CN114895711A - Automatic unmanned aerial vehicle flight path line planning method for fan blade inspection - Google Patents
Automatic unmanned aerial vehicle flight path line planning method for fan blade inspection Download PDFInfo
- Publication number
- CN114895711A CN114895711A CN202210666133.7A CN202210666133A CN114895711A CN 114895711 A CN114895711 A CN 114895711A CN 202210666133 A CN202210666133 A CN 202210666133A CN 114895711 A CN114895711 A CN 114895711A
- Authority
- CN
- China
- Prior art keywords
- fan
- unmanned aerial
- flight path
- aerial vehicle
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000007689 inspection Methods 0.000 title claims abstract description 52
- 238000005457 optimization Methods 0.000 claims abstract description 45
- 238000012545 processing Methods 0.000 claims abstract description 12
- 238000005070 sampling Methods 0.000 claims abstract description 11
- 230000008569 process Effects 0.000 claims description 19
- 238000001514 detection method Methods 0.000 claims description 16
- 239000013598 vector Substances 0.000 claims description 13
- 238000003384 imaging method Methods 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 10
- 230000001154 acute effect Effects 0.000 claims description 5
- 230000009194 climbing Effects 0.000 claims description 3
- 230000006872 improvement Effects 0.000 claims description 3
- 238000000926 separation method Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000001459 lithography Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000000746 purification Methods 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000001308 synthesis method Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 238000001931 thermography Methods 0.000 description 1
- 238000011179 visual inspection Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/70—Wind energy
- Y02E10/72—Wind turbines with rotation axis in wind direction
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses an unmanned aerial vehicle automatic flight path planning method for fan blade inspection, which comprises the following steps: establishing a fan model, carrying out gridding processing on the fan model according to the routing inspection requirement to obtain a fan surface graphical model, planning a candidate course point area on the basis of each grid course of the model, ensuring that at least one complete course area can be seen from the selected course points in the course point area, and ensuring that a complete fan surface image is obtained; carrying out track point sampling based on the candidate track point area to complete track point optimization; completing the planning of the flight path line; and identifying the attitude of the fan and finishing the adjustment of the flight path line. The method generates the flight path planning based on the model, and then adjusts according to the field fan information, so that the method has low requirement on hardware, high reliability of the planned path and strong robustness; the deflection angle of the fan blade is obtained by using a machine vision method, the adjustment of flight path planning is realized, the problem of shutdown angle of the fan is not required to be concerned during routing inspection, and the automation level is high.
Description
Technical Field
The invention belongs to the technical field of unmanned aerial vehicles, and particularly relates to an automatic flight path line planning method of an unmanned aerial vehicle for fan blade inspection.
Background
Currently, many new and highly skilled detection methods are being developed for detecting faults in fan blades. Aiming at static blade detection, the method mainly comprises methods such as laser Doppler vibration measurement, X-ray imaging, ultrasonic detection, laser speckle detection, infrared thermal imaging and the like. Aiming at dynamic blade detection, the method mainly comprises optical fiber sensing, vibration monitoring, resistance strain detection, digital image processing, acoustic emission and the like. However, these methods are still in the development and trial stage, and the most applied methods for detecting the fan blade are still the traditional detection methods.
The traditional fan blade inspection modes comprise an overhead circumambulation descending visual inspection method, high power telescope detection, a blade maintenance platform and the like. The traditional blade detection method is completed manually, so that the problems of poor safety, low detection efficiency, long detection time, high detection cost and the like are faced. Therefore, a set of inspection system with higher safety and efficiency needs to be developed.
In recent years, unmanned aerial vehicle technology and image recognition technology are developed, a fan image is obtained through an unmanned aerial vehicle, then a detection method for judging whether the fan breaks down or not through the image is more and more concerned by people, and people regard detection of power equipment according to the unmanned aerial vehicle image as an unprecedented technological innovation. At present, the unmanned aerial vehicle inspection technology is applied to the inspection fields of power lines, power towers and the like, and the inspection is advancing towards the direction of fan inspection.
At the present stage, the unmanned aerial vehicle is used for carrying out the inspection process of the fan blade, at least two workers are needed to carry out the control of the unmanned aerial vehicle, and the unmanned aerial vehicle is matched with the worker through sending instructions to finally complete the inspection task. The staff controls unmanned aerial vehicle and flies near the blade, uses the imaging device that unmanned aerial vehicle carried to acquire the image on fan blade surface, then transmits image information to ground station or unmanned aerial vehicle, accomplishes once to patrol and examine and flies back to ground after, reads out the fan blade image information that unmanned aerial vehicle obtained again. Use unmanned aerial vehicle to patrol and examine and replace traditional mode of patrolling and examining, will greatly reduced cost undoubtedly, improve efficiency and the security of patrolling and examining, even in the remote mountain area or close to the marine region also can be relatively safe completion and patrol and examine the task. However, when manually controlling unmanned aerial vehicle and patrolling and examining the fan, it is higher to flying the requirement of controlling the hand, and the human factor influence is big, patrols and examines efficiency and still need further promotion, and closely patrols and examines and be difficult to realize. Therefore, in order to solve the above problems, it is necessary to perform a research on the automated inspection by the unmanned aerial vehicle. Meanwhile, the inspection efficiency and the inspection safety of the fan blade can be greatly improved by realizing automatic inspection, and the imaging quality can also be ensured.
Disclosure of Invention
In order to solve the technical problems in the prior art, the invention aims to provide an automatic flight path line planning method of an unmanned aerial vehicle for fan blade inspection.
In order to achieve the purpose and achieve the technical effect, the invention adopts the technical scheme that:
an unmanned aerial vehicle automatic trajectory planning method for fan blade inspection comprises the following steps:
firstly, establishing a fan model, carrying out gridding processing on the fan model according to the routing inspection requirement to obtain a fan surface graphical model, planning candidate track point areas on the basis of each grid patch of the fan surface graphical model, ensuring that at least one complete patch area can be seen from selected track points in the track point areas, and ensuring that a complete fan surface image is obtained; then, carrying out course point sampling based on the candidate course point area to complete course point optimization; completing the planning of the flight path line; and identifying the attitude of the fan and finishing the adjustment of the flight path line.
Further, the fan model is established, gridding processing is carried out on the fan model according to the inspection requirement, and the fan surface graphical model obtaining step comprises the following steps:
selecting a modeling tool to establish a three-dimensional fan model according to fan three-dimensional size information, and deriving a fan model from the fan model, wherein an STL-format model file takes a small triangular patch as a basic unit and describes the surface of the fan model in a discrete approximate manner, so that the requirement of representing the model by imaging can be met;
and (3) processing the meshed fan model obtained by the modeling tool again by using materialises Magics 24.0 to obtain a fan surface graphical model meeting the requirement.
Furthermore, when planning a candidate track point area, considering safety, restricting a safety range, setting maximum and minimum safety spaces, considering visibility, and needing to enable the course of the unmanned aerial vehicle and a triangular mesh surface patch to form an acute angle, even if an included angle between the course of the unmanned aerial vehicle and a normal vector of a triangular hyperplane is larger than zero, the hyperplane of the triangular surface patch is determined by an incident angle, and a restriction condition is established, so that all triangular surface patches capable of covering a fan surface graphical model are selected, a high-quality complete fan surface image is ensured to be obtained, and a polling task is ensured to be completed with the highest quality.
In consideration of visibility, an acute angle is formed between the course of the unmanned aerial vehicle and the triangular patch, even if an included angle between the course of the unmanned aerial vehicle and a normal vector of a triangular hyperplane is larger than zero, the hyperplane of the triangular patch is determined by an incident angle of the triangular patch to obtain a candidate track area, and a constraint condition is expressed as plane constraint, and is shown in a formula (1):
where g ═ x, y, z denotes the spatial position of the course point, d max To the maximum safe space, d min As a minimum safe space, x i Is the corner of the mesh triangle, a N Is the normalized triangle normal, m is the center point of the triangle patch, n i Is a normal line of a separation hyperplane for the incident angle constraint, in 1 ,∠in 2 ,∠in 3 Represents a minimum angle of incidence;
the imaging equipment carried by the unmanned aerial vehicle is installed on the unmanned aerial vehicle at a fixed pitch angle and a relative course, the field angle FOV of the imaging equipment is divided into a vertical opening and a horizontal opening, the candidate track area is limited to a triangular section by utilizing the equivalent constraint of the upper limit of the FOV, and the total sampling space generated by the opening constraint of a vertical camera carried by the unmanned aerial vehicle is the union set of all the triangular sections in all vertical directions; dividing space into N C And (3) the constraint condition is expressed as formula (2) if the blocks are equal to each other:
wherein,
respectively representing the associated angles of the triangular patches,
n right 、n left respectively, the normal vectors of the respective separated hyperplanes.
Furthermore, a corresponding fan surface graphical model is constructed according to a fan model of the wind field, a candidate track point area is obtained through the model, and when a track point is selected in the candidate track point area, iterative optimization is needed, so that the optimization of the whole track point set is realized.
During course point optimization, a course point is randomly sampled in a course point candidate area, and then the current course point g is minimized k Track point g in current candidate track point area k-1 The previous track point of the current track point
And the next track point
The sum of the squares of the distances is targeted, as shown in equation (3), the first two parts shorten the course length by moving the course point closer together, while the last two parts limit the size of the improvement step because
And
and the distance may be moved closer, and the optimization of the whole track point set is realized by gradually optimizing the selection of the track points:
the constraint condition in the formula (4) is consistent with the selected candidate track point area;
the weighting matrix for neighbor distances is:
B＝diag(b const ,b const ,a const +b const )
in the formula, b const Is a general weight of the distance to the neighbor, a const Penalizing for weight changes;
the weighting matrix of the distance of the current viewpoint in the old tour is:
D＝diag(d const ,d const ,d const )
in the formula (d) const Is a weighted value;
after the track point is selected, the current track point g is selected k The direction to the center point m of the triangular patch is the course of the track point, namely:
ψ k ＝g k -m。
further, the step of completing the trajectory planning comprises:
planning the flight path line is similar to that when the flight path point is selected, the planning of the flight path line is constrained according to task information and operation cautions of the unmanned aerial vehicle, so that the optimal flight path line is obtained, and relevant constraint conditions comprise flight path line length constraint, climbing constraint, obstacle avoidance constraint and minimum turning radius constraint;
according to the constraint conditions, the shortest flight path line is optimal when the flight path line of the unmanned aerial vehicle is ensured to finish the task, and the shorter the flight time of the unmanned aerial vehicle is, namely the flight path line planning process covering the whole target is finished; in the process of solving the optimal flight path line, an iterative re-optimization mode is selected, so that the flight path line is gradually optimized until the optimal flight path line is reached.
Further, the planned route line passes through all route points without repetition, and the initial route line is obtained by adopting the following algorithm:
(1) selecting an initial course point v i
(2) Selecting the next course point v from the unselected course points according to the following 3 constraints j ：
1) Edge e ij Is a candidate edge;
2) edge e ij Alpha of (A) π -proximity α π (e ij )＝0；
3) Edge e ij Belonging to the current optimal trajectory line;
if v is selected j If all the 3 constraints cannot be met, preferentially selecting the track points meeting the constraint 1), and if all the track points cannot be met, randomly selecting the unselected track points;
(3) let i ═ j, i and j denote the number of the track point respectively, used for marking all selected track points, if there are still unselected track lines, return to step (2);
if a plurality of track points are available for selection in the step (2), randomly selecting one point from the plurality of track points;
after planning a flight path line, iterative optimization needs to be carried out on the flight path line, in the process of iterative optimization, 5-opt operation is taken as a basic unit and is called basic operation, the basic operation selects an edge added last time in the basic operation as an edge deleted first, the edge deleted last time cannot be an edge added in the basic operation, 2-opt, 3-opt or 4-opt operation which can be carried out is searched in the basic operation for carrying out flight path line optimization, if a proper flight path point is not searched after the basic operation, the searching is stopped, the 5-opt, 4-opt and 2-opt have the same meaning, the switching operation of the edge is carried out, and only the number of the exchanged edges is increased;
starting from a certain track point, if the track line cannot be optimized through any basic operation, a discontinuous and closed 5-opt operation is carried out, and the discontinuous and closed 5-opt operation consists of a non-closed 2-opt operation and another non-closed 2-opt operation or 3-opt operation, which is called an interruption operation;
judging whether the flight path line is optimized after one-time interruption operation, if the flight path line is optimized, continuing to perform basic operation optimization on the flight path line, and then performing interruption operation, and repeating the operation until the flight path line cannot be optimized, wherein the process is called to complete one-time iterative optimization; and then, after the track point is reselected, optimizing the track line again, and repeating the two steps to complete the iterative optimization of the track line.
Further, the identification of the wind turbine attitude comprises the following steps:
unmanned aerial vehicle flies to fan wheel hub department according to the route planned, shoots the fan and conveys to the processing end and handles, confirms fan shaft tower and fan blade, discerns the feathering shut down position of fan blade, obtains the deflection angle gamma that the fan blade shut down according to the straight line slope that represents fan blade edge:
wherein,
(x 1 ,y 1 )、(x 2 ,y 2 ) Respectively representing the coordinates of points at both ends of a straight line segment of the fan blade,
representing the mean deflection angle, α i The deflection angle of each straight line representing a fan blade is indicated, and n is the number of straight lines representing a fan blade generated by image detection.
Further, the step of adjusting the trajectory line comprises:
the structure of the fan blade is relatively fixed, and the deflection angle of the fan blade is estimated and used for adjusting the air track line of the unmanned aerial vehicle routing inspection fan. In three-dimensional space, the fan blade plane of place can be regarded as the fixed plane of z value, and the fan is rotatory certain angle gamma, then track point follows rotatory gamma, can realize that the blade under the different shut down states patrols and examines, according to the expression rule of ball coordinate, track point adjustment is:
and finally, the unmanned aerial vehicle automatically patrols and examines according to the corrected path data to finish the patrolling and examining work of the blade.
Compared with the prior art, the invention has the beneficial effects that:
1) firstly, a fan model is established, surface graph processing is carried out on the model, then a candidate track point area is planned for each triangular patch, and then a non-repetitive track line passing through all track points is finally obtained through an optimization stage of alternating track point selection and track line optimization;
2) the method generates the flight path planning based on the three-dimensional fan model, and then adjusts according to the field fan information, so that the method has low requirement on hardware, high reliability of the planned path and strong robustness;
3) the deflection angle of the fan blade is obtained by using a machine vision method, so that the automatic adjustment of fan track planning is realized, the problem of the shutdown angle of the fan is not required to be concerned during routing inspection, and the automation level is greatly improved.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a diagram of a candidate track area of the present invention;
FIG. 3 is a bump dividing view of the present invention;
FIG. 4 is a flight path tree diagram according to the present invention.
Detailed Description
The present invention is described in detail below so that the advantages and features of the present invention can be more easily understood by those skilled in the art, and thus the scope of the present invention can be clearly and clearly defined.
The following presents a simplified summary of one or more aspects in order to provide a basic understanding of such aspects. This summary is not an extensive overview of all contemplated aspects, and is intended to neither identify key or critical elements of all aspects nor delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later.
Unmanned aerial vehicle's automation is patrolled and examined that general selection all is rotor aircraft that has the ability of hovering steadily, and the process of patrolling and examining is generally: the unmanned aerial vehicle flies to the vicinity of the fan according to the planned flight path line, and obtains a complete fan blade surface image at a specified position along the flight path line according to self positioning. In the process, the unmanned aerial vehicle must be able to recognize the surrounding environment and solve the problems of where the unmanned aerial vehicle is, where the target is, how to reach the target and what posture the unmanned aerial vehicle detects the target on the basis of the surrounding environment. Therefore, the completion of the planning of the trajectory line required by the unmanned aerial vehicle inspection and the realization of the fact that the unmanned aerial vehicle can know the position of the unmanned aerial vehicle are important links for realizing automatic inspection, and the unmanned aerial vehicle inspection fan blade has important significance for realizing the automatic inspection of the unmanned aerial vehicle.
Therefore, the invention discloses an automatic flight path line planning method of an unmanned aerial vehicle for fan blade inspection, which designs an optimal flight path line of the fan blade for automatic inspection of the unmanned aerial vehicle and ensures the completion of an inspection task. Firstly, a fan model is established, the fan model is subjected to gridding processing according to the routing inspection requirement to obtain a fan surface graphical model, a candidate track point area is planned on the basis of each grid surface patch of the fan surface graphical model, at least one complete surface patch area is ensured to be visible for selected track points in the track point area, and the complete fan surface is ensured to be obtained. Secondly, track point sampling is carried out based on the candidate track point area, the NP-hard problem of track line planning is solved by an iterative resampling strategy, and the optimization of the track line is completed. Thirdly, the problem of insufficient GPS positioning accuracy in the wind power plant environment is solved by adopting a binocular vision and inertia measurement positioning method for extracting point-line characteristics, and a purification strategy is improved on an algorithm for extracting line characteristics so as to facilitate fan attitude identification and track line adjustment.
As shown in fig. 1 to 4, the method for planning the automatic trajectory of the unmanned aerial vehicle for fan blade inspection provided by the invention mainly comprises the following steps:
step 1: establishing a graphical model of a fan surface
First, a three-dimensional fan model needs to be established. The modeling tool SoildWorks is selected to establish the three-dimensional fan model according to the three-dimensional fan size information, the method has the advantages of simplicity, easiness in operation and the like, and the model file in the STL (stereo lithography) format of the fan model can be directly derived from the SoildWorks. The model in the STL format takes a small triangular patch as a basic unit, the surface of the three-dimensional fan model is discretely and approximately described, the requirement of representing the model by the image science can be met, the STL format three-dimensional model can record the number of all triangular patches, the vertex coordinates of each triangle and the normal vector of each patch, and the normal vector accords with the right-hand rule. In the model file, the number of triangular patches, normal vectors of each patch, vertex coordinates of each triangle and the like of the model can be obtained;
the quantity of triangular meshes on the gridding model obtained by Soildworks is distributed too much at partial non-key monitoring positions, such as the head of a cabin, the joint of a blade and the cabin and the like, and the generated flight path lines are too dense at the positions, so that the gridding fan model is processed again by using Materialise Magics 24.0 to obtain a fan surface graphical model meeting the requirement.
Step 2: track point design based on three-dimensional model
When planning a candidate track point area, by using a generation-test method and a synthesis method, not only the sampling of track points in a certain space outside a model from the model needs to be considered, but also a constraint function for selecting the candidate area needs to be constructed, and based on a fan surface graph model and the consideration of obtaining a complete high-quality image, a track point is planned for each triangular surface patch on the fan model, so that a track point area needs to be planned for each triangular surface patch.
When planning the candidate area of the track point, the safety range needs to be restricted from the safety consideration, and the maximum safety space and the minimum safety space are d respectively max And d min In consideration of visibility, an acute angle is formed between the heading of the unmanned aerial vehicle and the triangular patch, even if an included angle between the heading of the unmanned aerial vehicle and a normal vector of a triangular hyperplane is greater than zero, and the hyperplane of the triangular patch is determined by an incident angle, a candidate track area shown in fig. 2 can be obtained comprehensively, and a constraint condition can be expressed as a plane constraint, as shown in formula (1):
where g ═ x, y, z denotes the spatial position of the course point, x i Is the corner of the mesh triangle, a N Is the normalized triangle normal, m is the center point of the triangle patch, n i Is a normal line of a separation hyperplane for the incident angle constraint, in 1 ,∠in 2 ,∠in 3 Represents a minimum angle of incidence;
from an imaging device carried by a drone, its field of view FOV is known divided into vertical and horizontal openings and mounted at a fixed pitch angle and relative headingOn the drone. With equivalent constraints from the upper FOV limit, the candidate track area is limited to triangular sections. The total sampling space generated by the vertical camera opening constraint on the drone is the union of all these triangular sections in all vertical directions, rather than being convex optimized. To approximate and highlight this problem, the space is divided into N according to FIG. 3 C Equal bumps. To approximate and highlight the problem, an optimal value is calculated for each slice in order to find a globally optimal solution. The segment j constraint is derived as follows:
the left and right boundaries of the sampling space are the boundaries of the rotation section, the rotation section is a space section cut in the direction shown in fig. 2 when approximate convex optimization is performed, the top and the bottom of the cone are represented by a single plane tangent to the center of the slice, when the slice is convex optimization, the mapping of the triangular patch in the convex optimization space is performed, the angle camera constraint in the horizontal direction can be eliminated, and d min It is sufficient to choose high enough to allow the triangle to be fully visible, so the constraint can be expressed as equation (2):
wherein,
respectively representing the associated angles of the triangular patches,
n right 、n left respectively representing normal vectors of the respective separating hyperplanes;
all the triangular patches of the coverage surface graphic model can be selected by combining the two aspects, so that a high-quality complete fan surface image can be obtained, and the inspection task can be finished with the highest quality.
A fan surface graphical model can be constructed according to a fan model of a wind field, a candidate track point area can be obtained through the model, and iterative optimization is needed when track points are selected in the candidate track point area. Firstly, selecting a first track point in a candidate track point area by a random sampling method, selecting one track point in the candidate track point area corresponding to each triangle, and simply and quickly finishing the selection of a first group of track points. Then, random sampling and QP quadratic optimization can be combined in the optimization process of the flight path point set.
During course point optimization, a course point is randomly sampled in a course point candidate area, and then the current course point g is minimized k Track point g in current candidate track point area k-1 The previous track point of the current track point
And the latter track point
The sum of the squares of the distances is the target, as shown in equation (3). The first two parts shorten the track length by moving the track point closer, while the last two parts limit the size of the improvement step because
And
may also move closer. By gradually optimizing the selection of the track points, the optimization of the whole track point set is realized, and the track points which enable the track line to be shortened are selected as much as possible.
The constraint condition in the formula (4) is consistent with the selection of the candidate track point area.
And, the weighting matrix of the neighbor distance is:
B＝diag(b const ,b const ,a const +b const )
in the formula, b const Is a general weight of the distance to the neighbor, a const Penalizes for weight change.
The weighting matrix of the distance of the current viewpoint in the old tour is as follows:
D＝diag(d const ,d const ,d const )
in the formula (d) const Is a weight value.
After the track point is selected, the current track point g is selected k The direction to the center point m of the triangular patch is the course of the track point, namely:
ψ k ＝g k -m。
and step 3: automatic track planning
Similar to the process of selecting a track point, the planning of the track line needs to be constrained according to task information and operation cautions of the unmanned aerial vehicle, so as to obtain an optimal track line, and relevant constraint conditions include: flight path length constraint, climbing constraint, obstacle avoidance constraint, minimum turning radius constraint and the like.
According to the constraint conditions, the flight path line of the unmanned aerial vehicle ensures that the shorter the flight time of the unmanned aerial vehicle is, the better the flight path line is under the condition of completing the task, namely the shortest flight path line is the optimal flight path line in the process of completing the planning of the flight path line covering the whole target. In the process of solving the optimal flight path line, an iterative re-optimization mode is selected, the flight path line is gradually optimized until the optimal flight path line is reached, and the optimization process is a process of gradually approaching the optimal solution.
The planned course line does not repeatedly pass through all course points, so the following algorithm is adopted to obtain an initial course line:
(1) selecting an initial course point v i
(2) Selecting the next course point v from the unselected course points according to the following 3 constraints j ：
1) Edge e ij Is a candidate edge;
2)α π (e ij )＝0；
3) edge e ij Belonging to the current optimal trajectory line;
if v is selected j If all the 3 constraints cannot be met, preferentially selecting the track points meeting the constraint 1), and if all the track points cannot be met, randomly selecting the unselected track points;
let G ═ V, E be the weighted graph, V ═ V 1 ，v 2 ，...，v n The points are track point sets, E is a set of connecting lines between track points, and the slave track points v i To track point v j A distance of d ij 。
The 1-tree is composed of subtrees connected by a set of points V/{1} in the graph G ═ (V, E) excluding the special points {1}, and two shortest edges connecting from the points {1} to the subtrees, as shown in fig. 4, which contains a loop, and the minimum 1-tree is the 1-tree with the smallest sum of paths through all course points.
Based on the characteristics of the minimum 1-tree obtained from the connectivity graph, the relationship of the remaining optimal air traces can be determined as follows: if the minimum 1-tree is a complete loop without "bifurcation," it is the optimal trajectory line; the optimal trajectory is a minimum 1-tree with 2 degrees per point. Also, in the study it was found that the optimal trajectory line is 70% -80% the same as the minimum 1-tree edges, so the candidate set of optimal trajectory lines can be selected based on the minimum 1-tree.
Minimum 1-track point in tree v i 、v j The distance between them is recorded as d ij ，D n×n ＝{d ij Expressing a distance matrix between the track points, and adding the same value pi according to the cost value of calculating each track point without changing the principle of optimal track selection, and converting the matrix D into a matrix D n×n Conversion into matrix P n×n ＝{p ij |p ij ＝d ij +π i +π j Using T } π Represents a minimum of 1-tree, L π (T π ) Represents according to P n×n Calculated T π Total length pi ═ pi 1 ,π 2 ,...,π n The vector for increasing the cost value of each track point is represented, and the following can be obtained:
ω(Π)＝L π (T π )-2∑π i (3-7)
the vector II can be adjusted using sub-gradient optimization to change ω (Π), thereby causing T to change ω (Π) π Closer to the optimal trajectory line
The representation includes passing through track point v i 、v j Minimum 1-tree of edge, then edge e ij Alpha of (A) π -proximity is expressed as:
wherein alpha is π (e ij ) When e is more than or equal to 0 ij When belonging to the minimum 1-tree, alpha π (e ij )＝0。
(3) And (5) if the unselected route lines exist, returning to the step (2).
If a plurality of track points are available for selection in step (2), a point is randomly selected from the plurality of track points.
After planning the trajectory, iterative optimization needs to be performed on the trajectory. In the process of iterative optimization, 5-opt is taken as a basic unit, called basic operation (basic move), the basic operation selects an edge added last time by the basic operation last time as an edge deleted first time, and the edge deleted last time cannot be an edge added once by the basic operation this time. And searching out the 2-opt, 3-opt or 4-opt which can be performed in the basic operation to perform route optimization, and stopping searching if a proper route point is not searched after one basic operation.
If the course cannot be optimized through any basic operation from a certain course point, a discontinuous, closed 5-opt operation is performed, which consists of one non-closed 2-opt operation and another non-closed 2-opt operation or 3-opt operation, and is called a break operation (break).
And judging whether the flight path line is optimized after one interruption operation, if the flight path line is optimized, continuing to perform basic operation optimization on the flight path line, and then performing interruption operation, and repeating the steps until the flight path line cannot be optimized, wherein the step of completing one iteration optimization is called to complete. And then, after the track point is reselected, optimizing the track line again, and repeating the two steps to complete the iterative optimization of the track line.
And 4, step 4: fan shutdown state identification
And 3, performing global flight path planning on the three-dimensional model by utilizing the step 3 to obtain routing inspection path data, and importing the obtained path data into a routing inspection unmanned aerial vehicle, wherein the unmanned aerial vehicle is provided with a binocular camera, automatically flies to the hub of the fan according to the planned path, and after the feathering shutdown position of the fan blade is identified by utilizing computer vision, the angle value of the unmanned aerial vehicle can be obtained according to the slope representing the edge of the fan blade, and then the angle value is compared with the fan angle of the basic model to obtain the change estimation of the shutdown angle of the fan blade, and the positive value is set as clockwise deflection, and the negative value is set as anticlockwise deflection. The method comprises the following specific steps:
(1) confirm fan shaft tower
Because the whole size of the wind driven generator is large, in the wind driven generator image obtained from the front side of the unmanned aerial vehicle, the tower can exceed the image, namely, a straight line with an end point at the lower edge of the image and a slope close to 90 degrees can be confirmed as a straight line representing the tower;
(2) fan blade confirmation
After a straight line representing a fan tower is confirmed, the other end of the straight line, far away from the edge of an image, can be approximately considered as the position of an engine room, a circle is made by taking the end point as the center of the circle, then straight lines which are about 120 degrees are selected from the circular area according to the structural characteristics of the three-blade fan, if two other straight lines which are 120 degrees with the straight lines exist in the straight line l, the straight line l can be marked as a straight line representing a fan blade, all the straight lines are traversed, the straight lines with similar deflection angles are classified into the same group, and the average angle of the straight lines is the deflection angle of the representative blade;
(3) determining blade stall angle from linear slope
After the straight line group representing the blade is determined, the deflection angle of each straight line can be calculated according to two points, and points (x) at two ends of the straight line segment are taken 1 ,y 1 )、(x 2 ,y 2 ) Then the angle of the straight line representing the blade can be obtained:
wherein,
representing the mean deflection angle, α i The deflection angle of each line is indicated and n indicates the number of fan blades per group.
Then, a blade straight line of the highest pixel point is found, and according to the relation between the inner angle and the opposite vertex angle of the right triangle, the deflection angle gamma of the shutdown of the fan blade can be obtained:
and 5: route adjustment of fan blade for inspection
The structure of the wind driven generator blade is relatively fixed, and after the deflection angle of the wind driven generator is estimated, the angle can be used for adjusting the air track line of the unmanned aerial vehicle routing inspection fan. In three-dimensional space, the fan blade plane of place can be regarded as the fixed plane of z value, and the fan is rotatory certain angle gamma, then track point follows rotatory gamma, can realize that the blade under the different shut down states patrols and examines, so according to the expression rule of ball coordinate, track point adjustment is:
according to the method, the reference path data obtained according to the fan model is converted by using the obtained deflection angle to obtain the path data capable of adapting to the current fan state, and finally, the unmanned aerial vehicle automatically patrols according to the corrected path data to finish the polling work of the blades.
The parts or structures of the invention which are not described in detail can be the same as those in the prior art or the existing products, and are not described in detail herein.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by the present specification, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. An unmanned aerial vehicle automatic trajectory planning method for fan blade inspection is characterized by comprising the following steps:
firstly, establishing a fan model, carrying out gridding processing on the fan model according to the routing inspection requirement to obtain a fan surface graphical model, planning candidate track point areas on the basis of each grid surface patch of the fan surface graphical model, ensuring that at least one complete surface patch area can be seen from selected track points in the track point areas, and ensuring that a complete fan surface image is obtained; then, carrying out course point sampling based on the candidate course point area to complete course point optimization; completing the planning of the flight path line; and identifying the attitude of the fan and finishing the adjustment of the flight path line.
2. The automatic unmanned aerial vehicle trajectory planning method for fan blade inspection according to claim 1, wherein the step of establishing a fan model, and performing gridding processing on the fan model according to inspection requirements to obtain a fan surface graphical model comprises:
selecting a modeling tool to establish a three-dimensional fan model according to fan three-dimensional size information, and deriving a fan model from the fan model, wherein an STL-format model file takes a small triangular patch as a basic unit and describes the surface of the fan model in a discrete approximate manner, so that the requirement of representing the model by imaging can be met;
and (3) processing the meshed fan model obtained by the modeling tool again by using materialises Magics 24.0 to obtain a fan surface graphical model meeting the requirement.
3. The automatic trajectory planning method for unmanned aerial vehicle for fan blade inspection according to claim 1, wherein when planning a candidate trajectory point region, a safety range is restricted from safety consideration, a maximum safety space and a minimum safety space are set, and from visibility, an acute angle is required to be formed between the heading of the unmanned aerial vehicle and a triangular mesh patch, and even if an included angle between the heading of the unmanned aerial vehicle and a normal vector of a triangular hyperplane is larger than zero, the hyperplane of the triangular patch is determined by an incident angle of the triangular patch, a restriction condition is established, so that all triangular patches capable of covering a fan surface graphics model are selected, a high-quality complete fan surface image is ensured to be obtained, and an inspection task is ensured to be completed with the highest quality.
4. The automatic trajectory planning method for the unmanned aerial vehicle for the fan blade inspection according to claim 3, wherein in view of visibility, an acute angle is formed between the course of the unmanned aerial vehicle and a triangular patch, even if an included angle between the course of the unmanned aerial vehicle and a normal vector of a hyperplane of the triangle is greater than zero, the hyperplane of the triangular patch is determined by an incident angle of the hyperplane, so as to obtain a candidate trajectory region, and the constraint condition is expressed as a plane constraint, as shown in formula (1):
where g ═ x, y, z denotes the spatial position of the course point, d max To the maximum safe space, d min As a minimum safe space, x i Is the corner of the mesh triangle, a N Is the normalized triangle normal, m is the center point of the triangle patch, n i Is a normal line of a separation hyperplane for the incident angle constraint, in 1 ,∠in 2 ,∠in 3 Represents a minimum angle of incidence;
the imaging equipment carried by the unmanned aerial vehicle is installed on the unmanned aerial vehicle at a fixed pitch angle and a relative course, the field angle FOV of the imaging equipment is divided into a vertical opening and a horizontal opening, the candidate track area is limited to a triangular section by utilizing the equivalent constraint of the upper limit of the FOV, and the total sampling space generated by the opening constraint of a vertical camera carried by the unmanned aerial vehicle is the union set of all the triangular sections in all vertical directions; dividing space into N C And (3) the constraint condition is expressed as formula (2) if the blocks are equal to each other:
wherein,
respectively representing the associated angles of the triangular patches,
n right 、n left respectively, the normal vectors of the respective separated hyperplanes.
5. The automatic unmanned aerial vehicle trajectory planning method for fan blade inspection according to claim 1, wherein a corresponding fan surface graphical model is constructed according to a fan model of a wind field, a candidate trajectory point area is obtained through the model, and when a trajectory point is selected in the candidate trajectory point area, iterative optimization is required, so that optimization of an overall trajectory point set is realized.
6. The automatic unmanned aerial vehicle trajectory planning method for fan blade inspection according to claim 5, wherein during course point optimization, a course point is randomly sampled in a candidate course point area, and then the current course point is minimizedTracing point g k Track point g in current candidate track point area k-1 The previous track point of the current track point
And the latter track point
The sum of the squares of the distances is targeted, as shown in equation (3), the first two parts shorten the course length by moving the course point closer together, while the last two parts limit the size of the improvement step because
And
and the distance may be moved closer, and the optimization of the whole track point set is realized by gradually optimizing the selection of the track points:
the constraint condition in the formula (4) is consistent with the selected candidate track point area;
the weighting matrix for neighbor distances is:
B＝diag(b const ,b const ,a const +b const )
in the formula, b const Is a general weight of the distance to the neighbor, a const Penalizing for weight changes;
the weighting matrix of the distance of the current viewpoint in the old tour is:
D＝diag(d const ,d const ,d const )
in the formula (d) const Is a weighted value;
after the track point is selected, the current track point g is selected k The direction to the center point m of the triangular patch is the course of the track point, namely:
ψ k ＝g k -m。
7. the automatic trajectory planning method for the unmanned aerial vehicle for the fan blade inspection according to claim 1, wherein the step of completing the trajectory planning comprises:
planning the flight path line is similar to that when the flight path point is selected, the planning of the flight path line is constrained according to task information and operation cautions of the unmanned aerial vehicle, so that the optimal flight path line is obtained, and relevant constraint conditions comprise flight path line length constraint, climbing constraint, obstacle avoidance constraint and minimum turning radius constraint;
according to the constraint conditions, the shortest flight path line is optimal when the flight path line of the unmanned aerial vehicle is ensured to finish the task, and the shorter the flight time of the unmanned aerial vehicle is, namely the flight path line planning process covering the whole target is finished; in the process of solving the optimal flight path line, an iterative re-optimization mode is selected, so that the flight path line is gradually optimized until the optimal flight path line is reached.
8. The automatic unmanned aerial vehicle trajectory planning method for fan blade inspection according to claim 7, wherein the planned trajectory passes through all trajectory points without repetition, and the initial trajectory is obtained by adopting the following algorithm:
(1) selecting an initial course point v i
(2) Selecting the next course point v from the unselected course points according to the following 3 constraints j ：
1) Edge e ij Is a candidate edge;
2) edge e ij Alpha of (A) π -proximity α π (e ij )＝0；
3) Edge e ij Belonging to the current optimal trajectory line;
if v is selected j Can not all beIf the 3 constraints are met, the course points meeting the constraint 1) are preferentially selected, and if the two points cannot be met, the course points which are not selected are randomly selected;
(3) let i ═ j, i and j denote the number of the track point respectively, used for marking all selected track points, if there are still unselected track lines, return to step (2);
if a plurality of track points are available for selection in the step (2), randomly selecting one point from the plurality of track points;
after planning a flight path line, iterative optimization needs to be carried out on the flight path line, in the process of iterative optimization, 5-opt operation is taken as a basic unit and is called basic operation, the basic operation selects an edge added last time in the basic operation as an edge deleted first, the edge deleted last time cannot be an edge added once in the basic operation, 2-opt, 3-opt or 4-opt operation which can be carried out is searched in the basic operation for carrying out flight path line optimization, and if a proper flight path point is not searched after one time of basic operation, the search is stopped;
starting from a certain track point, if the track line cannot be optimized through any basic operation, a discontinuous and closed 5-opt operation is carried out, and the discontinuous and closed 5-opt operation consists of a non-closed 2-opt operation and another non-closed 2-opt operation or 3-opt operation, which is called an interruption operation;
judging whether the flight path line is optimized after one-time interruption operation, if the flight path line is optimized, continuing to perform basic operation optimization on the flight path line, and then performing interruption operation, and repeating the operation until the flight path line cannot be optimized, wherein the process is called to complete one-time iterative optimization; and then, after the track point is reselected, optimizing the track line again, and repeating the two steps to complete the iterative optimization of the track line.
9. The method for automatic trajectory planning for unmanned aerial vehicle for fan blade inspection according to claim 1, wherein the identification of the fan attitude comprises the following steps:
unmanned aerial vehicle flies to fan wheel hub department according to the route planned, shoots the fan and conveys to the processing end and handles, confirms fan shaft tower and fan blade, discerns the feathering shut down position of fan blade, obtains the deflection angle gamma that the fan blade shut down according to the straight line slope that represents fan blade edge:
wherein,
(x 1 ,y 1 )、(x 2 ,y 2 ) Respectively representing the coordinates of points at both ends of a straight line segment of the fan blade,
representing the mean deflection angle, α i The deflection angle of each straight line representing a fan blade is indicated, and n is the number of straight lines representing a fan blade generated by image detection.
10. The automatic unmanned aerial vehicle trajectory planning method for fan blade inspection according to claim 1, wherein the trajectory adjusting step comprises:
the structure of the fan blade is relatively fixed, and the deflection angle of the fan blade is estimated and used for adjusting the air track line of the unmanned aerial vehicle routing inspection fan. In three-dimensional space, the fan blade plane of place can be regarded as the fixed plane of z value, and the fan is rotatory certain angle gamma, then track point follows rotatory gamma, can realize that the blade under the different shut down states patrols and examines, according to the expression rule of ball coordinate, track point adjustment is:
and finally, the unmanned aerial vehicle automatically patrols and examines according to the corrected path data to finish the patrolling and examining work of the blade.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210666133.7A CN114895711B (en) | 2022-06-14 | 2022-06-14 | Unmanned aerial vehicle automatic path planning method for fan blade inspection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210666133.7A CN114895711B (en) | 2022-06-14 | 2022-06-14 | Unmanned aerial vehicle automatic path planning method for fan blade inspection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114895711A true CN114895711A (en) | 2022-08-12 |
| CN114895711B CN114895711B (en) | 2024-07-05 |
Family
ID=82727251
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210666133.7A Active CN114895711B (en) | 2022-06-14 | 2022-06-14 | Unmanned aerial vehicle automatic path planning method for fan blade inspection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114895711B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115713645A (en) * | 2022-10-19 | 2023-02-24 | 山东大学 | Lithology identification method and system based on spectral imaging technology |
| CN116501091A (en) * | 2023-06-27 | 2023-07-28 | 珠海优特电力科技股份有限公司 | Fan inspection control method and device based on unmanned aerial vehicle automatic adjustment route |
| CN116577981A (en) * | 2023-04-12 | 2023-08-11 | 西南石油大学 | Control distribution method for omni-directional driving underwater robot propelled by magnetic coupling vector |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112082552A (en) * | 2020-07-11 | 2020-12-15 | 西北工业大学太仓长三角研究院 | Unmanned aerial vehicle flight path planning method based on improved hybrid particle swarm optimization algorithm |
| CN113885580A (en) * | 2021-11-17 | 2022-01-04 | 国能定边新能源有限公司 | Route planning method and system for realizing automatic inspection of fan based on unmanned aerial vehicle |
| CN114442665A (en) * | 2022-01-20 | 2022-05-06 | 北京华能新锐控制技术有限公司 | Wind power blade inspection line planning method based on unmanned aerial vehicle |
| CN116680958A (en) * | 2023-06-15 | 2023-09-01 | 盐城工学院 | Robot spraying track optimization method for bilateral cooperation of fan blades |
-
2022
- 2022-06-14 CN CN202210666133.7A patent/CN114895711B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112082552A (en) * | 2020-07-11 | 2020-12-15 | 西北工业大学太仓长三角研究院 | Unmanned aerial vehicle flight path planning method based on improved hybrid particle swarm optimization algorithm |
| CN113885580A (en) * | 2021-11-17 | 2022-01-04 | 国能定边新能源有限公司 | Route planning method and system for realizing automatic inspection of fan based on unmanned aerial vehicle |
| CN114442665A (en) * | 2022-01-20 | 2022-05-06 | 北京华能新锐控制技术有限公司 | Wind power blade inspection line planning method based on unmanned aerial vehicle |
| CN116680958A (en) * | 2023-06-15 | 2023-09-01 | 盐城工学院 | Robot spraying track optimization method for bilateral cooperation of fan blades |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115713645A (en) * | 2022-10-19 | 2023-02-24 | 山东大学 | Lithology identification method and system based on spectral imaging technology |
| CN116577981A (en) * | 2023-04-12 | 2023-08-11 | 西南石油大学 | Control distribution method for omni-directional driving underwater robot propelled by magnetic coupling vector |
| CN116577981B (en) * | 2023-04-12 | 2023-12-01 | 西南石油大学 | Control distribution method for omni-directional driving underwater robot propelled by magnetic coupling vector |
| CN116501091A (en) * | 2023-06-27 | 2023-07-28 | 珠海优特电力科技股份有限公司 | Fan inspection control method and device based on unmanned aerial vehicle automatic adjustment route |
| CN116501091B (en) * | 2023-06-27 | 2023-11-07 | 珠海优特电力科技股份有限公司 | Fan inspection control method and device based on unmanned aerial vehicle automatic adjustment route |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114895711B (en) | 2024-07-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114895711B (en) | Unmanned aerial vehicle automatic path planning method for fan blade inspection | |
| CN106762451B (en) | Fan blade damage detection method, device and system based on unmanned aerial vehicle | |
| CN110282143B (en) | Inspection method for offshore wind farm unmanned aerial vehicle | |
| CN110603379B (en) | Inspection tool control device for wind power equipment inspection tool | |
| CN107729808B (en) | Intelligent image acquisition system and method for unmanned aerial vehicle inspection of power transmission line | |
| CN108894933B (en) | Method and system for tracking, losing and re-capturing fan blade tips during tracking detection of fan blade tips through unmanned aerial vehicle | |
| US20160017866A1 (en) | Wind tower and wind farm inspections via unmanned aircraft systems | |
| CN112904877A (en) | Automatic fan blade inspection system and method based on unmanned aerial vehicle | |
| CN108915959B (en) | Method and system for detecting bypassing of fan blade tip area through unmanned aerial vehicle | |
| EP3850455B1 (en) | Control and navigation systems | |
| EP3850456B1 (en) | Control and navigation systems, pose optimisation, mapping, and localisation techniques | |
| CN109032175B (en) | Unmanned aerial vehicle power inspection path calculation method based on Archimedes spiral | |
| WO2020014951A1 (en) | Method and apparatus for building local obstacle map, and unmanned aerial vehicle | |
| CN108869196B (en) | Method and system for detecting rear side blade root area of fan through unmanned aerial vehicle | |
| CN110687919B (en) | Unmanned aerial vehicle autonomous flight control method for hydropower station water pipeline inspection | |
| CN113942616B (en) | Inspection mechanism and method for offshore wind farm | |
| JP2024027907A (en) | Unmanned aerial vehicle, and, inspection system of wind power generation equipment, and inspection method of wind power generation equipment | |
| CN113885580A (en) | Route planning method and system for realizing automatic inspection of fan based on unmanned aerial vehicle | |
| JP2019523363A (en) | How to record wind turbine measurements | |
| CN110104167A (en) | A kind of automation search and rescue UAV system and control method using infrared thermal imaging sensor | |
| CN116501091B (en) | Fan inspection control method and device based on unmanned aerial vehicle automatic adjustment route | |
| CN115326075A (en) | Path planning method for realizing wind field global automatic inspection based on unmanned aerial vehicle | |
| CN111984025A (en) | Unmanned aerial vehicle path planning method for wind turbine generator blades | |
| GB2581403A (en) | Pose optimisation, mapping, and localisation techniques | |
| JP2023538423A (en) | Method and system for object detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |