Hazard analysis methods, systems, equipment, and media based on UAV lidar

Abstract

This invention relates to the field of power transmission line environmental hazard detection technology, and discloses a hazard analysis method, system, equipment, and medium based on UAV lidar, including: S1, acquiring raw point cloud data and trajectory data of moving equipment; S2, parsing the raw point cloud data for one period using the trajectory data and tower ledger data to obtain intermediate point cloud data between adjacent towers; S3, summarizing and analyzing the parsed intermediate point cloud data; S4, searching for points within a radius of r meters in the power line point cloud data point by point; points within this range are identified as hazard points; S5, rendering and displaying all point cloud data, and distinguishing hazard points from non-hazard points using different colors; S6, determining whether the raw point cloud data parsing is complete; if so, the process ends; otherwise, returning to S2 to parse the raw point cloud data for the next period. This invention enables real-time classification and hazard analysis.

Description

Technical Field

[0001] This invention relates to the field of environmental hazard detection technology for power transmission lines, specifically to a hazard analysis method, system, equipment, and medium based on UAV lidar. Background Technology

[0002] Overhead transmission lines are a crucial infrastructure component of national economic development. Maintaining these lines is an essential part of building a modern power grid. Trees near overhead transmission lines that may pose a safety hazard are designated as potential hazards. To ensure a stable and reliable power supply and prevent accidents, power grid management departments need to regularly deploy significant personnel to conduct hazard inspections of transmission lines. With the construction and operation of ultra-high-voltage (UHV) power lines, their mileage is constantly expanding, and their connections and the geographical environments they traverse are becoming increasingly complex. Traditional manual on-site inspections are arduous, inefficient, pose certain safety risks, and lack accuracy. They are insufficient to meet the demands of rapidly developing power grid construction.

[0003] With the rapid development of lightweight drone technology, lightweight drone operations have become a new means of power line inspection. During inspections, measuring the distance between non-power elements such as trees and ground features and power lines using drone payloads is crucial for determining potential hazards. As LiDAR scanners continue to evolve, drone-based LiDAR technology can quickly acquire high-precision 3D spatial information of power transmission line corridors at a lower cost. The laser point cloud data it collects offers unparalleled advantages in spatial positioning and measurement accuracy compared to other remote sensing methods. This enables high-precision, rapid, and automatic extraction of power transmission lines and hazard analysis reports, significantly improving the efficiency of power line inspections. Furthermore, based on the point cloud data, 3D models of power transmission line corridors can be created, making power line management more intuitive and reliable. However, the power transmission line corridor presents challenges such as complex point cloud scenarios, diverse ground features, and varying point density. Existing UAV LiDAR systems for analyzing potential hazards during power transmission line inspections are cumbersome, requiring a series of post-processing steps on the point cloud data. After acquiring the data, the hazard analysis results cannot be obtained immediately. The raw data needs to be parsed and transformed, followed by a series of post-processing steps, including but not limited to: point cloud filtering, point cloud cropping, point cloud classification, pole and tower ledger information entry, hazard detection, result screening, and report export. This post-processing process is complex, inefficient, tedious, and time-consuming, significantly impacting the efficiency of inspection operations. Summary of the Invention

[0004] To overcome the shortcomings of the above measurement methods, this invention proposes a hazard analysis method, system, equipment, and medium based on UAV LiDAR. It establishes a hardware framework for UAV-based 3D LiDAR measurement and provides a UAV-based hazard measurement solution, overcoming the cumbersome and time-consuming nature of traditional UAV data analysis and processing. The real-time nature of this solution is reflected in the fact that analysis is completed during the data generation process, rather than through a series of post-processing methods after point cloud data is calculated. Compared with traditional hazard detection methods, this invention can automatically complete real-time hazard analysis, possessing the timeliness and automation lacking in traditional methods.

[0005] This invention is achieved through the following technical solution:

[0006] A hazard analysis method based on UAV lidar includes:

[0007] S1. Acquire radar data and trajectory data of the motion device, and perform joint calculations on the radar data and trajectory data to generate raw point cloud data;

[0008] S2. Based on the tower ledger data, the trajectory data is restricted, and the original point cloud data for one period is parsed using the trajectory data and the tower ledger data to obtain the intermediate point cloud data between adjacent towers;

[0009] S3. After parsing the original point cloud data for one cycle, the intermediate point cloud data obtained from the parsing are summarized and analyzed to obtain power line point cloud data and non-power element point cloud data.

[0010] S4. Calculate the three-dimensional spatial positional relationship between the power line point cloud data and the non-power point cloud data. By searching point by point within a radius of r meters of the power line point cloud data, the points within this range are the potential hazard points.

[0011] S5. Render and display all point cloud data, and distinguish between potential and non-potential points by rendering them with different colors;

[0012] S6. Obtain the data packet end marker by reading the data packet content in the radar data. Determine whether the original point cloud data has been parsed completely by using the data packet end marker. If yes, end the process; otherwise, jump to S2 to parse the original point cloud data for the next cycle.

[0013] As an optimization, the sports equipment is a drone.

[0014] As an optimization, before performing real-time parsing of the original point cloud data using the trajectory data and tower ledger data, it is necessary to convert the trajectory data into a format that is *.hmr, *.imr, or *.hex, converting the collected trajectory data in *.PosT format.

[0015] As an optimization, the specific steps of S2 are as follows:

[0016] S2.1 Read the tower ledger data and find the coordinates of two adjacent tower locations respectively;

[0017] S2.2. Based on the trajectory data, find the nearest neighbor trajectory points corresponding to the coordinates of the two adjacent towers, that is, the two trajectory points closest to the coordinates of the two towers respectively.

[0018] S2.3. Retain the first trajectory data within the range of the trajectory points corresponding to the position coordinates of the two adjacent towers. The original point cloud data within the range of the first trajectory data as the origin and the diameter as H is the intermediate point cloud data.

[0019] As an optimization, the specific steps for S3 are as follows:

[0020] S3.1 Power line identification is performed from several intermediate point cloud data by calculating the linear features of the point cloud and the horizontal angle of the direction, to obtain power element point cloud data. The horizontal angle of the direction is calculated and obtained from the coordinates of the poles in the ledger.

[0021] S3.2 Identify pole and tower point cloud data from the power element point cloud data in conjunction with the pole and tower ledger data. All point cloud data other than the pole and tower point cloud data in the power element point cloud data are preliminary power line point cloud data.

[0022] S3.3. From the preliminary power line point cloud data, vegetation point cloud data that is mistakenly identified as power line point cloud data is removed by Euclidean clustering to finally obtain power line point cloud data.

[0023] S3.4 All point cloud data other than the power line point cloud data and the tower point cloud data are non-power element point cloud data.

[0024] As an optimization, the specific process of identifying pole point cloud data from the power element point cloud data in combination with the pole ledger data is as follows:

[0025] The pole latitude and longitude coordinates of the poles are obtained from the pole ledger data and converted into position coordinates in the point cloud coordinate system. Points within a radius R = 60 meters with the position coordinates as the center are determined to be the pole point cloud.

[0026] As an optimization, it also includes S7, generating hazard reports, and exporting them.

[0027] This invention also discloses a hazard analysis system based on UAV LiDAR, used to implement the aforementioned hazard analysis method based on UAV LiDAR, comprising:

[0028] The acquisition module is used to acquire radar data and trajectory data of moving equipment, and to generate raw point cloud data by jointly solving the radar data and trajectory data.

[0029] The parsing module is used to restrict the trajectory data based on the tower ledger data, and to parse the original point cloud data through the trajectory data and the tower ledger data to obtain the intermediate point cloud data between adjacent towers;

[0030] The analysis module is used to summarize and analyze the intermediate point cloud data obtained after parsing the original point cloud data to obtain power line point cloud data and non-power element point cloud data.

[0031] The hazard point identification module is used to calculate the three-dimensional spatial positional relationship between the power line point cloud data and the non-power point cloud data. By searching point by point within a radius of r meters of the power line point cloud data, the points within this range are identified as hazard points.

[0032] The rendering module is used to render and display all point cloud data, and to distinguish between potential and non-potential points by rendering them with different colors.

[0033] The judgment module is used to obtain the data packet end symbol by reading the data packet content in the radar data, and to determine whether the original point cloud data has been completely parsed based on the data packet end symbol. If it has, the process ends; otherwise, the parsing module is used to parse the original point cloud data for the next cycle.

[0034] The present invention also discloses an electronic device, including at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the aforementioned method for hazard analysis based on UAV lidar.

[0035] The present invention also discloses a storage medium storing a computer program, which, when executed by a processor, implements a hazard analysis method based on UAV lidar as described above.

[0036] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0037] During point cloud data parsing, real-time classification of point cloud data can be performed. Relying on trajectory data and pole / tower ledger data, the accuracy and efficiency of point cloud classification can be improved, enabling real-time analysis of potential hazards in power corridors (current technologies cannot perform simultaneous data collection and analysis due to limitations in satellite signal processing time). This invention can perform real-time classification and hazard analysis during data parsing and generation; the parsed data is the analyzed hazard data.

[0038] Tree obstacle analysis requires absolute location positioning, which relies on satellite signals. Currently, satellite signal processing can only be completed after the flight is over. The solution proposed in this invention targets real-time analysis during point cloud parsing. The real-time nature of this solution is reflected in the fact that the analysis is completed during the data generation process. That is, tree obstacle analysis is completed at the same time as data parsing (i.e., point cloud data generation). Traditional solutions require a series of post-processing and manual processing after parsing before tree obstacle analysis and output of results can be performed. Attached Figure Description

[0039] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0040] Figure 1 A schematic diagram of a system for implementing the hazard analysis method based on UAV lidar described in this invention;

[0041] Figure 2 This is a flowchart of a hidden danger analysis method based on UAV lidar according to the present invention;

[0042] Figure 3 A schematic diagram for determining the location of potential hazards.

[0043] Figure 4 This is a schematic diagram of the analysis report. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0045] Example 1 provides a method for hazard analysis based on UAV lidar, such as... Figure 1-2 As shown, it includes:

[0046] S1. Acquire radar data and trajectory data of the motion device, and perform joint calculations on the radar data and trajectory data to generate raw point cloud data;

[0047] The motion device is a drone, specifically a multi-rotor drone platform. For example... Figure 1 As shown, radar and trajectory data can be acquired through the CBI-300 lidar system mounted on the UAV. After acquisition, the raw radar data (*.PCAP), pole and tower ledger data (*.KML), and trajectory data (*.PosT / *.APOS / *.PCD) are imported into the software system. The software system automatically processes the data and performs real-time hazard analysis. It automatically identifies pole and tower ledger names, identifies power channel elements, calculates hazards, and renders and displays them in real time. Hazard calculation is completed simultaneously with the processing. Users can choose whether to export a standardized hazard report according to their actual needs.

[0048] The hardware device (LiDAR system) of the UAV LiDAR real-time hazard analysis system transmits raw radar data and trajectory data through a 2.4GHz communication module, and transmits raw point cloud data (*.PCAP) and raw flight trajectory data (*.hmr, imr, *.hex) to the software device of the UAV LiDAR real-time hazard analysis system.

[0049] In other words, the UAV LiDAR real-time hazard analysis system collects data through hardware-based radar equipment. After data recording, it automatically transmits the data to the real-time hazard analysis software. The software first automatically parses the raw flight trajectory data to obtain flight trajectory data (.*PosT), which is then used for real-time hazard analysis. The specific real-time hazard analysis process is attached. Figure 2 As shown.

[0050] S2. Based on the tower ledger data, the trajectory data is restricted, and the original point cloud data for one period is parsed using the trajectory data and the tower ledger data to obtain the intermediate point cloud data between adjacent towers;

[0051] Before performing real-time analysis of the raw point cloud data using the trajectory data and tower ledger data, it is necessary to convert the trajectory data into a format that is *.hmr, *.imr, or *.hex.

[0052] After receiving the raw point cloud data and flight trajectory data, the software converts the raw trajectory data into processable trajectory data (*.PosT) through cloud base station services.

[0053] The point cloud is analyzed in real time using raw point cloud data, tower ledger data, and trajectory data. The analyzed trajectory is restricted by the ledger data. The process is as follows: read the tower location coordinates from the ledger, find the nearest neighbor trajectory point at its corresponding location, and perform real-time analysis of the point cloud data by retaining the trajectory data within the range of the corresponding trajectory points of adjacent towers.

[0054] The specific steps of S2 are as follows:

[0055] S2.1 Read the tower ledger data and find the coordinates of two adjacent tower locations respectively;

[0056] S2.2. Based on the trajectory data, find the nearest neighbor trajectory points corresponding to the coordinates of the two adjacent towers, that is, the two trajectory points closest to the coordinates of the two towers respectively.

[0057] S2.3. Retain the first trajectory data within the range of the trajectory points corresponding to the position coordinates of the two adjacent towers. The original point cloud data within the range of the first trajectory data as the origin and the diameter as H is the intermediate point cloud data, where H is a positive number.

[0058] A processing cycle T = 2 seconds is set, which means that 2 seconds of data parsing is considered one cycle. The data within one cycle is collected and analyzed in real time. The real-time analysis process is divided into two processes: power channel point cloud classification (S3) and hidden danger monitoring (S4).

[0059] S3. After parsing the original point cloud data for one cycle, the intermediate point cloud data obtained from the parsing are summarized and analyzed to obtain power line point cloud data and non-power element point cloud data.

[0060] The purpose of point cloud channel classification is to distinguish between two power element points: power poles and power lines, and non-power element points such as vegetation. Based on the characteristic that the direction of power lines is highly consistent with the transmission corridor, power lines are identified by calculating the linear characteristics of the point cloud and the horizontal angle of their direction.

[0061] The specific steps are as follows:

[0062] S3.1 Power line identification is performed from several intermediate point cloud data by calculating the linear features of the point cloud and the horizontal angle of the direction, to obtain power element point cloud data. The horizontal angle of the direction is calculated and obtained from the coordinates of the poles in the ledger.

[0063] The formula for calculating the linear features of a point cloud is:

[0064] L=(λ1-λ2) / λ1;

[0065] Wherein, λ1 and λ2 are the eigenvalues ​​of the XOY plane (horizontal plane), respectively. The eigenvalues ​​are calculated by analyzing the geometric characteristic relationship of the point cloud coordinates. They are distinguishable for features with obvious characteristics such as power lines and towers. L is the linear eigenvalue of each point cloud data. The specific values ​​of λ1 and λ2 can be calculated by constructing a covariance matrix. They are closely related to the point cloud data. This is existing technology and will not be elaborated here.

[0066] S3.2 Identify pole and tower point cloud data from the power element point cloud data in conjunction with the pole and tower ledger data. All point cloud data other than the pole and tower point cloud data in the power element point cloud data are preliminary power line point cloud data. Pole and tower point cloud data can be constrained and identified through the pole and tower ledger data. Specifically, the pole and tower latitude and longitude are obtained from the pole and tower ledger data and converted into position coordinates in the point cloud coordinate system. Points within a radius of R = 60 meters within the coordinate position are determined to be pole and tower point clouds.

[0067] S3.3. From the preliminary power line point cloud data, vegetation point cloud data that is mistakenly identified as power line point cloud data is removed by Euclidean clustering to finally obtain power line point cloud data.

[0068] S3.4 All point cloud data other than the power line point cloud data and the tower point cloud data are non-power element point cloud data.

[0069] Misclassified vegetation points were removed using Euclidean clustering, and the remaining point cloud points were uniformly identified as non-electrical element points.

[0070] S4. Calculate the three-dimensional spatial positional relationship between the power line point cloud data and the non-power point cloud data. By searching point by point within a radius of r meters of the power line point cloud data, the points within this range are the potential hazard points.

[0071] The process iterates through the power line point cloud within a single cycle, calculating the three-dimensional spatial relationship between power lines and non-power element points. This is achieved by searching for points within a radius of r meters on each power line point. The radius is determined based on the relevant clearance distances specified in the "Operation Regulations for Overhead Transmission Lines" DL / T 741-2019. The specific principle is detailed in the appendix. Figure 3 As shown, the lines in the middle are power lines composed of power line point cloud data, and the points inside the dashed lines are potential hazard points.

[0072] S5. Render and display all point cloud data, and distinguish between potential and non-potential points by rendering them with different colors;

[0073] S6. Obtain the data packet end marker by reading the data packet content in the radar data. Determine whether the original point cloud data has been parsed completely by using the data packet end marker. If yes, end the process; otherwise, jump to S2 to parse the original point cloud data for the next cycle.

[0074] In this embodiment, after completing the point cloud hazard analysis within one cycle, the point cloud is rendered and displayed through the software system. Hazardous areas will be distinguished and displayed in red. The software system will then perform data processing and real-time analysis for the next time cycle, repeating this process until the data parsing is complete.

[0075] It also includes S7, generating hazard reports, and exporting them.

[0076] After completing all data analysis, users can choose whether to export the hazard report. If exported, the software system will automatically write the hazard parameters, point cloud screenshots, and coordinate location information into a pre-set Word template to generate an analysis report, as shown in the attached document. Figure 4 As shown.

[0077] Next, we will illustrate this method through specific implementation cases.

[0078] First scenario:

[0079] Data collection for the power transmission channel is conducted using a UAV equipped with a CBI-300 radar system. After the data collection is completed, the data is transmitted via the radar system's data transmission module. The transmitted data includes, but is not limited to, the following: raw lidar data, i.e., raw point cloud data (*.PCAP), and raw flight trajectory data (*.hmr, imr, *.hex).

[0080] The software system automatically identifies the received files, processes the acquired raw flight trajectory data on the cloud platform, and automatically uploads the HMR (*.hmr), IMR (*.imr), and HEX (*.hex) files to the cloud platform. The cloud platform automatically processes the data and converts it into a flight trajectory file (*.PosT). After the conversion is complete, the cloud platform automatically sends the data to the software system.

[0081] After receiving the returned flight trajectory file (*.PosT), the software system automatically calculates the point cloud data. The default calculation distance is within a range of 5m to 300m from the scanner, and the software system automatically performs real-time hazard analysis on the data within this range.

[0082] Real-time hazard analysis requires reading the pole and tower ledger data (*.kml) reserved by the system to obtain the pole and tower coordinates and names. By comparing this data with the trajectory file, the closest trajectory point to the pole in planar distance is obtained. The specific calculation formula is as follows:

[0083]

[0084] In the formula, (x p y p (x) represents the coordinates of the trajectory point; t y t ) represents the coordinates of the tower ledger; d xy This is the horizontal distance between the trajectory point and the tower coordinates.

[0085] The trajectory file contains the time period recorded by the aircraft. After calculating and obtaining the trajectory point closest to the pole on the horizontal plane, the trajectory file is cropped, and trajectories within the time intervals of non-starting and ending poles are segmented and deleted. For the starting pole trajectory point, trajectories with a time less than 5 seconds are automatically deleted (i.e., trajectory points within 5 seconds of the starting pole are deleted). For the ending pole trajectory point, trajectories exceeding 5 seconds of its time are automatically deleted (i.e., trajectory points within 5 seconds after the ending pole are deleted). In short, only power line data within the corresponding time range is retained for analysis. The point cloud region corresponding to the retained trajectory data is analyzed, and the positional relationship between each point and the flight trajectory is determined point by point. If all power line point clouds are contained within a 60-meter range of the flight trajectory, then that point is identified as a pre-selected power facility point. During power line inspection, the flight trajectory will be consistent with the direction of the power lines, and the drone will maintain a certain safe distance from the power line towers, set at a 60-meter range, ensuring that all power line point clouds on the line are included.

[0086] This paper utilizes covariance features to achieve refined extraction of power element point clouds (power line point clouds, tower point clouds). The covariance feature calculation process is as follows: Point cloud data is stored in a kd-tree; a covariance matrix is ​​constructed, and eigenvalues ​​λ1, λ2, λ3 (λ1>λ2>λ3) and eigenvectors v1, v2, v3 are calculated to obtain different covariance features for power line and tower point cloud identification. The covariance features include vertical angle H, linearity L, and curvature C, and their calculation formulas are as follows:

[0087]

[0088]

[0089]

[0090] Power facilities can be categorized into three main types: 1. Transmission towers and their ancillary facilities; 2. Transmission lines; and 3. Non-power facilities. Non-pre-selected power facilities are uniformly classified as non-power facilities. The corresponding towers in the ledger are named and numbered according to the tower name at the corresponding coordinates in the ledger data.

[0091] Hazard Calculation: The system calculates the 3D spatial relationship between power line point cloud data and non-power facilities, comparing it to preset national standards. (A distance ≤ 2.5 meters is considered a critical defect; 2.5-3.6 meters is a serious defect; 3.6-4.5 meters is a general defect; and 4.5-5.5 meters is another defect.) Defect judgment values ​​can be preset according to actual needs. After all point cloud hazard analysis is completed, hazard information is automatically recorded and saved in SQL database format (*.db3). The software can optionally automatically export a power analysis report.

[0092] The second scenario:

[0093] The difference from the first case is that the input file lacks KML format data for the tower ledger. After the software system receives the corresponding flight trajectory file (*.PosT), it does not have KML format data for the tower ledger. It parses the point cloud area corresponding to the retained trajectory data, identifies the transmission towers and their ancillary facilities, and assigns tower names using the default naming method. The default name is "T" + number, that is, it starts recording from the number "T001".

[0094] Example 2 also discloses a hazard analysis system based on UAV LiDAR, used to implement the hazard analysis method based on UAV LiDAR described in Example 1, including:

[0095] The acquisition module acquires radar data and trajectory data of the moving equipment, and performs joint calculations on the radar data and trajectory data to generate raw point cloud data.

[0096] The parsing module is used to restrict the trajectory data based on the tower ledger data, and to parse the original point cloud data through the trajectory data and the tower ledger data to obtain the intermediate point cloud data between adjacent towers;

[0097] The analysis module is used to summarize and analyze the intermediate point cloud data obtained after parsing the original point cloud data to obtain power line point cloud data and non-power element point cloud data.

[0098] The hazard point identification module is used to calculate the three-dimensional spatial positional relationship between the power line point cloud data and the non-power point cloud data. By searching point by point within a radius of r meters of the power line point cloud data, the points within this range are identified as hazard points.

[0099] The rendering module is used to render and display all point cloud data, and to distinguish between potential and non-potential points by rendering them with different colors.

[0100] The judgment module is used to obtain the data packet end symbol by reading the data packet content in the radar data, and to determine whether the original point cloud data has been completely parsed based on the data packet end symbol. If it has, the process ends; otherwise, the parsing module is used to parse the original point cloud data for the next cycle.

[0101] An electronic device according to Embodiment 3 is characterized in that it includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a hazard analysis method based on UAV lidar as described in Embodiment 1.

[0102] Example 4 provides a storage medium storing a computer program, characterized in that, when the computer program is executed by a processor, it implements a hazard analysis method based on UAV lidar as described in Example 1.

[0103] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A hazard analysis method based on unmanned aerial vehicle lidar, characterized in that, include: S1. Acquire radar data and trajectory data of the motion device, and perform joint calculations on the radar data and trajectory data to generate raw point cloud data; S2. Based on the tower ledger data, the trajectory data is restricted, and the original point cloud data for one period is parsed using the trajectory data and the tower ledger data to obtain the intermediate point cloud data between adjacent towers; S3. After parsing the original point cloud data for one cycle, the intermediate point cloud data obtained from the parsing are summarized and analyzed to obtain power line point cloud data and non-power element point cloud data. S4. Calculate the three-dimensional spatial positional relationship between the power line point cloud data and the non-power point cloud data. By searching point by point within a radius of r meters of the power line point cloud data, the points within this range are the potential hazard points. S5. Render and display all point cloud data, and distinguish between potential and non-potential points by rendering them with different colors; S6. Obtain the data packet end marker by reading the data packet content in the radar data. Determine whether the original point cloud data has been parsed completely by using the data packet end marker. If yes, end the process; otherwise, jump to S2 to parse the original point cloud data for the next cycle.

2. The method for hazard analysis based on UAV lidar according to claim 1, characterized in that, The sports equipment mentioned is a drone.

3. The method for hazard analysis based on UAV lidar according to claim 1, characterized in that, Before performing real-time analysis of the raw point cloud data using the trajectory data and tower ledger data, it is necessary to convert the trajectory data into a format that is *.hmr, *.imr, or *.hex.

4. The method for hazard analysis based on UAV lidar according to claim 1, characterized in that, The specific steps of S2 are as follows: S2.1 Read the tower ledger data and find the coordinates of two adjacent tower locations respectively; S2.

2. Based on the trajectory data, find the nearest neighbor trajectory points corresponding to the coordinates of the two adjacent towers, that is, the two trajectory points closest to the coordinates of the two towers respectively. S2.

3. Retain the first trajectory data within the range of the trajectory points corresponding to the position coordinates of the two adjacent towers. The original point cloud data within the range of the first trajectory data as the origin and the diameter as H is the intermediate point cloud data.

5. The method for hazard analysis based on UAV lidar according to claim 4, characterized in that, The specific steps for S3 are as follows: S3.1 Power line identification is performed from several intermediate point cloud data by calculating the linear features of the point cloud and the horizontal angle of the direction, to obtain power element point cloud data. The horizontal angle of the direction is calculated and obtained from the coordinates of the poles in the ledger. S3.2 Identify pole and tower point cloud data from the power element point cloud data in conjunction with the pole and tower ledger data. All point cloud data other than the pole and tower point cloud data in the power element point cloud data are preliminary power line point cloud data. S3.

3. From the preliminary power line point cloud data, vegetation point cloud data that is mistakenly identified as power line point cloud data is removed by Euclidean clustering to finally obtain power line point cloud data. S3.4 All point cloud data other than the power line point cloud data and the pole point cloud data are non-power element point cloud data.

6. The method for hazard analysis based on UAV lidar according to claim 5, characterized in that, The specific process of identifying pole point cloud data from the power element point cloud data and combining it with the pole ledger data is as follows: The pole latitude and longitude coordinates are obtained from the pole ledger data and converted into position coordinates in the point cloud coordinate system. The power element points within a radius R = 60 meters with the position coordinates as the center are determined to be the pole point cloud.

7. The method for hazard analysis based on UAV lidar according to claim 1, characterized in that, It also includes S7, generating hazard reports, and exporting them.

8. A hazard analysis system based on UAV lidar, used to implement the hazard analysis method based on UAV lidar as described in any one of claims 1-7, characterized in that, include: The acquisition module is used to acquire radar data and trajectory data of moving equipment, and to generate raw point cloud data by jointly solving the radar data and trajectory data. The parsing module is used to restrict the trajectory data based on the tower ledger data, and to parse the original point cloud data through the trajectory data and the tower ledger data to obtain the intermediate point cloud data between adjacent towers; The analysis module is used to summarize and analyze the intermediate point cloud data obtained after parsing the original point cloud data to obtain power line point cloud data and non-power element point cloud data. The hazard point identification module is used to calculate the three-dimensional spatial positional relationship between the power line point cloud data and the non-power point cloud data. By searching point by point within a radius of r meters of the power line point cloud data, the points within this range are identified as hazard points. The rendering module is used to render and display all point cloud data, and to distinguish between potential and non-potential points by rendering them with different colors. The judgment module is used to obtain the data packet end symbol by reading the data packet content in the radar data, and to determine whether the original point cloud data has been completely parsed based on the data packet end symbol. If it has, the process ends; otherwise, the parsing module is used to parse the original point cloud data for the next cycle.

9. An electronic device, characterized in that, It includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a hazard analysis method based on UAV lidar as described in any one of claims 1 to 7.

10. A storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements a method for hazard analysis based on UAV lidar as described in any one of claims 1 to 7.

Images