establishing module, an obstacle extraction module, an obstacle map generating module, a second inspection path planning module, and a multi-path fusion decision module; The environment modeling module models the towers and surrounding environment of the transmission and distribution line to determine the obstacles and detection points in the transmission and distribution line; The first inspection path planning module plans a first unmanned aerial vehicle inspection path according to the obstacles and detection points in the transmission and distribution line; The three-dimensional model establishing module collects three-dimensional point cloud data of the transmission and distribution line and establishes a three-dimensional model; The obstacle extraction module filters the three-dimensional point cloud data using a point cloud filtering algorithm to obtain the obstacles in the three-dimensional model; The obstacle map generating module updates the confidence of the obstacle in the three-dimensional model based on Bayesian estimation to obtain an obstacle map; The second inspection path planning module calculates the Euclidean distance between each obstacle in the obstacle map using a fast marching algorithm, and plans a second unmanned aerial vehicle inspection path according to the obstacle map and the Euclidean distance; The multi-path fusion decision module determines a final unmanned aerial vehicle inspection path by combining the first unmanned aerial vehicle inspection path and the second unmanned aerial vehicle inspection path.