A Method for Measuring the Clearance of Wind Turbine Blades by UAV Machine Vision

Abstract

The present invention relates to the technical field of wind turbine blade clearance monitoring, and provides a method for measuring the clearance of wind turbine blades by using unmanned aerial vehicle (UAV) machine vision. The UAV carrying the machine vision system flies to a suitable position on the side of the nacelle, aligns the picture and then stays; the machine vision system records videos to obtain image data, and performs real-time clearance calculation according to preset parameters and the identification results of the blade and the tower barrel; within a continuous time t, if the proportion of the duration during which the machine vision measurement optical axis is perpendicular to the nacelle symmetry plane or the perpendicular deviation angle is within the perpendicular deviation angle threshold range exceeds 80%, the measurement of the blade clearance is completed. This method measures the clearance of wind turbine blades based on UAV machine vision, and determines whether the machine vision measurement optical axis is perpendicular to the nacelle symmetry plane during the measurement process, ensuring the accuracy and effectiveness of the clearance measurement results.

Description

Technical Field

[0001] The present invention relates to the technical field of clearance monitoring for wind turbine blades, and particularly to a method for measuring the clearance of wind turbine blades by using unmanned aerial vehicle (UAV) machine vision. Background Art

[0002] The blades of wind power industry units are developing towards the trends of being longer, more flexible and lighter. With the development of onshore wind turbines, the proportion of wind farm construction under conditions such as high temperature and low temperature, high altitude, high humidity and complex terrain that exceed the standard design environment of the original wind turbines is increasing day by day, and the blade clearance problem has become increasingly severe. The blade clearance actually refers to the minimum clearance between the blade tip and the tower surface. The GL certification specification requires that during the operation of the unit, this minimum clearance shall not be less than 30% of the undeformed state of the blade. Blade clearance is very important in the development and certification stage of wind turbines, and is related to the stable and safe operation of the unit. Therefore, the detection and monitoring of blade clearance have become more and more important. At present, the technical routes for measuring clearance based on machine vision mainly include the nacelle clearance measurement scheme and the ground clearance measurement scheme. The nacelle clearance measurement scheme is to install clearance equipment in the nacelle and calculate the clearance result through machine vision ranging technology. The ground clearance measurement scheme is generally to take a video of the impeller rotation perpendicular to the axis of the nacelle at a distance of several hundred meters from the tower and calculate the clearance.

[0003] At present, there are also some methods that use UAVs to take pictures of the operation of the wind turbine rotor and then calculate the clearance. For example, Chinese Patent CN116950850A, a wind turbine blade clearance detection device and method based on UAV. In the method for measuring blade clearance based on UAV, the UAV can measure the blade clearance at the same height as the blade tip, with fewer influencing factors, which is convenient for accurately capturing the moment when the blade tip is at the lowest point, and can effectively eliminate the elevation deviation caused by ground clearance measurement. The blade clearance measurement is more accurate. However, in the scheme of measuring the blade clearance with the UAV at the same height as the blade tip, it is necessary to solve the technical problem that the optical axis of machine vision measurement is perpendicular to the symmetry plane of the nacelle. Otherwise, when the wind turbine yaws or the blade deflects due to the action of wind force, once the measurement optical axis is not perpendicular to the left-right symmetry plane of the nacelle, it will cause a large deviation in the measured clearance, seriously affecting the effectiveness and accuracy of the measurement. Summary of the Invention

[0004] In order to solve the above-mentioned deficiencies in the prior art, the present invention provides a method for measuring the clearance of wind turbine blades by using UAV machine vision. This method measures the clearance of wind turbine blades based on UAV machine vision and determines whether the optical axis of machine vision measurement is perpendicular to the symmetry plane of the nacelle during the measurement process to ensure the accuracy and effectiveness of the clearance measurement result.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] The present invention provides a method for a drone to measure the clearance of a wind turbine blade using machine vision, comprising the following steps:

[0007] S1. The drone carrying the machine vision system flies to a suitable position on the side of the nacelle, aligns the picture, and then stays. The requirements for picture alignment are as follows:

[0008] a. The optical axis of the machine vision measurement is kept horizontal;

[0009] b. When the tip of the blade rotates to the lowest point, the tower barrel and the blade should appear simultaneously in the picture taken by the machine vision system, and the optical axis of the machine vision measurement directly faces the tip of the blade;

[0010] S2. The machine vision system records videos to obtain image data, and performs real-time clearance calculation based on preset parameters and the recognition results of the blade and the tower barrel;

[0011] S3. Determine whether the clearance shooting duration and conditions in step S2 meet the requirements. If they meet the requirements, end the video recording and complete the measurement of the blade clearance; if they do not meet the requirements, go to step S4;

[0012] The clearance shooting duration and conditions are as follows: within a continuous time t, the proportion of the duration during which the optical axis of the machine vision measurement is perpendicular to the symmetry plane of the nacelle or the perpendicular deviation angle is within the perpendicular deviation angle threshold range exceeds 80%;

[0013] S4. Control the drone to rotate in the direction opposite to the perpendicular deviation direction, with the rotation angle being the same as the perpendicular deviation angle value, and then repeat steps S2 and S3. If the requirements are still not met, end the clearance measurement and determine that this clearance measurement is invalid.

[0014] The determination steps for the optical axis of the machine vision measurement in step S3 to be perpendicular to the symmetry plane of the nacelle or the perpendicular deviation angle being within the perpendicular deviation angle threshold range are as follows:

[0015] S31. Based on the characteristics of the perpendicular deviation angle between the optical axis of the machine vision measurement and the symmetry plane of the nacelle and the arc drawn by the tip of the blade, simulate the three-dimensional model of the arc drawn by the tip of the blade at different perpendicular deviation angles through 3D software;

[0016] S32. Set the perpendicular deviation angle threshold α 0 , establish the three-dimensional model of the arc drawn by the tip of the blade under this perpendicular deviation angle threshold α 0 , and based on this model, export the width of the arc drawn by the tip of the blade at a specified height h 0 below the lowest point of the tip of the blade d 0 ;

[0017] S33. Extract the video data of the machine vision system in step S2 and obtain the image data of the actual trajectory of the blade tip, so as to obtain the specified height from the lowest point of the blade tip h 0 of the actual arc width of the blade tip d s , if d s < d 0 , it is determined that the machine vision measurement optical axis is perpendicular to the nacelle symmetry plane or the perpendicular deviation angle is within the perpendicular deviation angle threshold range.

[0018] The variables required for simulation in step S31 include blade length, tower width, tower height, nacelle and impeller dimension information, elevation cone angle, perpendicular deviation angle between the machine vision measurement optical axis and the nacelle symmetry plane, and the distance between the machine vision measurement optical axis and the nacelle symmetry plane.

[0019] Extract the video data of the machine vision system in step S2 and obtain the image data of the actual trajectory of the blade tip. The specified height from the lowest point of the blade tip is h 0 the actual arc width of the blade tip d s , based on the 3D software simulation, establish the 3D model of the blade tip arc at different perpendicular deviation angles, and export the specified height from the lowest point of the blade tip as h 0 the arc width of the blade tip when d s corresponding to the actual perpendicular deviation angle α s , which is the calculated perpendicular deviation angle.

[0020] Compared with the prior art, the beneficial effects of the present invention are: (1) In the present invention, the clearance of the wind turbine blade is measured based on the unmanned aerial vehicle machine vision, and the perpendicularity of the machine vision measurement optical axis and the nacelle symmetry plane during the measurement process is determined to ensure the accuracy and effectiveness of the clearance measurement result.

[0021] (2) In the present invention, through the arc characteristics of the blade tip when the machine vision measurement optical axis deviates perpendicularly from the nacelle symmetry plane, the 3D model of the blade tip arc at different perpendicular deviation angles is simulated by 3D software, and based on this model, the specified height from the lowest point of the blade tip α 0 when the set perpendicular deviation angle threshold h 0 is reached, the arc width of the blade tip d 0, so as to quickly determine whether the optical axis of the machine vision measurement is perpendicular to the symmetry plane of the nacelle or whether the vertical deviation angle is within the range of the vertical deviation angle threshold, and based on the simulation model, quickly calculate and adjust the shooting angle of the UAV, and effectively suppress the influence of yaw or shooting angle deviation on the measurement clearance in a timely manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 Schematic diagram of the stopping position of the UAV and the captured image when the UAV of the present invention measures the clearance of the wind turbine blade by machine vision;

[0023] Figure 2 Schematic diagram of the trajectory of the blade tip when the optical axis of the machine vision measurement forms different angles with the symmetry plane of the nacelle in the present invention;

[0024] Figure 3 Schematic diagram of the model of the trajectory of the blade tip established in the present invention;

[0025] Figure 4 Vertical deviation angle threshold in the present invention α 0 Three-dimensional simulation model of the blade tip arc drawing under;

[0026] Figure 5 Vertical deviation angle in the case study α is 0°, schematic diagram of the trajectory of the blade tip;

[0027] Figure 6 Vertical deviation angle in the case study α is 1.6°, schematic diagram of the trajectory of the blade tip;

[0028] Figure 7 Vertical deviation angle in the case study α is 3.2°, schematic diagram of the trajectory of the blade tip. DETAILED DESCRIPTION OF THE INVENTION

[0029] The present invention will be further described below with reference to the drawings and embodiments.

[0030] The present invention provides a method for a UAV to measure the clearance of a wind turbine blade by machine vision. In this embodiment, when the machine vision system captures the clearance of the blade, the UAV carries the machine vision system and flies to a suitable position on the side of the nacelle, stops after aligning the picture; the requirements for picture alignment are:

[0031] a. The optical axis of the machine vision measurement is kept horizontal;

[0032] b. When the blade tip rotates to the lowest point, both the tower barrel and the blade appear in the captured image of the machine vision system, and the optical axis of the machine vision measurement directly faces the blade tip, that is, the blade tip is placed at or near the middle line position of the picture, as Figure 1 shown.

[0033] After the machine vision system is aligned, image data is acquired through video recording, and real-time clearance calculation is performed based on the preset parameters and the identification results of the blade and tower barrel.

[0034] To avoid large deviations in the measured clearance and affect the effectiveness and accuracy of the clearance measurement value, during the clearance measurement process of this embodiment, it is necessary to determine the perpendicularity of the machine vision measurement optical axis and the symmetry plane of the nacelle. By observing the trajectory diagram of the blade tip when the machine vision measurement optical axis and the symmetry plane of the nacelle form different angles through 3D software simulation, as Figure 2 shown, it can be clearly seen that according to the established shooting method, if the machine vision measurement optical axis is not perpendicular to the symmetry plane of the nacelle or the unit yaws, it will cause the blade tip to draw an arc effect, and the greater the perpendicular deviation angle, the more pronounced the blade tip arc effect. This can be used as an important reference basis for determining the perpendicularity of the machine vision measurement optical axis and the symmetry plane of the nacelle.

[0035] Based on the characteristics of the perpendicular deviation angle between the machine vision measurement optical axis and the symmetry plane of the nacelle and the blade tip arc, a blade tip trajectory model can be established, as Figure 3 shown. In the figure, P is the position of the machine vision measurement optical axis on the unmanned aerial vehicle, A is the midpoint of the impeller rotation, O is the lowest point of the blade tip rotation (the origin of the coordinate system), the X-axis is the direction perpendicular to the symmetry plane of the nacelle, the Y-axis is the direction parallel to the symmetry plane of the nacelle, the plane XOZ is the blade rotation plane, the machine vision measurement optical axis P is located in the XOY plane, the plane YOZ is the symmetry plane of the nacelle, G is the projection of P on the Y-axis, H is the projection of P on the X-axis, D is the truncation height point, DO is the truncation height (i.e., the height from the lowest point of the blade tip h ), B and C are the horizontal intersection points of D in the blade tip trajectory, that is, BC is the blade tip arc width at the height h from the lowest point of the blade tip d , and the angle ∠HOP in the figure is the perpendicular deviation angle between the machine vision measurement optical axis and the symmetry plane of the nacelle α。 From Figure 2 and Figure 3 it can be seen that when the height h from the lowest point of the blade tip is certain, the greater the perpendicular deviation angle α , the greater the blade tip arc width d . Based on this, a perpendicular deviation angle threshold α 0 can be set, and then a three-dimensional simulation model of the blade tip arc under this perpendicular deviation angle threshold α 0 is established, and the blade tip arc width h 0 at a specified height from the lowest point of the blade tip is derived based on this model d 0 , as Figure 4As shown; then obtain the specified height from the lowest point of the leaf tip from the actually captured trajectory image of the leaf tip h 0 of the actual arc width of the leaf tip below d s , if d s < d 0 , it is determined that the machine vision measurement optical axis is perpendicular to the symmetry plane of the nacelle or the vertical deviation angle is within the vertical deviation angle threshold range.

[0036] The variables required for the above simulation include blade length, tower width, tower height, nacelle and impeller size information, elevation cone angle, vertical deviation angle between the machine vision measurement optical axis and the symmetry plane of the nacelle, and the distance between the machine vision measurement optical axis and the symmetry plane of the nacelle.

[0037] Based on the above analysis, the following method for the unmanned aerial vehicle (UAV) machine vision to measure the clearance of the wind turbine blade is formed, as Figure 5 shown, including the following steps:

[0038] S1. The UAV carries the machine vision system and flies to a suitable position on the side of the nacelle, stops after aligning the picture; the requirements for picture alignment are:

[0039] a. The machine vision measurement optical axis remains horizontal;

[0040] b. When the leaf tip rotates to the lowest point, both the tower and the blade appear in the picture captured by the machine vision system, and the machine vision measurement optical axis directly faces the leaf tip;

[0041] S2. The machine vision system records video to obtain image data, and performs real-time clearance calculation according to the preset parameters and the recognition results of the blade and tower;

[0042] S3. Judge whether the clearance shooting duration and conditions in step S2 meet the requirements. If they meet the requirements, end the video recording and complete the measurement of the blade clearance; if they do not meet the requirements, enter step S4;

[0043] The clearance shooting duration and conditions are: within the continuous time t, the proportion of the duration during which the machine vision measurement optical axis is perpendicular to the symmetry plane of the nacelle or the vertical deviation angle is within the vertical deviation angle threshold range exceeds 80%; specifically, the continuous time t is determined according to the clearance shooting duration requirement, and can be selected as 20 s in this embodiment. If the above clearance shooting duration and conditions are met, the effectiveness and accuracy of the measured clearance value can be guaranteed.

[0044] Specifically, the determination steps for the machine vision measurement optical axis to be perpendicular to the symmetry plane of the nacelle or the vertical deviation angle within the vertical deviation angle threshold range are:

[0045] S31. Based on the characteristics of measuring the vertical deviation angle between the optical axis of machine vision and the symmetry plane of the nacelle and the arc drawn by the blade tip, simulate the three-dimensional model of the arc drawn by the blade tip at different vertical deviation angles through 3D software; the variables required for simulation include blade length, tower width, tower height, nacelle and impeller size information, elevation cone angle, the vertical deviation angle between the optical axis of machine vision and the symmetry plane of the nacelle, and the distance between the optical axis of machine vision and the symmetry plane of the nacelle.

[0046] S32. Set the vertical deviation angle threshold α 0 , and establish the three-dimensional model of the arc drawn by the blade tip under this vertical deviation angle threshold α 0 , and based on this model, export the width of the arc drawn by the blade tip at a specified height from the lowest point of the blade tip h 0 ; in this embodiment, the vertical deviation angle threshold is set to 5°. d 0

[0047] S33. Extract the video data of the machine vision system in step S2 and obtain the image data of the actual trajectory of the blade tip, so as to obtain the actual width of the arc drawn by the blade tip at a specified height h 0 from the lowest point of the blade tip d s . If d s < d 0 , it is determined that the optical axis of machine vision is perpendicular to the symmetry plane of the nacelle or the vertical deviation angle is within the vertical deviation angle threshold range.

[0048] S4. Control the UAV to rotate in the direction opposite to the vertical deviation direction, and the rotation angle is the same as the vertical deviation angle value. Then repeat steps S2 and S3. If the requirements are still not met, end the clearance measurement and determine that this clearance measurement is invalid;

[0049] The calculation method of the vertical deviation angle is as follows: extract the video data of the machine vision system in step S2 and obtain the image data of the actual trajectory of the blade tip, and obtain the actual width of the arc drawn by the blade tip when the specified height from the lowest point of the blade tip is h 0 ; based on the three-dimensional model of the arc drawn by the blade tip simulated at different vertical deviation angles through 3D software, export the width of the arc drawn by the blade tip when the specified height from the lowest point of the blade tip is d s ; the actual vertical deviation angle corresponding to h 0 when the width of the arc drawn by the blade tip is d s α s, which is the calculated vertical deviation angle.

[0050] In this embodiment, the vertical deviation angle between the established machine vision measurement optical axis and the symmetry plane of the nacelle α and the specified height from the lowest point of the blade tip h The arc width of the lower blade tip d The relationship model between them was demonstrated by examples, and the arc width of the blade tip during the continuous yawing process of the unit was measured on site d The blade length of this unit is 81m, the measurement distance is 120m, the rated speed is 10rpm, the yaw rate is 0.4deg / s, and the specified height from the lowest point of the blade tip here h =1.23m, the arc width of the blade tip d The results are as Figures 5 - 7 and the following table shows, Figure 5 The vertical deviation angle in α is 0°, and the arc width of the blade tip d =0; Figure 6 The vertical deviation angle in α is 1.6°, and the arc width of the blade tip d =0.75m; Figure 7 The vertical deviation angle in α is 3.2°, and the arc width of the blade tip d =1.68.

[0051] The simulation data in the above table is calculated according to the relationship model between the vertical deviation angle between the established machine vision measurement optical axis and the symmetry plane of the nacelle α and the specified height from the lowest point of the blade tip h The arc width of the lower blade tip d The deviation between the simulation model and the actual data is basically controlled within 8%, and the simulation model is relatively reliable.

Claims

1. A method for measuring the clearance of a wind turbine blade by using machine vision of an unmanned aerial vehicle, characterized in that It includes the following steps: S1. The drone carries the machine vision system and flies to an appropriate position on the side of the nacelle. After aligning the picture, it stays. The requirements for picture alignment are: a. The optical axis of the machine vision measurement remains horizontal; b. When the tip of the blade rotates to the lowest point, both the tower barrel and the blade appear in the picture taken by the machine vision system, and the optical axis of the machine vision measurement directly faces the tip of the blade; S2. The machine vision system records videos to obtain image data, and performs real-time clearance calculation based on the preset parameters and the recognition results of the blade and the tower barrel; S3. Determine whether the clearance shooting duration and conditions in step S2 meet the requirements. If they meet the requirements, end the video recording and complete the measurement of the blade clearance. If they do not meet the requirements, go to step S4; The clearance shooting duration and conditions are: within a continuous time t, the proportion of the duration during which the optical axis of the machine vision measurement is perpendicular to the symmetry plane of the nacelle or the vertical deviation angle is within the vertical deviation angle threshold range exceeds 80%; The determination steps for the optical axis of the machine vision measurement to be perpendicular to the symmetry plane of the nacelle or the vertical deviation angle to be within the vertical deviation angle threshold range are: S31. Based on the characteristics of the vertical deviation angle between the optical axis of the machine vision measurement and the symmetry plane of the nacelle and the arc drawn by the tip of the blade, simulate the three-dimensional model of the arc drawn by the tip of the blade at different vertical deviation angles through 3D software; S32. Set the vertical deviation angle threshold α 0 , and establish the three-dimensional model of the tip arc drawing under this vertical deviation angle threshold α 0 , and derive the width of the tip arc drawing at a specified height from the lowest point of the tip based on this model h 0 ; d 0 ; S33, extract the video data of the machine vision system in step S2 and obtain the image data of the actual trajectory of the blade tip, so as to obtain the specified height from the lowest point of the blade tip h 0 The actual blade tip arc width d s ,like d s < d 0 , it is determined that the machine vision measurement optical axis is perpendicular to the cabin symmetry plane or the vertical deviation angle is within the vertical deviation angle threshold range; S4. Control the drone to rotate in the direction opposite to the vertical deviation direction, and the rotation angle is the same as the vertical deviation angle value. Then repeat steps S2 and S3. If the requirements are still not met, end the clearance measurement and determine that this clearance measurement is invalid.

2. The method for measuring the clearance of a wind turbine blade by using unmanned aerial vehicle machine vision according to claim 1, wherein: The variables required for simulation in step S31 include the blade length, tower barrel width, tower barrel height, nacelle and impeller dimension information, elevation cone angle, vertical deviation angle between the optical axis of the machine vision measurement and the symmetry plane of the nacelle, and the distance between the optical axis of the machine vision measurement and the symmetry plane of the nacelle.

3. The method for measuring the clearance of a wind turbine blade by using a drone's machine vision according to claim 1, characterized in that: The calculation method of the vertical deviation angle in step S4 is as follows: extract the video data of the machine vision system in step S2 and obtain the image data of the actual trajectory of the blade tip, and obtain the specified height from the lowest point of the blade tip as h 0 The actual blade tip arc width d s Based on the 3D software simulation of the blade tip arc 3D model at different vertical deviation angles, the specified height from the lowest point of the blade tip is derived as h 0 When the blade tip arc width is d s The corresponding actual vertical deviation angle α s , which is the calculated vertical deviation angle.

Images