CN-121977566-A - Autonomous planning method for inspection waypoints of unmanned aerial vehicle of transformer substation

CN121977566ACN 121977566 ACN121977566 ACN 121977566ACN-121977566-A

Abstract

The invention discloses an autonomous planning method for a patrol navigation point of a substation unmanned aerial vehicle, which comprises the following steps of mapping three-dimensional space positions of patrol equipment; the method comprises the following steps of generating an initial routing inspection waypoint autonomously, detecting and adjusting the waypoint collision, adaptively adjusting the shooting angle, and planning the routing inspection route autonomously. The autonomous degree is high, the cost is saved, the picture shooting quality is high, and the inspection efficiency is improved. By utilizing the technologies such as the autonomous programming algorithm of the waypoints, the three-dimensional collision detection technology and the target recognition technology, the automatic and flexible programming of the unmanned aerial vehicle inspection waypoints is realized, the inspection efficiency is improved, the flight safety of the unmanned aerial vehicle is ensured, and the diversified inspection requirements of the transformer substation are met.

Inventors

WANG HAILEI
FANG YONGFENG
HUANG YAOQUN
WANG CHANGJUN
ZHANG QINGQING

Assignees

安徽优航遥感信息技术有限公司

Dates

Publication Date: 20260505
Application Date: 20260126

Claims (8)

1. The automatic planning method for the inspection waypoints of the unmanned aerial vehicle of the transformer substation is characterized by comprising the following steps of: mapping the three-dimensional space position of the inspection equipment; Step two, the initial patrol navigation points are generated autonomously; Thirdly, detecting and adjusting the collision of the waypoints; Step four, shooting angle self-adaptive adjustment; fifthly, independently planning a routing inspection route; Wherein: in the third step, a three-dimensional collision detection model needs to be constructed firstly, and the model comprises the following strategies: The navigation point data and the three-dimensional environment are respectively stored in a three-dimensional coordinate form and respectively marked as: Calculating the distance from the three-dimensional model point to the center of the navigation point by using Euclidean distance formula, ; Set the collision distance as ; The safe adjustment of the waypoint comprises the adjustment of the vertical direction and the adjustment of the horizontal direction, and is that When the vertical ascent adjustment strategy is executed, a new waypoint Is updated to be When (when) When the method is used, a horizontal direction adjustment flow is started, a rotation matrix is constructed through a Rodrigues rotation formula, and vectors are obtained Rotating Obtaining new direction vector by angle Further calculate the new waypoint coordinates 。
2. The autonomous planning method for patrol waypoints of unmanned aerial vehicle of transformer substation according to claim 1, wherein the method for mapping three-dimensional space positions of patrol equipment comprises the following steps Analyzing bin position topological relation in the electric main wiring diagram, and establishing an association index with the three-dimensional physical space through interval numbers; Positioning a device datum point in a three-dimensional coordinate system, wherein the GIS combined electrical apparatus integrated device needs to take a center point of a gas chamber as a coordinate origin, and AIS device marks a framework hanging point or a bracket base coordinate independently; The generated digital ledger comprises equipment logic bin numbers, absolute coordinates, orientation angles and adjacent equipment interval check values.
3. The autonomous planning method for the patrol waypoints of the unmanned aerial vehicle of the transformer substation according to claim 2, wherein the method for autonomously generating the initial patrol waypoints comprises the following steps of By analyzing the three-dimensional coordinates (X, Y, Z) of the selected equipment to be inspected in the standing account and the outline dimension of the equipment, the geometric center of the equipment is adopted Based on the combination of the safety distance And the equipment inspection angle is regular, and an equipment space bounding volume is constructed; For the key inspection part of the equipment, the coordinates under the local coordinate system are marked respectively After the conversion to the global coordinate system, a key inspection position set is established Wherein Forming refined inspection parameters comprising equipment space distribution, safety distance constraint and key inspection positions; According to parameters of visible light and infrared cameras carried by the inspection unmanned aerial vehicle, the parameters comprise optical center coordinates Horizontal viewing angle Vertical viewing angle Focal length Calculating an optimal observation view angle of each device by adopting a view cone collision detection algorithm, and ensuring that a key part is in a view angle range of a camera; generating waypoints capable of being patrolled and examined in all directions of the equipment, including front, rear, left, right and overlooking angles of the equipment, and determining initial waypoint coordinates of each waypoint Obtaining the initial height Initial aircraft yaw angle Thereby forming a patrol waypoint set containing initial waypoint coordinates, initial flying height, initial cradle head pitch angle and initial airplane yaw angle information, and the initial height And an initial aircraft yaw angle Calculated from the following formula: ; ;
4. the substation unmanned aerial vehicle inspection waypoint autonomous planning method according to claim 3, wherein the calculation method of the view angle range comprises the following steps Set its distance from the optical center And the equations for the six planes of the view cone are calculated by the following steps, Calculating half field angles in the horizontal direction and the vertical direction: ; left upper corner point Calculating coordinates of four corner points of the near clipping plane: ; Using three non-collinear corner points, according to the planar point French equation Three points are respectively set as And calculate the coefficients of the near clipping plane equation, and similarly calculate the equations of the other five planes of the view cone, ; For each plane of the view cone Checking eight vertices of object axis alignment bounding box Calculating the distance from the vertex to the plane If all the vertices And when all planes pass the test, the condition that all vertexes are on the invisible side of a certain plane does not exist, and the object is judged to be in the view cone.
5. The autonomous planning method for the inspection waypoints of the unmanned aerial vehicle of the transformer substation according to claim 4, wherein the vertical direction adjustment method comprises the following steps: when the navigation point A kind of electronic device The axis coordinates being greater than the collision point A kind of electronic device Axis coordinates, i.e. Executing a vertical ascent adjustment strategy; First calculate the collision distance The distance ensures that a safe buffer space exists between the adjusted navigation point and the collision point; Before the ascending operation is executed, the height feasibility is verified, and whether the ascending waypoint height exceeds the maximum allowable height of the flying or moving equipment and is collided with other obstacles above the flying or moving equipment is checked; if the conflict exists, gradually adjusting according to a preset rule until the safety condition is met; Novel waypoint Is updated to be Meanwhile, the adjustment record is stored in a system log, so that subsequent tracing and analysis are facilitated; Horizontal direction adjustment When (when) Starting a horizontal direction adjustment process, and calculating a target shooting point To the waypoint Direction vector of (a) And collision point P to waypoint Direction vector of (a) In order to ensure the accuracy of vector calculation, the input coordinates are subjected to accuracy verification, and abnormal values are removed; ; the normal vector of the plane where the two vectors are is obtained through vector cross multiplication operation ; The direction of the normal vector follows the right hand rule, and the modular length reflects the area size of the plane formed by the two vectors; ; In the normal vector In a determined plane, in navigation points Adjusting the datum point; Distance of Is usually taken The self-adaptive adjustment can be performed according to the complexity of the environment; Angle of The selection of the navigation points adopts a dynamic strategy, can be randomly selected in a half range of an included angle between zero and two vectors at the initial stage, and gradually optimizes the navigation points in a direction away from the obstacle along with the increase of iteration times; Constructing a rotation matrix by a Rodrigues rotation formula, and vector Rotating Obtaining new direction vector by angle Further calculate the new waypoint coordinates
6. The unmanned aerial vehicle routing inspection waypoint autonomous planning method of claim 5, wherein the waypoint collision detection adjustment step further comprises an iterative optimization process; after finishing one-time navigation point adjustment, immediately starting a new round of collision detection; the system traverses all the points in the three-dimensional environment model again, and calculates the distance between the system and the new navigation point according to the Euclidean distance formula; If a new collision point is detected, repeatedly executing a corresponding navigation point adjustment strategy; if no collision is detected, marking the waypoint as a safe waypoint, and incorporating the safe waypoint into a path planning sequence; In a single iteration, detection is suspended and the detected region is preferentially processed when the number of detection points reaches a threshold value and collision detection is not completed yet.
7. The autonomous planning method for the inspection waypoints of the unmanned aerial vehicle of the transformer substation according to claim 6, wherein in the field of waypoint path planning, an a-path planning algorithm is adopted: preprocessing and space modeling are carried out on initial waypoint data; Discretizing a two-dimensional or three-dimensional space into a regular grid unit by using a grid method, enabling the waypoints to correspond to grid nodes, and constructing a discretized topological space model; Or a Voronoi diagram method is adopted, space division is generated according to the distribution of the obstacles, and the inter-waypoint path is ensured to be far away from the obstacles as far as possible; mathematical modeling of various constraints, e.g. fly-height constraints Speed constraint And converting the polygonal boundary constraint of the no-fly zone into a constraint equation which can be processed by an algorithm, and integrating the constraint equation into a subsequent path planning process.
8. The substation unmanned aerial vehicle inspection waypoint autonomous programming method according to claim 7, wherein in the waypoint path planning field, a bidirectional a algorithm can be adopted: searching from the initial waypoint and the target waypoint respectively, merging paths when the two search trees meet, and combining a JPS algorithm to further improve the searching efficiency by searching jump points to skip unnecessary node evaluation, thereby being particularly suitable for a large-scale waypoint path planning scene; after the initial waypoints are led into a path planning system, setting related parameters of an A-algorithm according to scene requirements; the algorithm executes search operation in the topological space model, and continuously calculates the path cost function in the process Wherein In order to be a path distance, In order to predict the time of flight of the light, In order to be able to consume energy, Is a corresponding weight coefficient; after the search is completed, outputting the route which meets the constraint condition and has the optimal cost, and presenting the route in a series of continuous track point sequences.

Description

Autonomous planning method for inspection waypoints of unmanned aerial vehicle of transformer substation Technical Field The invention relates to the technical field of unmanned aerial vehicle control systems, in particular to an autonomous planning method for a patrol navigation point of a unmanned aerial vehicle of a transformer substation. Background At present, substation inspection work plays a vital role in guaranteeing stable operation of a power system. Traditional substation inspection is mainly dependent on manual work, and the mode is low in efficiency and has a plurality of limitations. For example, the defects such as cracks of an upper insulator, pin loss, metal corrosion at a wire connection part, loosening of a nut and the like of the upper equipment of the transformer substation are difficult to be detected in time, and hidden danger is brought to the safe operation of the transformer substation. Meanwhile, the maintenance staff climbs to the structural support for investigation, so that personal safety risks are increased, and total station equipment cannot be covered. With the development of unmanned aerial vehicle technology, the application of unmanned aerial vehicle in substation inspection is gradually increased. The unmanned aerial vehicle can be provided with visible light and infrared double-light lenses, fly to the air to be close to required inspection equipment, take pictures of the equipment from multiple angles, and can realize timely monitoring of the running condition of substation equipment by combining an image recognition technology. However, most of the conventional unmanned aerial vehicle inspection waypoint planning has defects. Part of the method realizes the inspection of the fixed equipment by manually presetting the fixed equipment route, the waypoints cannot be flexibly adjusted according to the real-time inspection requirements in the station, and the diversified inspection requirements are difficult to meet. When the inspection requirements for different equipment are met in the transformer substation, the different interval airlines are required to be executed in sequence, and photos of the required inspection equipment are selected for analysis, so that a great amount of time is consumed. According to the artificial intelligence-based building inspection unmanned aerial vehicle path planning method and system in the prior art CN202410361429.7, the efficiency and accuracy of determining the optimal flight path are improved by adopting a genetic algorithm, so that the whole building inspection efficiency is improved, and meanwhile, the encountered obstacles are identified by adopting an image analysis algorithm in the unmanned aerial vehicle flight process, so that the unmanned aerial vehicle can timely avoid the obstacles, and the situation that the unmanned aerial vehicle collides with the obstacles is prevented. According to the prior art CN202211280780.0, a substation virtual image scene is firstly obtained and established according to building information of a substation, points to be inspected are marked in the virtual image scene, then a test flight route is planned according to coordinate positions of all the points to be inspected, then the unmanned aerial vehicle is controlled to fly in the substation according to the test flight route, meanwhile whether collision occurs or not is detected, if yes, all collision points are recorded and marked in the substation virtual image scene, the test flight route is adjusted according to all the points to be inspected and the collision points, so that an actual inspection route is obtained, and otherwise, the test flight route is taken as the actual inspection route. The unmanned aerial vehicle can realize autonomous flight without collision, manual operation is not needed, and the inspection operation efficiency is improved while the flight safety of the unmanned aerial vehicle is ensured. In the prior art, CN2018101111063. X, a multi-rotor track planning system and a method for power transmission line inspection can realize the on-line real-time adjustment of the flight path of the preset flight path of an unmanned aerial vehicle. In the process of automatically generating an offline reference track and inspecting, the aircraft coordinates are positioned according to the high-precision RTK, and the inspection track is adjusted on line by combining the obstacle detection module, so that collision of the unmanned aerial vehicle is avoided. The prior art provides a planning inspection scheme through coordinates, but self-help adjustment cannot be achieved, and the prior art is especially low in applicability because the coordinates are preset. In addition, in a complex transformer substation environment, unmanned aerial vehicle flight faces the risk of collision with various devices and obstacles, and the existing waypoint planning method is not perfect in the aspect of collision detection and cannot fully guarant