CN-121141669-B - Blade dust accumulation and damage monitoring system and method based on image recognition and laser radar

CN121141669BCN 121141669 BCN121141669 BCN 121141669BCN-121141669-B

Abstract

The invention relates to the technical field of onshore wind power generation, in particular to a blade dust accumulation and damage monitoring system and method based on image recognition and a laser radar. The device comprises a shooting module, an image selection module, an image identification module, a laser radar module, a flight module, a judging module and a control module, wherein the shooting module and the laser radar module are mounted on the flight module, the control module is used for sending a first flight instruction and a shooting instruction to the flight module, the flight module flies to the positive side of a fan, after the flight module flies to the positive side of the fan, the shooting module receives the shooting instruction, shoots the fan to obtain a plurality of first images and sends the first images to the image selection module, and because the flight module flies according to the first distance, when a problem occurs to the blade, the second distance obtained by the flight module is different from the second distance obtained under the normal condition, so that whether dust accumulation or damage occurs to the blade can be judged through the difference of the second distance.

Inventors

WANG HAONAN
YANG HAO
Jian Yangyang
TANG YUAN
Lv Shuxian
TANG JINGJING

Assignees

华能吐鲁番风力发电有限公司
明阳智慧能源集团股份公司

Dates

Publication Date: 20260512
Application Date: 20251027

Claims (8)

1. The blade dust accumulation and damage monitoring system based on the image recognition and the laser radar is characterized by comprising a shooting module, an image selection module, an image recognition module, a laser radar module, a flight module, a judging module and a control module; The flying module is provided with the shooting module and the laser radar module; The control module is used for sending a first flight instruction and a shooting instruction to the flight module; after receiving a first flight instruction, the flight module flies to the positive side of the fan, and after flying to the positive side of the fan, the shooting module receives the shooting instruction, shoots the fan, obtains a plurality of first images and sends the first images to the image selection module; The image selection module is used for receiving the first images, selecting the first images with the longest blades from the first images as second images, and sending the second images to the image recognition module; the image recognition module is used for receiving the second image, judging whether dust collection or damage exists or not based on the angle relation between the blades and the tower in the second image, and sending a second flight instruction to the flight module if the dust collection or damage exists; After the flight module receives a second flight instruction and waits for the fan to stop, the flight module moves along the length direction of the blade according to a fixed first distance, and sends a first control instruction to the laser radar module, wherein the first control instruction comprises the steps of obtaining the distance between the laser radar and the blade as the first distance, obtaining the distance between the laser radar and the wheel border as the second distance, and sending the first distance and the second distance to the judging module; The judging module judges that dust accumulation or damage occurs to the blade based on the first distance and the second distance; The image selection module comprises a first model detection sub-module, a judgment value acquisition sub-module and a separation sub-module: the first model detection submodule is used for carrying out model detection on the first image to obtain a blade frame in the first image; The judging value acquisition sub-module is used for acquiring the length of the blade frame body positioned at the lowest part as a first length, sorting the first length in size, and acquiring a plurality of first lengths before sorting as a second length; Calculating an average value of the second length as a judgment value; the separation module is used for separating the first image with the first length larger than the judgment value and marking the first image as a second image.
2. The blade dust and damage monitoring system based on image recognition and laser radar of claim 1, wherein the image recognition module comprises a second model detection sub-module, an extension line acquisition sub-module and a judgment sub-module; The second model detection submodule is used for acquiring two detection frames with the largest volume in the second image, taking the two detection frames as a first frame body, and sending the first frame body to the extension line acquisition submodule; The extension line acquisition submodule is used for respectively acquiring first extension lines of a wide central line in the first frame body, taking an included angle formed by the two first extension lines as a first included angle, and sending the first included angle to the judging submodule; The judging submodule is used for judging whether the first included angle is in a preset angle range, if the first included angle is in the preset angle range, judging that the blade has no problem, and if the first included angle is not in the preset angle range, judging that the blade has dust collection or damage.
3. The blade dust and damage monitoring system based on image recognition and laser radar according to claim 1, wherein the judging module comprises a moving distance obtaining sub-module, a calculating sub-module and a confirming sub-module; the moving distance acquisition sub-module is used for acquiring the moving distance of the flight module as a third distance and sending the third distance to the calculation sub-module; the calculation submodule is used for receiving the third distance, confirming the vertical position of the flight module and the blade according to the third distance, acquiring a fourth distance based on the blade parameter and the vertical position, and adding the fourth distance to the first distance to acquire a fifth distance; calculating a sixth distance according to the triangular relation between the third distance and the fifth distance, and sending the sixth distance to the confirmation sub-module; The confirmation sub-module is used for receiving the fourth distance and judging that dust accumulation or damage occurs to the blade according to the magnitude relation between the sixth distance and the second distance.
4. A blade dust and damage monitoring system based on image recognition and lidar as recited in claim 3, wherein the blade is damaged when the sixth distance is greater than the second distance; When the sixth distance is less than the second distance, dust accumulation occurs on the blade.
5. The method for monitoring the dust and damage of the blade based on the image recognition and the laser radar is applied to the system for monitoring the dust and damage of the blade based on the image recognition and the laser radar, and is characterized by comprising the following steps: Step S1, driving an unmanned aerial vehicle to fly to the positive side of a fan, shooting the fan, and obtaining a plurality of first images; S2, selecting a first image with a plurality of blades in the longest state from a plurality of first images as a second image; Step S3, judging whether dust collection or damage exists based on the angle relation between the blades and the tower in the second image, if dust collection or damage exists, driving the unmanned aerial vehicle to move along the length direction of the blades according to a fixed first distance after the fan is stopped, and sending a first control instruction to the laser radar module, wherein the first control instruction comprises the steps of acquiring the distance between the laser radar and the blades as a first distance and acquiring the distance between the laser radar and a wheel border as a second distance; and S4, judging that dust accumulation or damage occurs to the blade based on the first distance and the second distance.
6. The method for monitoring blade dust and damage based on image recognition and lidar as recited in claim 5, wherein the step S2 comprises the steps of: performing model detection on the first image to obtain a blade frame in the first image; The length of the blade frame body positioned at the lowest part is obtained and used as a first length, the first lengths are subjected to size sorting, a plurality of first lengths before sorting are obtained and used as a second length, and the average value of the second lengths is calculated and used as a judgment value; and separating the first image with the first length larger than the judgment value and marking the first image as a second image.
7. The method for monitoring dust and damage on a blade based on image recognition and lidar according to claim 6, wherein the step of determining whether dust or damage exists based on the angular relationship between the blade and the tower in the second image in step S3 is as follows: acquiring two detection frames with the largest volume in a second image as a first frame body; Respectively acquiring first extension lines of a wide central line in a first frame body, and taking an included angle formed by the two first extension lines as a first included angle; Judging whether the first included angle is within a preset angle range, if the first included angle is within the preset angle range, judging that the blade has no problem, and if the first included angle is not within the preset angle range, judging that the blade has dust collection or damage.
8. The method for monitoring blade dust and damage based on image recognition and lidar as recited in claim 5, wherein the step of step S4 is as follows: acquiring the moving distance of the flight module as a third distance; Confirming the vertical position of the flight module and the blade according to the third distance, obtaining a fourth distance based on the blade parameter and the vertical position, and obtaining a fifth distance by adding the fourth distance and the first distance; calculating a sixth distance through a triangular relation between the third distance and the fifth distance; When the sixth distance is larger than the second distance, judging that the blade is damaged; And when the sixth distance is smaller than the second distance, judging that dust accumulation occurs on the blade.

Description

Blade dust accumulation and damage monitoring system and method based on image recognition and laser radar Technical Field The invention relates to the technical field of onshore wind power generation, in particular to a blade dust accumulation and damage monitoring system and method based on image recognition and a laser radar. Background With the rapid development of the renewable energy field, land wind power generation is an important clean energy form, and has been applied on a large scale nationwide, for example, in the turkey area, where there is a bai wind area and wind resources are abundant, and in these areas, wind power generation has become a main means for obtaining electric energy. However, the failure problem of the wind generating set severely restricts the continuous stable operation and the power generation efficiency, not only causes the loss of the generated energy, but also increases the maintenance cost. Under the background, the development of an efficient and accurate fault diagnosis technology is particularly critical to improving the operation reliability and economy of the wind turbine. For example, in areas with special geographical and climatic conditions such as turbinate, the day and night temperature differences are different due to the huge day and night temperature differences and the unique basin topography effect, dew is easily condensed on the blades due to the lower night temperature, and the environmental wind is less and larger, so that a large amount of dust is easily accumulated on the blades after long-time work. The surface smoothness of the dust-deposited blade is damaged due to the change of the surface morphology, so that the airflow cannot cling to the surface of the blade. The resulting captured wind energy is less and the rotor speed decreases, resulting in a reduced power generation. In addition, due to erosion of wind and sand, part of blades can be damaged, and at the moment, the phenomenon of rotor rotation speed reduction and power generation capacity reduction can also occur. Therefore, how to effectively identify blade dust collection and damage monitoring becomes a problem to be solved in onshore wind power generation. Disclosure of Invention In view of the above drawbacks, the present invention is directed to a system and a method for monitoring dust and damage of a blade based on image recognition and lidar. Whether dust accumulation or damage occurs on the blade is effectively distinguished. The blade dust accumulation and damage monitoring system based on the image recognition and the laser radar comprises a shooting module, an image selection module, an image recognition module, a laser radar module, a flight module, a judging module and a control module; The flying module is provided with the shooting module and the laser radar module; The control module is used for sending a first flight instruction and a shooting instruction to the flight module; after receiving a first flight instruction, the flight module flies to the positive side of the fan, and after flying to the positive side of the fan, the shooting module receives the shooting instruction, shoots the fan, obtains a plurality of first images and sends the first images to the image selection module; The image selection module is used for receiving the first images, selecting the first images with the longest blades from the first images as second images, and sending the second images to the image recognition module; the image recognition module is used for receiving the second image, judging whether dust collection or damage exists or not based on the angle relation between the blades and the tower in the second image, and sending a second flight instruction to the flight module if the dust collection or damage exists; After the flight module receives a second flight instruction and waits for the fan to stop, the flight module moves along the length direction of the blade according to a fixed first distance, and sends a first control instruction to the laser radar module, wherein the first control instruction comprises the steps of obtaining the distance between the laser radar and the blade as the first distance, obtaining the distance between the laser radar and the wheel border as the second distance, and sending the first distance and the second distance to the judging module; the judging module judges that dust accumulation or damage occurs to the blade based on the first distance and the second distance. Preferably, the image selecting module includes a first model detecting sub-module, a judgment value obtaining sub-module, and a separating sub-module: the first model detection submodule is used for carrying out model detection on the first image to obtain a blade frame in the first image; The judging value acquisition sub-module is used for acquiring the length of the blade frame body positioned at the lowest part as a first length, sorting the first length in size, and acquir