CN-122002131-A - Method and system for dynamically focusing unmanned aerial vehicle inspection of fan

CN122002131ACN 122002131 ACN122002131 ACN 122002131ACN-122002131-A

Abstract

The invention discloses a method for dynamically focusing a fan unmanned aerial vehicle inspection, which comprises the steps of constructing a motion prediction model of a fan component based on fan operation parameters and unmanned aerial vehicle flight track planning, obtaining target core parameters at the next moment, wherein the target core parameters comprise fan component prediction position coordinates, linear distance between the unmanned aerial vehicle and the fan component and distance change acceleration, calculating a target focal length required by a pan-tilt camera at the next moment based on the target core parameters at the next moment, constructing a mapping relation, obtaining and adjusting motion advance of a camera focusing motor, and implementing dynamic driving adjustment of the camera focusing motor based on the motion advance and completing dynamic focusing adaptation based on encoder real-time feedback correction. Based on target motion track prediction and advanced focusing control, dynamic targets such as fan rotating blades and the like can be accurately tracked, focusing response speed is high, the 'out-of-focus' rate is reduced, and image definition in a dynamic scene is ensured.

Inventors

LI YONGJUN
Chai dongyuan

Assignees

尚特杰电力科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (9)

1. The method for dynamically focusing the unmanned aerial vehicle inspection of the fan is characterized by comprising the following steps of: Based on fan operation parameters and unmanned aerial vehicle flight path planning, a motion prediction model of a fan component is constructed, and a target core parameter at the next moment is obtained, wherein the target core parameter comprises a fan component prediction position coordinate, a linear distance between the predicted unmanned aerial vehicle and the fan component and a distance change acceleration; calculating and acquiring a target focal length required by the cradle head camera at the next moment based on the target core parameters at the next moment; Constructing a mapping relation, and acquiring and adjusting the motion advance of a camera focusing motor; Based on the motion advance, the camera focusing motor implements dynamic driving adjustment and real-time feedback correction based on the encoder to complete dynamic focusing adaptation.
2. The method of claim 1, further comprising, prior to said constructing a motion prediction model for a fan component: The method comprises the steps of collecting images of fans of different types, extracting outline features, texture features and color features of fan components, constructing a three-dimensional model and a feature database of the fan components through computer vision, wherein the outline features comprise streamline edges of the fan blades and round structures of hubs, the texture features comprise coating textures on the surfaces of the fan blades, the color features comprise specific coating colors of cabins, and the fan components comprise the blades, the hubs and the cabins.
3. The method for dynamically focusing on inspection of a fan unmanned aerial vehicle according to claim 2, wherein the constructing a motion prediction model of a fan component based on fan operation parameters and unmanned aerial vehicle flight path planning, and obtaining a target core parameter at a next moment comprises: Acquiring position information of a fan component in real time by the unmanned aerial vehicle through a mounted GPS, recording position change of the fan component in continuous frame images, and acquiring flying speed, acceleration and attitude angle of the unmanned aerial vehicle based on the IMU, and acquiring linear distance and distance change acceleration between the next moment of the unmanned aerial vehicle and the fan component; based on fan operation parameters and unmanned aerial vehicle flight path planning, a motion prediction model of a fan part is constructed, a motion trend of the fan relative to the unmanned aerial vehicle is obtained, and a motion trend of the fan part relative to the unmanned aerial vehicle is obtained; Based on Kalman filtering, unmanned aerial vehicle motion data and fan component displacement data are fused, the position of the fan component in the image at the next moment is predicted, and predicted position coordinates are output.
4. The method for dynamically focusing on inspection of a fan unmanned aerial vehicle according to claim 3, wherein the calculating, based on the target core parameter at the next moment, the target focal length required by the pan-tilt camera at the next moment comprises: Based on a pinhole imaging geometric relationship, combining the linear distance predicted at the next moment to obtain a basic focal length at the next moment of the camera, wherein a calculation formula is as follows: Wherein, the As the focal length at the current moment in time, For the predicted straight line distance at the next time, As the straight line distance at the current moment, Is a calibration coefficient; based on the position offset of the fan component in the image, position compensation is added on the basic focal length, and a compensation focal length is obtained, wherein the calculation formula of the compensation focal length is as follows: wherein, the For the center coordinates of the camera sensor, The coordinates of the fan in the image at the current moment, For the predicted coordinates of the fan in the image at the next time, The position offset between the next moment and the current moment; , ; The acceleration of the distance change is combined, the target focal length which is finally required to be adjusted is obtained, and the calculation formula of the target focal length is as follows: Wherein, the For varying acceleration by distance Judging whether the fan moves at a uniform speed or accelerates or decelerates, Is a dynamic weight factor; Dynamic weighting factor The value rule of (2) is as follows: At (constant speed) ; (Acceleration away/approach) ; At (speed-down) 。
5. The method for dynamically focusing on a fan unmanned aerial vehicle inspection according to claim 4, wherein the constructing a mapping relation and adjusting the motion advance of a camera focusing motor comprise: Based on the early calibration, constructing a mapping relation between the focal length and the step length of a camera focusing motor; Based on the mapping relation, acquiring the position of a camera focusing motor at the next moment corresponding to the target focal length; Based on the position and distance change acceleration of the camera focusing motor at the next moment, the motion advance of the camera focusing motor is calculated, wherein the motion advance comprises a start advance time and a motion speed.
6. The method of claim 5, wherein the camera focusing motor performs dynamic driving adjustment based on motion advance and completes dynamic focusing adaptation based on real-time feedback correction of an encoder, and the method comprises: Based on the motion advance, controlling a camera focusing motor to execute motion by outputting a PWM driving signal with an adjustable duty ratio; The method comprises the steps of reading the actual position of an encoder every 0.5ms, calculating deviation from a target position, fine-adjusting PWM duty ratio through a PID algorithm when the position deviation is larger than a threshold value to realize deviation compensation, keeping a current driving signal when the position deviation is smaller than the threshold value, judging that a camera focusing motor reaches the target position when the position deviation is smaller than 1/2 of the threshold value, stopping the driving signal, and completing dynamic focusing adaptation.
7. A fan unmanned aerial vehicle patrols and examines dynamic focusing system which characterized in that includes: The model construction module is used for constructing a motion prediction model of the fan component based on the fan operation parameters and the unmanned aerial vehicle flight path planning, and acquiring target core parameters at the next moment, wherein the target core parameters comprise a fan component prediction position coordinate, a linear distance between the predicted unmanned aerial vehicle and the fan component and a distance change acceleration; The calculation module is used for calculating and acquiring a target focal length required by the pan-tilt camera at the next moment based on the target core parameters at the next moment; the acquisition module is used for constructing a mapping relation and acquiring and adjusting the motion advance of the camera focusing motor; and the control module is used for implementing dynamic driving adjustment of the camera focusing motor based on the motion advance and completing dynamic focusing adaptation based on real-time feedback correction of the encoder.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor, when executing the program, implements the method for dynamically focusing a fan drone inspection as claimed in any one of claims 1 to 6.
9. A computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements a method of dynamic focus for inspection of a fan drone according to any one of claims 1-6.

Description

Method and system for dynamically focusing unmanned aerial vehicle inspection of fan Technical Field The patent application relates to the technical field of fan inspection, in particular to a method and a system for dynamically focusing fan unmanned aerial vehicle inspection. Background At present, in the field of fan inspection, unmanned aerial vehicle has widely replaced traditional manual inspection mode by virtue of advantages such as its flexibility is high, simple operation, can cover complicated operation environment. In the existing unmanned aerial vehicle inspection technology, the main stream technology is automatic focusing based on contrast detection or phase detection. The contrast detection is used for searching a focus position with highest contrast by analyzing the brightness contrast change of pixels in an image, and the phase detection is used for calculating the focus offset by the phase difference of light received by a sensor so as to realize quick focusing. The existing automatic focusing technology is better in static shooting or scenes with slower target moving speed, and is applied to part of fan unmanned aerial vehicle inspection equipment, for example, part of cameras carried by commercial inspection unmanned aerial vehicles can automatically focus and shoot the surface of a fan cabin. The prior art has the following problems that 1, the adaptability of a dynamic scene is weak, a fan continuously rotates at a certain rotating speed (the rotating speed is 10-20 revolutions per minute in general) in the running process of the fan, and an unmanned aerial vehicle needs to fly around the fan to shoot part images at different angles, so that relative motion exists between a shooting target (such as the rotating blade) and the unmanned aerial vehicle. The existing automatic focusing technology is designed aiming at static or low-speed moving targets, when the targets move dynamically, a focusing system is difficult to track the position change of the targets rapidly, the phenomenon of 'out-of-focus' is easy to occur, and the definition of images is greatly reduced. 2. The distance adaptability is insufficient, and the unmanned aerial vehicle can change along with the flight path with the distance of fan part (for example from keeping away from the cabin to being close to the blade pointed end) in the inspection process, and the focusing distance adjustment range and the response speed of current automatic focusing system are difficult to match this kind of distance dynamic variation, especially when closely shooting the blade detail, easily appear focusing hysteresis, can't in time clearly catch target feature. The existing focusing technology is not fully combined with the scene characteristics of 'dynamic target+distance change' of fan inspection, and lacks accurate focusing control logic aiming at the fan part motion rule and the association of unmanned aerial vehicle flight track and target distance, so that the focusing precision and response speed cannot meet the requirements of high precision and high efficiency of fan inspection. Therefore, we propose a method and a system for dynamically focusing the inspection of the unmanned aerial vehicle of the fan. Disclosure of Invention The invention aims to solve one of the technical problems in the related art to at least a certain extent, and therefore, the first aim of the invention is to provide a method for dynamically focusing the inspection of the unmanned fan, which is based on target motion track prediction and advanced focusing control, can accurately track dynamic targets such as rotating blades of the fan, has high focusing response speed, ensures image definition in a dynamic scene, does not need an operator to manually adjust a focal length, realizes automatic clear imaging of fan components through automatic target identification and dynamic focusing control, reduces manual intervention, shortens the inspection time of a single fan by 30% -50%, and greatly improves the inspection efficiency of a large-scale wind power plant. The second aim of the invention is to provide a fan unmanned aerial vehicle inspection dynamic focusing system. A third object of the present invention is to provide an electronic device. A fourth object of the present invention is to propose a computer readable storage medium. In order to achieve the above object, an embodiment of a first aspect of the present invention provides a method for dynamically focusing a fan unmanned aerial vehicle inspection, including the following steps: Based on fan operation parameters and unmanned aerial vehicle flight path planning, a motion prediction model of a fan component is constructed, and a target core parameter at the next moment is obtained, wherein the target core parameter comprises a fan component prediction position coordinate, a linear distance between the predicted unmanned aerial vehicle and the fan component and a distance change acceleration; calcu