An online evaluation method for wind turbine tower service performance based on machine vision

Abstract

The invention relates to an online evaluation method for the service performance of a wind turbine tower based on machine vision, comprising the following steps: S1, equipment installation: setting a tower acceleration detection device, a load monitoring device and a control processing system, and installing a visual monitoring device on a tower segment; S2, initial calibration: determining the corresponding relationship between the physical size and the pixel size of an end identification object and an intermediate identification object when the tower is not deformed; S3, online evaluation: when the tower acceleration is less than a set value, a static data analysis method is used to determine the physical distance of the feature points of the intermediate identification object during sampling, and calculate the deformation of the tower segment; when the tower acceleration is greater than or equal to the set value, a dynamic data analysis method is used to determine the displacement of the end identification object, calculate the actual relative displacement of the two ends of the tower segment, and calculate the theoretical displacement based on the consideration of only the linear segment; and the actual displacement is compared with the theoretical displacement to determine whether the tower has entered a nonlinear state.

Description

Technical Field

[0001] The present invention relates to the technical field of offshore wind power generation, and in particular to an online evaluation method for the service performance of a wind turbine tower based on machine vision. Background Art

[0002] As the global energy structure shifts toward low-carbonization and the energy consumption structure continues to optimize, the trend of continued growth in demand for renewable energy is certain. With its outstanding advantages such as abundant resources, environmental protection, high degree of automation in operation and management, and continuously reducing costs, wind energy has become one of the most widely developed and applied renewable energy sources. It is an important component of global renewable energy development and utilization, and its development is gradually shifting from supplementary energy to alternative energy.

[0003] However, the service environment of wind turbines is usually very harsh. Wind turbines usually serve in the field and are affected by extreme weather, climate change and natural disasters, such as lightning strikes, storms, freezing, earthquakes, etc. These factors may damage wind turbine equipment or affect its normal operation. In addition, since wind turbines are usually installed in remote areas, maintenance and overhaul are difficult and require professional technical and equipment support. At the same time, due to the large height and size of wind turbines, maintenance operations require special equipment and measures. With the continuous growth of global wind power installed capacity, the operation and maintenance of wind turbines has become a problem that needs to be solved urgently.

[0004] Machine vision technology is a technology that uses computers, image processing, artificial intelligence and other technologies to simulate human visual functions and realize various applications. Machine vision systems can quickly obtain a large amount of information, process complex data, and accurately detect and identify targets. They have been increasingly widely used in various fields.

[0005] Wind turbine tower deformation monitoring is an important part of wind power generation projects, because the deformation of wind turbine towers may affect the normal operation and safety of wind turbines. Traditional wind turbine tower deformation monitoring methods mainly rely on manual inspections and regular maintenance, which is not only time-consuming and labor-intensive, but may also fail to detect potential safety hazards in a timely manner.

[0006] Therefore, how to realize intelligent online evaluation of the service performance of wind turbine tower is an urgent problem to be solved. Summary of the invention

[0007] In view of the above-mentioned shortcomings of the prior art, the technical problem to be solved by the present invention is to provide an online evaluation method for the service performance of a wind turbine tower based on machine vision, which can achieve non-contact, non-destructive, fast and accurate measurement and evaluation.

[0008] To achieve the above object, the present invention provides a method for online evaluation of wind turbine tower service performance based on machine vision, comprising the following steps:

[0009] S1. Equipment installation, including:

[0010] S11. Set up a tower acceleration detection device, a load monitoring device and a control processing system; the tower acceleration detection device can detect the acceleration of the tower when it shakes, and communicate with the control processing system; the load monitoring device can obtain the external load information of the wind turbine when it is running, and communicate with the control processing system;

[0011] S12. Installation of visual monitoring device: the tower includes multiple tower segments, and a visual monitoring device is installed on the tower segments. The visual monitoring device includes a camera, an end recognition object and an intermediate recognition object. A camera is arranged at at least one end of the tower segment; multiple height installation positions at different heights are selected on the outer wall or inner wall of the tower segment, and multiple marking points are selected along the circumferential direction for each height installation position. An intermediate recognition object is fixedly arranged at each marking point, and the end recognition object and the intermediate recognition object are both within the shooting range of the camera; the control processing system is connected to the camera for communication, and can receive and process the image data shot by the camera. The control processing system can also control the adjustment focus of the camera;

[0012] S2. Initial calibration: In the natural state of the tower without deformation, the camera is used to shoot, and the collected image is processed. The end identification object and the middle identification object in the image are used as feature points to determine the pixel coordinates and pixel distances of the feature points. According to the actual physical coordinates and physical distances of the feature points, the corresponding relationship between the physical coordinates and the pixel coordinates, and the corresponding relationship between the physical distance and the pixel distance are determined;

[0013] S3. Online evaluation: During the operation of the wind turbine, select the data analysis method based on the tower acceleration response obtained by the tower acceleration detection device:

[0014] S31. When the tower acceleration is less than the set value, a static data analysis method is used, including the following steps:

[0015] S311, the camera takes pictures and samples according to the set sampling frequency f1, processes the collected images, determines the pixel distance of each intermediate recognition object feature point in the image, and determines the physical distance of the intermediate recognition object feature point during sampling according to the correspondence between the pixel distance and the physical distance of the intermediate recognition object feature point obtained by the initial calibration in step S2; calculates the deformation of the tower segment;

[0016] S32, when the tower acceleration is greater than or equal to the set value, a dynamic data analysis method is used, including the following steps:

[0017] S321, the control processing system adjusts the focus of the camera to focus on the position of the end recognition object;

[0018] S322, the camera takes pictures and samples according to the set sampling frequency f2, processes the collected images, determines the pixel coordinates of the feature points of the end recognition object in the image, and determines the displacement of the end recognition object according to the correspondence between the pixel coordinates of the feature points of the end recognition object obtained by the initial calibration in step S2 and the physical coordinates, and calculates the actual relative displacement of the two ends of the tower segment on the horizontal plane (D x ,D y );

[0019] S323, based on the external load information obtained by the load monitoring device, combined with the geometric information and material information of the tower, the relative theoretical displacement (d x ,d y ); compared with the actual displacement (D x ,D y ) and theoretical displacement (d x ,d y ) to determine whether the tower has entered a nonlinear state.

[0020] Furthermore, in the step S12 of installing the visual monitoring device, a camera is respectively arranged at the top and the bottom of the tower segment, and the two cameras are simultaneously used as end recognition objects, and the middle recognition object and the camera are both arranged inside the tower segment; in the step S32, the two cameras independently perform shooting work, and independently calculate the displacement data corresponding to each other, and then take the average value as the actual displacement (D x ,D y ).

[0021] Furthermore, in the installation of the visual monitoring device in step S12, the marking points on each installation height are evenly distributed along the circumference of the tower, and the straight-line distance between adjacent marking points at the same installation height is 50 cm.

[0022] Furthermore, in the installation of the visual monitoring device in step S12, the height difference of the height installation positions on the tower segment is 2m.

[0023] Furthermore, in the step S1 of monitoring the installation of the equipment, the intermediate identified object is a screw, and the surface of the screw is coated with colored paint.

[0024] Further, in step S311, the sampling frequency f1 is once per hour.

[0025] Furthermore, in step S31, when the deformation of the tower exceeds 5%, it is determined that the service performance of the wind turbine tower is abnormal.

[0026] Furthermore, in step S322, the sampling frequency F1 in the dynamic data analysis is 20 Hz.

[0027] Further, in step S323, when the actual displacement (D x ,D y ) exceeds the theoretical displacement (d x ,d y ) is 5%, judging that the tower has entered a nonlinear state.

[0028] As described above, the online evaluation method for the service performance of a wind turbine tower based on machine vision according to the present invention has the following beneficial effects:

[0029] Machine vision technology is used to quickly obtain a large amount of information, and the displacement data of multiple characteristic points inside the tower are collected at the same time, so as to identify the long-term deformation of the tower and provide a basis for the evaluation of the long-term service performance of the tower. In addition, for extreme environmental loads, machine vision technology is used to identify the deformation of each tower segment, and by comparing it with the numerical analysis results that only consider the linear stage of the material, it is determined whether the tower has entered the plastic extreme under extreme conditions, thereby providing a reference for the operation and maintenance decision of the wind turbine; monitoring the wind turbine tower in an automated, real-time and high-precision manner can achieve non-contact, non-destructive, fast and accurate measurement and evaluation, and improve the operating efficiency and safety of the wind farm. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic flow chart of the online evaluation method of the service performance of a wind turbine tower according to the present invention.

[0031] Figure 2 It is a schematic diagram of the installation of the visual monitoring device on the tower segment in the present invention.

[0032] Figure 3 It is a schematic diagram of the distribution of intermediate identification objects on the tower segment in the present invention.

[0033] Figure 4 It is a flowchart of the static data analysis method in the present invention.

[0034] Figure 5 It is a schematic diagram of the flow of the dynamic data analysis method in the present invention.

[0035] Description of Figure Numbers

[0036] 1 Tower

[0037] 11 Tower Segments

[0038] 2 Cameras

[0039] 3 Intermediate object recognition DETAILED DESCRIPTION

[0040] The following is a description of the implementation of the present invention by means of specific embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification.

[0041] It should be noted that the structures, proportions, sizes, etc. drawn in the drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the limiting conditions for the implementation of the present invention, so they have no substantial technical significance. Any modification of the structure, change of the proportional relationship or adjustment of the size should still fall within the scope of the technical content disclosed by the present invention without affecting the effects and purposes that can be achieved by the present invention. At the same time, the terms such as "upper", "lower", "left", "right", "middle" and so on quoted in this specification are only for the convenience of description, and are not used to limit the scope of the implementation of the present invention. The change or adjustment of their relative relationship should also be regarded as the scope of the implementation of the present invention without substantially changing the technical content.

[0042] See also Figures 1 to 5 The present invention provides a method for online evaluation of wind turbine tower service performance based on machine vision, comprising the following steps:

[0043] S1. Equipment installation, including:

[0044] S11. Set up the tower acceleration detection device, load monitoring device and control processing system. The tower acceleration detection device can detect the acceleration of the tower when it is shaking, and communicate with the control processing system. The acceleration detection mechanism can use an acceleration sensor, which is preferably set at the top of the tower, and can measure the shaking of the tower well. The acceleration sensor can use the sensor provided by the wind turbine. The load monitoring device can obtain the external load information of the wind turbine during operation, including wind load, seismic load, etc., and communicate with the control processing system to transmit the obtained external load information to the control processing system. The load monitoring device can use existing measuring instruments.

[0045] S12. Installation of visual monitoring device: see Figure 2 and Figure 3The tower 1 includes multiple tower segments 11, and a visual monitoring device is installed on the tower segments 11. Specifically, the visual monitoring device can be installed on part or all of the tower segments 11 as needed; the visual monitoring device includes a camera 2, an end identification object and an intermediate identification object 3. A camera is arranged at at least one end of the tower segment 11. Specifically, one or more cameras can be arranged at the top or bottom of the tower segment 11, and one or more cameras can also be arranged at the top or bottom at the same time; multiple height installation positions at different heights are selected on the outer wall or inner wall of the tower segment 11, and multiple marking points are selected along the circumferential direction for each height installation position, and an intermediate identification object is fixedly arranged at each marking point, and the end identification object and the intermediate identification object 3 are both within the shooting range of the camera; the control processing system is connected to the camera for communication, and can receive and process the image data captured by the camera, and the control processing system can also control the adjustment focus of the camera.

[0046] In this embodiment, see Figure 2 and Figure 3 As a preferred design, camera 2 uses a high-resolution HD camera to ensure the shooting effect. A camera 2 is respectively set at the top and bottom of the tower segment 11. Camera 2 is set at the middle position of the top and the bottom. The two cameras 2 are used as end identification objects at the same time, and there is no need to set other objects as end identification objects. The two cameras 2 work independently and can shoot better. In the subsequent processing and calculation process, the calculation results obtained by the two cameras 2 can be averaged to make the results more accurate. After the camera 2 is installed, it needs to be protected, especially the camera 2 at the bottom, to avoid damage to the camera 2 when the subsequent staff performs tower 1 maintenance.

[0047] In this embodiment, see Figure 2 and Figure 3 As a preferred design, the tower segment 11 selects multiple height installation positions from bottom to top, and the height difference between adjacent height installation positions is 2m. The number of marking points on each height installation position is the same, and they are evenly distributed along the circumference of the tower 1, and the straight-line distance between adjacent marking points in the same height installation position is 50cm. The marking points in the same direction on each height installation position are located on the same straight line, that is, from a top-down or top-down perspective, the marking points on all height installation positions are arranged in a radial shape, see Figure 3 Of course, the marking points of the adjacent height installation positions can also be arranged in a staggered manner. In this embodiment, the intermediate identification object 3 is a screw, and the surface of the screw is coated with bright color paint to facilitate identification in the captured image. The screw is fixedly connected to the inner wall of the tower 1 by welding.

[0048] S2. Initial calibration: In the natural state of the tower without deformation, the camera is used to shoot, the collected images are processed, and the end identification objects and middle identification objects in the image are used as feature points to determine the pixel coordinates and pixel distances of the feature points. According to the actual physical coordinates and physical distances of the feature points, the corresponding relationship between the physical coordinates and the pixel coordinates, as well as the corresponding relationship between the physical distance and the pixel distance, are determined.

[0049] Specifically, after the camera shoots, the collected image is preprocessed, including denoising, enhancement and other operations to improve the image quality. The image data is transmitted to the control processing system, which automatically processes the image. The end recognition objects and the middle recognition objects are extracted from the preprocessed image as feature points, and the pixel coordinates and pixel distances of the feature points are obtained. The actual physical coordinates and physical distances of the feature points such as the end recognition objects and the middle recognition objects are obtained by actual measurement. In this way, the correspondence between the physical coordinates and the pixel coordinates of the feature points such as the end recognition objects and the middle recognition objects, as well as the correspondence between the physical distance and the pixel distance, can be established, and the calibration is completed accordingly.

[0050] Over a long period of time, the visual monitoring device needs to be calibrated regularly to correct the correspondence between physical size and pixel size to ensure the reliability of machine vision recognition results.

[0051] S3. Online evaluation: During the operation of the wind turbine, select the data analysis method based on the tower acceleration response obtained by the tower acceleration detection device:

[0052] S31. When the tower acceleration is less than the set value, it indicates that the service environment is normal and the tower shakes slightly. The static data analysis method is adopted, including the following steps:

[0053] S311, the camera takes pictures and samples according to the set sampling frequency f1, and the sampling time interval can be selected to a relatively large appropriate value according to actual needs to perform static collection;

[0054] The collected images are processed to determine the pixel size of each feature point of the intermediate identified object in the image, and the physical distance of the intermediate identified object during sampling is determined based on the correspondence between the pixel distance and physical distance of the feature points obtained in the initial calibration S2; the deformation of the tower segment is calculated.

[0055] In this embodiment, the sampling frequency f1 is preferably once per hour. The image is preprocessed first, including denoising, enhancement and other operations to improve the image quality, and then formally processed. The control processing system identifies the intermediate identification object in the drawing as a feature point, determines the pixel coordinates and pixel distance of the feature point, and then combines the correspondence between the pixel distance and the physical distance of the feature point of the intermediate identification object obtained by the initial calibration in step S2, so as to convert the physical distance of the intermediate identification object when the image is sampled, and compare it with the physical distance of the intermediate identification object during the initial calibration to determine the deformation. Specifically, by detecting the change in the spacing between the intermediate identification objects in the same height installation position, the radial deformation of the tower segment at the height can be determined. For the change in the spacing between the intermediate identification objects in the installation positions of different heights, the deformation of the tower segment in the height direction can be detected. At the same time, the deformation of each tower segment can be evaluated through the deformation of the entire tower.

[0056] S312. When the calculated deformation degree of the tower exceeds 5% of the natural state, it indicates that the service performance of the wind turbine is abnormal, and the staff is reminded to perform maintenance in time.

[0057] Under normal service environment, static data analysis is used to collect images and detect calculations statically to simplify operations for online evaluation of long-term service performance.

[0058] S32. When the tower acceleration is greater than or equal to the set value, it means that the service environment is extreme and the tower shakes violently. At this time, a dynamic data analysis method is used, including the following steps:

[0059] S321. The control processing system adjusts the focus of the camera so that it focuses on the position of the end recognition object. Specifically in this embodiment, the upper and lower cameras simultaneously serve as the end recognition objects in the images taken by each other, so the focus of the upper and lower cameras is adjusted to focus on each other.

[0060] S322, the camera takes pictures and samples according to the set sampling frequency f2, processes the collected images, determines the pixel coordinates of the feature points of the end recognition object in the image, and determines the displacement of the end recognition object according to the correspondence between the pixel coordinates of the feature points of the end recognition object obtained by the initial calibration in step S2 and the physical coordinates, and calculates the actual relative displacement of the two ends of the tower segment on the horizontal plane (D x ,D y ), where D x and D y They are displacements in the x-direction and y-direction respectively. The x-direction and y-direction are perpendicular to each other. The x-direction can be selected in the east-west direction, and the y-direction can be selected in the north-south direction.

[0061] Since frequent sampling and detection are required in extreme service environments, a large sampling frequency f2 is required, preferably 20Hz. Each frame of the image is pre-processed, including denoising, enhancement and other operations, to improve the image quality before formal processing. In this embodiment, two upper and lower cameras are used. The two cameras shoot independently and independently calculate the corresponding displacement data. The image processing method of the two cameras is the same, and then the average value is taken as the actual displacement (Dx, Dy).

[0062] S323, based on the external load information obtained by the load monitoring device, combined with the geometric information and material information of the tower, the relative theoretical displacement (d x ,d y ), d x and d y are the displacements in the x and y directions respectively; compared with the actual displacement (D x ,D y ) and theoretical displacement (d x ,d y ) to determine whether the tower has entered a nonlinear state.

[0063] Among them, the theoretical relative displacement (d x ,d y ), is the numerical analysis result considering only the linear stage. According to the geometric information and material information of the tower, the linear geometric model of the tower is established according to the finite element software, the finite element mesh is divided, the model constraints are added, and the external load information (including wind load, seismic load, etc.) obtained by the load monitoring device is combined to solve and calculate, and finally the tower response under the external load is obtained, and the displacement difference (d x ,d y ), which is an existing finite element linear solution method, and will not be described in detail. The linear stage analysis calculation method can be automatically performed by a calculation program built into the control system.

[0064] In this embodiment, preferably, when the actual displacement (D x ,D y ) exceeds the theoretical displacement (d x ,d y ) is 5%, that is, the actual displacement (D x ,D y ) and theoretical displacement (d x ,d y) exceeds 5% of the theoretical displacement, it is judged that the wind turbine tower has entered a nonlinear state and there are certain safety hazards. After the extreme event, the wind turbine equipment should be maintained to ensure its normal operation.

[0065] In extreme service environments, dynamic data analysis, dynamic image collection and detection calculation are used to timely reflect the deformation of the tower, thereby realizing online evaluation of the service performance in extreme service environments.

[0066] The online evaluation method of wind turbine tower service performance based on machine vision of the present invention has the following beneficial effects:

[0067] Machine vision technology is used to quickly obtain a large amount of information, and the displacement data of multiple characteristic points inside the tower are collected at the same time, so as to identify the long-term deformation of the tower and provide a basis for the evaluation of the long-term service performance of the tower. In addition, for extreme environmental loads, machine vision technology is used to identify the deformation of each tower segment, and by comparing it with the numerical analysis results that only consider the linear stage of the material, it is determined whether the tower has entered the plastic extreme under extreme conditions, thereby providing a reference for the operation and maintenance decision of the wind turbine; monitoring the wind turbine tower in an automated, real-time and high-precision manner can achieve non-contact, non-destructive, fast and accurate measurement and evaluation, and improve the operating efficiency and safety of the wind farm.

[0068] In summary, the present invention effectively overcomes various shortcomings in the prior art and has high industrial utilization value.

[0069] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A method for online evaluation of wind turbine tower service performance based on machine vision, characterized in that: The following steps are involved: S1. Equipment installation, including: S11. Set up a tower acceleration detection device, a load monitoring device and a control processing system; the tower acceleration detection device can detect the acceleration of the tower when it shakes, and communicate with the control processing system; the load monitoring device can obtain the external load information of the wind turbine when it is running, and communicate with the control processing system; S12. Installation of visual monitoring device: the tower includes multiple tower segments, and a visual monitoring device is installed on the tower segments. The visual monitoring device includes a camera, an end recognition object and an intermediate recognition object. A camera is arranged at at least one end of the tower segment; multiple height installation positions at different heights are selected on the outer wall or inner wall of the tower segment, and multiple marking points are selected along the circumferential direction for each height installation position. An intermediate recognition object is fixedly arranged at each marking point, and the end recognition object and the intermediate recognition object are both within the shooting range of the camera; the control processing system is connected to the camera for communication, and can receive and process the image data shot by the camera. The control processing system can also control the adjustment focus of the camera; S2. Initial calibration: In the natural state of the tower without deformation, the camera is used to shoot, and the collected image is processed. The end identification object and the middle identification object in the image are used as feature points to determine the pixel coordinates and pixel distances of the feature points. According to the actual physical coordinates and physical distances of the feature points, the corresponding relationship between the physical coordinates and the pixel coordinates, and the corresponding relationship between the physical distance and the pixel distance are determined; S3. Online evaluation: During the operation of the wind turbine, select the data analysis method based on the tower acceleration response obtained by the tower acceleration detection device: S31. When the tower acceleration is less than the set value, a static data analysis method is used, including the following steps: S311, the camera takes pictures and samples according to the set sampling frequency f1, processes the collected images, determines the pixel distance of each feature point of the intermediate identified object in the image, and determines the physical distance of the intermediate identified object during sampling according to the correspondence between the pixel distance and the physical distance of the feature point of the intermediate identified object obtained by the initial calibration in step S2; calculates the deformation of the tower segment; S32, when the tower acceleration is greater than or equal to the set value, a dynamic data analysis method is used, including the following steps: S321, the control processing system adjusts the focus of the camera to focus on the position of the end recognition object; S322, the camera takes pictures and samples according to the set sampling frequency f2, processes the collected images, determines the pixel coordinates of the feature points of the end recognition object in the image, and determines the displacement of the end recognition object according to the correspondence between the pixel coordinates of the feature points of the end recognition object obtained by the initial calibration in step S2 and the physical coordinates, and calculates the actual relative displacement of the two ends of the tower segment on the horizontal plane (D x ,D y ); S323, based on the external load information obtained by the load monitoring device, combined with the geometric information and material information of the tower, the relative theoretical displacement (d x ,d y ); compared with the actual displacement (D x ,D y ) and theoretical displacement (d x ,d y ) to determine whether the tower has entered a nonlinear state.

2. The online evaluation method for wind turbine tower service performance based on machine vision according to claim 1 is characterized in that: In step S12, during the installation of the visual monitoring device, a camera is provided at the top and bottom of the tower segment, respectively. The two cameras are used as end recognition objects at the same time, and the middle recognition object and the camera are both provided inside the tower segment. In step S32, the two cameras independently perform shooting work, and independently calculate the displacement data corresponding to each other, and then take the average value as the actual displacement (D x ,D y ).

3. The online evaluation method for wind turbine tower service performance based on machine vision according to claim 2 is characterized in that: In the installation of the visual monitoring device in step S12, the marking points on each height installation position are evenly distributed along the circumference of the tower, and the straight-line distance between adjacent marking points at the same height installation position is 50 cm.

4. The online evaluation method for wind turbine tower service performance based on machine vision according to claim 1 is characterized in that: In step S12 of installing the visual monitoring device, the height difference of the height installation positions on the tower segment is 2m.

5. The online evaluation method for wind turbine tower service performance based on machine vision according to claim 1 is characterized in that: In the step S1 monitoring device installation, the intermediate identified object is a screw, and the surface of the screw is coated with color paint.

6. The online evaluation method for wind turbine tower service performance based on machine vision according to claim 1 is characterized in that: In step S311 , the sampling frequency f1 is once per hour.

7. The online evaluation method for wind turbine tower service performance based on machine vision according to claim 1 is characterized in that: In step S31, when the deformation of the tower exceeds 5%, it is determined that the service performance of the wind turbine tower is abnormal.

8. The online evaluation method for wind turbine tower service performance based on machine vision according to claim 1 is characterized in that: In step S322, the sampling frequency F1 in the dynamic data analysis is 20 Hz.

9. The online evaluation method for wind turbine tower service performance based on machine vision according to claim 1 is characterized in that: In step S323, when the actual displacement (D x ,D y ) exceeds the theoretical displacement (d x ,d y ) is 5%, judging that the tower has entered a nonlinear state.

Images