Method for automatically generating inspection route of unmanned aerial vehicle of transformer substation

Through three-dimensional point cloud data processing and machine learning technology, substation drone inspection routes are automatically generated, solving the problems of low manual planning efficiency and major safety hazards, and achieving efficient and safe drone inspections.

CN118938983APending Publication Date: 2024-11-12国网黑龙江省电力有限公司牡丹江供电公司 +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410981306.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing drone inspection system requires manual planning of flight paths, resulting in inefficiency, high safety risks, high costs, and it is difficult to ensure the comprehensiveness and accuracy of inspections in complex environments.

Method used

Using three-dimensional point cloud data acquisition, preprocessing, device type identification, shooting direction calculation and waypoint position generation methods, automatic generation of drone inspection routes, RTK real-time differential technology and lidar data acquisition, combining machine learning and features Extract technology to realize automated route planning.

Benefits of technology

It greatly reduces the time of route planning, avoids manual operation errors, enhances safety, reduces manpower investment and inspection costs, and improves the efficiency and accuracy of inspections.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118938983A_ABST
    Figure CN118938983A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of unmanned aerial vehicles, in particular to a transformer substation unmanned aerial vehicle inspection route automatic generation method, which comprises the following steps of S1, acquiring three-dimensional point cloud data of a transformer substation; s2, preprocessing the acquired three-dimensional point cloud data of the transformer substation; s3, carrying out equipment type identification on the preprocessed three-dimensional point cloud data; s4, extracting the positions of geometric center points of vertical and horizontal projections of the equipment as the position of the equipment according to the identification result of the equipment; s5, marking equipment information on the point cloud model; s6, acquiring a shooting direction based on the equipment orientation and the equipment type information; s7, calculating a corresponding waypoint position; and S8, according to the obtained multiple waypoint positions, the related waypoint positions are connected in series to generate a route. According to the invention, automatic unmanned aerial vehicle route generation can be realized, and the substation inspection efficiency and safety are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles (UAVs), and in particular to a method for automatically generating inspection routes for substation UAVs. Background Art

[0002] As the scale of the power grid continues to expand, substations, as an important part of the power system, are of vital importance for safe and stable operation. Traditional manual inspection methods can no longer meet the needs of the rapid development of modern power grids. Drones are widely used in inspections and monitoring applications in forestry, agriculture, water conservancy, mining, electricity and other fields because of their advantages such as mobility, flexibility, low cost, less constraints from the natural environment, and the ability to carry different payloads and obtain observation data such as images and videos in real time, efficiently and safely. Of course, they have also become a new choice for substation inspections. However, how to automatically plan a reasonable inspection route for drones is still a major technical problem currently faced.

[0003] Traditional substation inspections rely on manual labor, which is inefficient, has great safety risks, and is costly. With the development of drone technology, although drone inspections have gradually replaced some manual inspections, most existing drone inspection systems require manual flight path planning, which is not only time-consuming and labor-intensive, but also difficult to ensure the comprehensiveness and accuracy of inspections in complex environments. Summary of the invention

[0004] In order to solve the problems in the background technology, the present invention proposes a method for automatically generating a substation drone inspection route. The present invention can realize automated drone route generation and improve the efficiency and safety of substation inspection.

[0005] To solve the above problems, the present invention adopts the following technical solution: a method for automatically generating a substation drone inspection route, comprising the following steps:

[0006] S1: Obtain three-dimensional point cloud data of substation;

[0007] S2: preprocessing the acquired substation three-dimensional point cloud data;

[0008] S3: Identify the device type based on the preprocessed 3D point cloud data;

[0009] S4: extracting the position of the geometric center point of the vertical and horizontal projections of the device as the position of the device according to the device recognition result;

[0010] S5: marking device information on the point cloud model, including device type, device location, and device orientation;

[0011] S6: acquiring a shooting direction based on the device orientation and device type information;

[0012] S7: Calculate the corresponding waypoint position based on the device position, the shooting direction and the input corresponding shooting distance;

[0013] S8: According to the obtained multiple waypoint positions, connect related waypoint positions in series to generate a route.

[0014] Furthermore, the three-dimensional point cloud data acquisition method in step S1 includes: utilizing RTK real-time differential technology and laser radar acquisition technology.

[0015] Furthermore, the preprocessing in step S2 includes: denoising, filtering and segmentation.

[0016] Furthermore, the device type identification on the preprocessed point cloud data in step S3 includes: an identification method based on feature extraction and machine learning.

[0017] Furthermore, the location information of the geometric center point in step S4 includes: longitude, latitude and altitude information.

[0018] Furthermore, the acquisition of the shooting direction in step S6 is based on the matching device shooting direction table.

[0019] Furthermore, the waypoint information of the waypoint position in step S7 includes longitude, latitude, altitude, azimuth, pitch angle and waypoint type.

[0020] Furthermore, the method for calculating the waypoint position in step S7 includes shooting a direction vector, converting the direction vector to a geographic angle, calculating the coordinates of the shooting point, and calculating the altitude of the shooting point.

[0021] Furthermore, the method for automatically generating a substation drone inspection route also includes obtaining safe shooting distance information of the drone and establishing a drone object model with a minimum safe distance as a radius.

[0022] Beneficial effects of the present invention: The present invention selects the target area to be surveyed in real time based on the electronic map of three-dimensional point cloud data. In view of the problem that most drone inspection systems in the prior art require manual planning of flight routes, which is time-consuming and labor-intensive, and it is difficult to ensure the comprehensiveness and accuracy of inspections in complex environments, the solution of the present invention does not require on-site measurement of the location information of the target area and manual planning of routes. It can automatically generate routes, greatly reducing the time for route planning, while avoiding manual operation errors, reducing accident risks, and enhancing safety. In addition, it reduces manpower investment and saves inspection costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0024] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0026] The present invention selects the target area to be surveyed in real time based on the electronic map of three-dimensional point cloud data. In view of the problem that most drone inspection systems in the prior art need to manually plan the flight route, which is time-consuming and labor-intensive, and it is difficult to ensure the comprehensiveness and accuracy of the inspection in complex environments, the solution of the present invention does not require on-site measurement of the location information of the target area and manual planning of the route, but can automatically generate the route, greatly reducing the time for route planning, while avoiding manual operation errors, reducing the risk of accidents, enhancing safety, reducing manpower investment, and saving inspection costs.

[0027] Specifically, Figure 1 As shown, a method for automatically generating a substation drone inspection route includes the following steps:

[0028] S1: Obtain three-dimensional point cloud data of substation;

[0029] S2: preprocessing the acquired substation three-dimensional point cloud data;

[0030] S3: Identify the device type based on the preprocessed 3D point cloud data;

[0031] S4: extracting the position of the geometric center point of the vertical and horizontal projections of the device as the position of the device according to the device recognition result;

[0032] S5: marking device information on the point cloud model, including device type, device location, and device orientation;

[0033] S6: acquiring a shooting direction based on the device orientation and device type information;

[0034] S7: Calculate the corresponding waypoint position based on the device position, the shooting direction and the input corresponding shooting distance;

[0035] S8: According to the obtained multiple waypoint positions, connect related waypoint positions in series to generate a route.

[0036] Specifically, in order to further improve the accuracy of device identification and positioning, the present invention supports manual marking of device information on the point cloud model. Users can select devices in the point cloud model through the interactive interface and input information such as the device type, location, and shooting direction. The manually marked information can be used for subsequent point cloud data training and optimization to improve the accuracy and efficiency of device identification and positioning.

[0037] Furthermore, the method for acquiring three-dimensional point cloud data in step S1 includes: using RTK real-time differential technology and laser radar acquisition technology. Through RTK real-time differential technology and laser radar acquisition technology, a centimeter-level three-dimensional point cloud model of the substation is acquired to improve the accuracy of the three-dimensional point cloud model.

[0038] Furthermore, the preprocessing in step S2 includes: denoising, filtering and segmentation. The acquired point cloud data is preprocessed to improve the quality and usability of the point cloud data, reduce the influence of noise and abnormal points on the recognition results, and improve the accuracy of recognition.

[0039] Furthermore, the device type identification of the preprocessed point cloud data in step S3 includes: an identification method based on feature extraction and machine learning. The feature information of the device, including shape, size and texture, is extracted from the processed point cloud data, and then the extracted features are trained and learned using a machine learning algorithm to build a device identification model, and the point cloud data to be identified is input into the trained model to perform device identification and classification to obtain the type information of the device. Specifically, machine learning algorithms include support vector machines, random forests and neural networks. Among them, the type information of the equipment that can be obtained includes oil level gauges, knife gates and converter valves.

[0040] Furthermore, the location information of the geometric center point in step S4 includes: longitude, latitude and altitude information. By setting the location coordinates (x_i, y_i, z_i) of the geometric center point, the center of the device (X_c, Y_c, Z_c) can be calculated by the following formula:

[0041] X_c=(Σx_i) / N

[0042] Y_c=(Σy_i) / N

[0043] Z_c=(Σz_i) / N

[0044] Where Σ is the sum, and N is the number of geometric center points, that is, the coordinates of multiple geometric center points are summed and the average is taken to obtain the center position of the device.

[0045] Furthermore, the acquisition of the shooting direction in step S6 is based on the matching device shooting direction table, wherein the shooting direction table is a database containing the correspondence between device types and shooting directions, and by querying the database, the device shooting direction information matching the recognition result can be acquired.

[0046] Furthermore, the waypoint information of the waypoint position in step S7 includes longitude, latitude, altitude, azimuth, pitch angle and waypoint type.

[0047] Furthermore, the method for calculating the waypoint position in step S7 includes shooting a direction vector, converting the direction vector to a geographic angle, calculating the coordinates of the shooting point, and calculating the altitude of the shooting point.

[0048] Specifically, the shooting direction vector is a unit vector (dx, dy, dz) in the geographic coordinate system (north, east, up); the conversion from (dx, dy, dz) to the geographic direction angle (azimuth and elevation) can be performed by the following formula:

[0049] Azimuth (Az): The horizontal angle measured clockwise from the north direction, ranging from 0° to 360°. Az = atan2(dy,dx)*(180 / π);

[0050] Elevation angle (El): The angle measured upward from the horizontal plane, ranging from -90° to 90°. El = atan2(dz, sqrt(dx2+dy2))*(180 / π);

[0051] The latitude and longitude of the device are (Lon_device, Lat_device), the altitude is Alt_device, the shooting distance is Dist (unit: meter), and the radius of the earth is R

[0052] ΔLon=(Dist / R)*(cos(El)*sin(Az))*(180 / π)

[0053] ΔLat=(Dist / R)*(cos(El)*cos(Az))*(180 / π)

[0054] Where ΔLon and ΔLat are the changes in longitude and latitude relative to the device's location.

[0055] The latitude and longitude of the photo taking point are:

[0056] Lon_camera=Lon_device+ΔLon

[0057] Lat_camera=Lat_device+ΔLat

[0058] The altitude of the photo point can be approximated as:

[0059] Alt_camera=Alt_device+Dist*sin(El)

[0060] Furthermore, the method for automatically generating a substation drone inspection route also includes obtaining the safe shooting distance information of the drone, and establishing a drone object model with the minimum safe distance as the radius. Specifically, when the simulated drone is flying along the route and detects that the object model size collides with the point cloud model, it automatically obtains the nearest point that can bypass the obstacle, marks this point as a detour point, obtains the longitude, latitude and altitude information of the detour point, and the nose direction and pitch angle inherit the previous waypoint information.

[0061] The present invention is based on equipment identification of three-dimensional point cloud models. In view of the problems that most drone inspection systems in the prior art require manual planning of flight routes, which is time-consuming and labor-intensive, and it is difficult to ensure the comprehensiveness and accuracy of inspections in complex environments, the technical solution of the present invention does not require on-site measurement of the location information of the target area and manual planning of routes. It can automatically generate routes, greatly reducing the time for route planning, while avoiding manual operation errors, reducing accident risks, enhancing safety, reducing manpower investment, and saving inspection costs.

[0062] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for automatically generating a substation drone inspection route, characterized in that: The method for automatically generating a substation drone inspection route comprises the following steps: S1: Obtain three-dimensional point cloud data of substation; S2: preprocessing the acquired substation three-dimensional point cloud data; S3: Identify the device type based on the preprocessed 3D point cloud data; S4: extracting the position of the geometric center point of the vertical and horizontal projections of the device as the position of the device according to the device recognition result; S5: marking device information on the point cloud model, including device type, device location, and device orientation; S6: acquiring a shooting direction based on the device orientation and device type information; S7: Calculate the corresponding waypoint position based on the device position, the shooting direction and the input corresponding shooting distance; S8: According to the obtained multiple waypoint positions, connect related waypoint positions in series to generate a route.

2. The method for automatically generating a substation drone inspection route according to claim 1 is characterized in that: The three-dimensional point cloud data acquisition method in step S1 includes: using RTK real-time differential technology and laser radar acquisition technology.

3. The method for automatically generating a substation drone inspection route according to claim 1 is characterized in that: The preprocessing in step S2 includes: denoising, filtering and segmentation.

4. The method for automatically generating a substation drone inspection route according to claim 1 is characterized in that: The device type identification on the preprocessed point cloud data in step S3 includes: an identification method based on feature extraction and machine learning.

5. The method for automatically generating a substation drone inspection route according to claim 1 is characterized in that: The location information of the geometric center point in step S4 includes: longitude, latitude and altitude information.

6. The method for automatically generating a substation drone inspection route according to claim 1 is characterized in that: The acquisition of the shooting direction in step S6 is based on the matching device shooting direction table.

7. The method for automatically generating a substation drone inspection route according to claim 1 is characterized in that: The waypoint information of the waypoint position in step S7 includes longitude, latitude, altitude, azimuth, pitch angle and waypoint type.

8. The method for automatically generating a substation drone inspection route according to claim 7 is characterized in that: The method for calculating the waypoint position in step S7 includes shooting a direction vector, converting the direction vector to a geographic angle, calculating the coordinates of the shooting point, and calculating the altitude of the shooting point.

9. A method for automatically generating a substation drone inspection route according to any one of claims 1 to 8, characterized in that: The method for automatically generating a substation drone inspection route also includes obtaining safe shooting distance information of the drone and establishing a drone object model with a minimum safe distance as a radius.