Cabin displacement monitoring method based on cabin video and dynamic tilt angle sensor

Abstract

The application discloses a kind of cabin displacement monitoring methods based on cabin video and dynamic inclination sensor, including real-time shooting ground picture using camera installed in cabin front, real-time collection inclination using two-axis dynamic inclination sensor installed in cabin interior;Static data calibration is carried out when wind turbine is stationary;Running data calculation is carried out when wind turbine is running;Cabin displacement solution.The cabin displacement monitoring method based on cabin video and dynamic inclination sensor provided by the application realizes the measurement of cabin absolute displacement by machine vision technology and inertial sensor, and gives clear analysis and formula, can accurately measure cabin plane displacement, for relevant analysis and early warning protection.

Description

Technical Field

[0001] This invention relates to the field of wind turbine safe operation monitoring technology, and in particular to a method for monitoring nacelle displacement based on nacelle video and dynamic tilt sensor. Background Technology

[0002] Measuring the displacement of the nacelle in wind turbines is of great value and significance for the safe operation and life management of wind turbine generator sets.

[0003] By monitoring and recording the displacement of wind turbine nacelles in real time, abnormal displacement caused by extreme wind speeds, structural fatigue, foundation settlement, and other factors can be detected promptly, preventing safety hazards such as excessive tower bending and torsion, and avoiding serious safety accidents. Nacelle displacement data helps assess the structural health of wind turbines, particularly the degree of fatigue damage at the tower and flange connections. By comparing design allowable values ​​with actual displacement trends, potential structural failure risks can be identified early. Long-term stable nacelle displacement monitoring data can be incorporated into the remaining life assessment model of wind turbines to predict future structural deterioration trends, providing a scientific basis for rationally scheduling maintenance, repairs, and equipment replacement. In the event of a fault, such as excessive tower vibration or tower resonance, nacelle displacement data is a crucial source of diagnostic information, helping engineers quickly locate the cause of the fault, shorten repair time, and reduce downtime losses.

[0004] Current research on nacelle displacement monitoring solutions mainly focuses on the following implementation methods: 1. Deploying complex remote measurement equipment on the ground to measure nacelle displacement. This method is inconvenient to install and implement, and yaw has a significant impact, requiring manual intervention and adjustment. 2. Measuring and calculating turbine displacement using sensors such as tilt angle or accelerometers combined with a turbine dynamics model. This method heavily relies on the simulation accuracy of the dynamics model, resulting in discrepancies with actual displacement measurements. 3. Using camera equipment to photograph pre-set markers at the bottom of the wind turbine tower and monitoring nacelle displacement through image processing. However, relying solely on images for calculation is prone to errors and cannot yield reliable results. 4. Suspending a reference object, similar to a steel cable, from the nacelle to the bottom of the tower, and then using tilt angle sensors at the bottom of the reference object to monitor the data in real time and further calculate nacelle displacement. However, this method is difficult to install and is currently limited to theoretical research, making it difficult to implement in engineering. Furthermore, the presence of inertia can introduce delays and errors into the system.

[0005] Therefore, how to monitor the displacement of wind turbine nacelles safely, accurately, and at low cost is a technical problem that urgently needs to be solved. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a cabin displacement monitoring method based on cabin video and dynamic tilt sensors. By using machine vision technology and inertial sensors, the absolute displacement of the cabin is measured, and clear analysis and formulas are provided. This method can accurately measure the planar displacement of the cabin for related analysis and early warning protection.

[0007] The embodiments of the present invention provide the following solutions:

[0008] This invention provides a method for monitoring cabin displacement based on cabin video and a dynamic tilt sensor, the method comprising:

[0009] Step 1: Use the camera installed at the front of the cabin to capture real-time ground images, and use the two-axis dynamic tilt sensor installed inside the cabin to collect the tilt angle in real time.

[0010] Step 2: Perform static data calibration while the wind turbine is stationary;

[0011] Step 3: Perform operational data calculations while the wind turbine is running;

[0012] Step 4: Calculate the displacement of the engine room.

[0013] In one optional embodiment, the parameters for static data calibration in step two include: the initial pitch angle of the cabin and the initial calibration values ​​of the coordinates of the equivalent point of the reference point at the bottom of the tower.

[0014] In an optional embodiment, the equivalent point of the tower base reference mentioned in step two is obtained through the following process:

[0015] S2.1, Image Acquisition: Capture an image containing the reference object using a camera;

[0016] S2.2 Region Selection: Preprocess the image to determine the region of interest;

[0017] S2.3 Edge Recognition: Applying edge detection methods to identify the contour edges of reference objects within the region of interest;

[0018] S2.4 Reference object contour point picking: For the identified edges, extract the contour points containing all pixel coordinate information of the object edges;

[0019] S2.5 Curve Fitting: Perform shape analysis on the extracted contour point set, perform shape fitting, and obtain the fitted curve;

[0020] S2.6, Locating the equivalent point: Select a point on the fitted curve as the equivalent point of the reference object at the bottom of the tower.

[0021] In an optional embodiment, the preprocessing described in step S2.2 includes forming a region of interest through threshold segmentation, edge detection, and identification of specific color and / or texture features.

[0022] In an alternative embodiment, the edge recognition in step S2.3 includes identifying the contour edges of a reference object within the region of interest using an edge detection method.

[0023] In an optional embodiment, the reference object contour point picking in step S2.4 is achieved using a contour detection method.

[0024] In an optional embodiment, the curve fitting in step S2.5 uses the least squares method to fit the tower profile into an elliptical curve.

[0025] In an optional embodiment, step S2.6 selects the point on the fitted curve that is farthest from the tower as the equivalent point of the tower bottom reference.

[0026] In one optional embodiment, step three, the calculation of operational data, includes analyzing images captured in real time during wind turbine operation to calculate the forward and backward direction pixels of the equivalent point of the tower base reference, as well as the pitch angle during nacelle operation.

[0027] In an optional embodiment, step four uses the following formula to calculate the cabin displacement:

[0028] d i =-(y i -y0)·rh·(θ i -θ0)

[0029] Where, d i Indicates cabin displacement, y i y0 represents the initial calibration value of the equivalent point of the reference object at the bottom of the tower during operation, r represents the spatial distance represented by a unit pixel in the image, h represents the camera installation height, and θ represents the coordinates of the equivalent point of the reference object at the bottom of the tower. i θ0 represents the pitch angle during cabin operation, and θ0 represents the initial calibration value of the cabin pitch angle.

[0030] The beneficial effects of this invention based on its technical solution are as follows:

[0031] This invention provides a cabin displacement monitoring method based on cabin video and a dynamic tilt sensor. The method calculates the cabin's horizontal displacement, including both forward / backward and left / right displacements, using cabin video image technology, cabin tilt data, and cabin motion relationships. The reference point can be a clearly visible ground feature, as long as it is within the field of view at each yaw position. Measuring absolute cabin displacement using cabin video and a dynamic tilt sensor achieves low-cost, high-precision measurements. With appropriate selection of a high-resolution camera, displacement measurement accuracy can reach the centimeter level. Furthermore, it avoids the influence of yaw position and the development difficulties associated with complex algorithms. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 This is a diagram showing the deployment of monitoring equipment.

[0034] Figure 2 This is a diagram showing the cabin moved backward.

[0035] Figure 3 This is a schematic diagram of the unit's rotation.

[0036] Figure 4 This is an image of the unit when it is shut down.

[0037] Figure 5 This is a schematic diagram of contour recognition.

[0038] Figure 6 This is a schematic diagram of ellipse fitting during calibration.

[0039] Figure 7 This is a schematic diagram of elliptic fitting during unit operation.

[0040] In the diagram: 1-cabin, 2-camera, 3-two-axis dynamic tilt sensor. Detailed Implementation

[0041] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the protection scope of the embodiments of the present invention.

[0042] This invention provides a method for monitoring cabin displacement based on cabin video and a dynamic tilt sensor, the method comprising:

[0043] Step 1, refer to Figure 1 The system uses a camera installed at the front of the cabin to capture real-time images of the ground, and a two-axis dynamic tilt sensor installed inside the cabin to collect real-time tilt angles. The two-axis dynamic tilt sensor must keep its axis parallel to the front and rear and left and right sides of the cabin.

[0044] During operation, wind turbines will shift backward due to wind thrust, accompanied by changes in the nacelle pitch angle. A schematic diagram of the nacelle shift is shown below. Figure 2 As shown, the dashed line represents the original static state of the nacelle, and the solid line represents the displacement of the nacelle after being loaded. At this time, the change in the nacelle's rotation angle is ignored. OC is the tower. From this schematic diagram, it can be seen that the reference point P moves forward a distance of d in the image.

[0045] A schematic diagram of the cabin rotation is shown below. Figure 3 As shown, the dashed line represents the original static state of the nacelle, and the solid line represents the displacement of the nacelle after being loaded. At this time, the change in the translation of the nacelle is ignored. OC is the tower. At this time, the rotation angle of the nacelle is θ, and the installation height of the camera is h. It can be seen that the distance that the reference point P moves backward is h*θ.

[0046] Therefore, the displacement calculation formula for cabin movement is as follows (taking the displacement in the forward and backward direction as an example):

[0047]

[0048] In the formula:

[0049] d g : The distance the camera's viewpoint at the base of the tower changes from its previous position to its current position, in meters;

[0050] y i : Equivalent point coordinates of the reference object at the bottom of the tower, in pixels;

[0051] y0: Initial calibration value of the equivalent point coordinates of the reference object at the bottom of the tower, in pixels;

[0052] r: Spatial distance represented by a unit pixel in an image, in m / pixel;

[0053] d i Forward and backward displacement of the cabin, in meters;

[0054] h: Camera installation height, in meters;

[0055] θ: The change in pitch angle during cabin operation, in rad;

[0056] θ i : Pitch angle value during cabin operation, in rad;

[0057] θ0: Initial calibration value of cabin pitch angle, in rad.

[0058] Therefore, by obtaining the pitch angle and key coordinate values ​​in real time, the cabin displacement can be calculated.

[0059] Step 2: Perform static data calibration while the wind turbine is stationary. Calibration parameters include: the initial pitch angle of the nacelle and the initial coordinates of the equivalent point of the tower base reference. The equivalent point of the tower base reference is obtained through the following process:

[0060] S2.1 Image Acquisition: Capture an image containing the reference object using a camera. Ensure good image quality, uniform lighting, and that the reference object is clearly visible. Figure 4 The image shown is of the unit when it is shut down. The outline of the white cement platform at the bottom of the tower is selected as the identification object (the location selected by the box in the figure).

[0061] S2.2 Region of Interest (ROI) Selection: The original image is preprocessed to determine the Region of Interest (ROI), which is the portion of the image containing the target object. Attention is focused on the target object through thresholding, edge detection, or recognition of specific color / texture features, reducing the complexity and computational burden of subsequent processing.

[0062] S2.3 Edge Recognition: Apply edge detection algorithms (such as Canny, Sobel, Laplacian, etc.) to identify the contour edges of the reference object (the circular cement platform at the base of the tower) within the ROI region, such as... Figure 5 As shown. This allows us to obtain the set of pixels at the object's edge, preparing for further extraction of the object's shape features.

[0063] S2.4 Reference Object Contour Point Extraction: For the identified edges, further extract contour points. This is usually achieved through contour detection algorithms, such as the findContours function in the OpenCV library. Contour points contain all pixel coordinate information of the object's edge.

[0064] S2.5 Curve Fitting: Perform shape analysis on the extracted contour point set, and select special markers or shapes. Figure 6 The image shows the fitting result of the circular cement platform at the bottom of the tower when the unit is shut down; its shape is approximately a perfect circle.

[0065] When the unit is running, since it is distorted into an ellipse in the imaging, an ellipse fitting algorithm (such as the least squares method) can be used to fit the contour point data into an optimal ellipse curve. This step will obtain parameters such as the center coordinates, major and minor axes, and rotation angle of the ellipse.

[0066] S2.6, Locating the Equivalent Point: The point furthest from the tower cylinder in the fitted shape is taken as the target equivalent point. In this embodiment, the coordinates of the equivalent point of the reference object at the bottom of the tower are determined based on the ellipse fitting result. Generally, the point furthest from the tower cylinder in the ellipse is selected as the target point. The advantage of ellipse fitting is that the results of more contour recognition will be more accurate, avoiding and reducing the interference and error of single-point recognition.

[0067] Step 3: Perform operational data calculations while the wind turbine is running, including analyzing real-time images captured during wind turbine operation to calculate the forward and backward direction pixels of the equivalent point of the tower base reference, as well as the pitch angle of the nacelle during operation.

[0068] Step 4: Calculate the cabin displacement using the following formula:

[0069] d i =-(y i -y0)·rh·(θ i -θ0)

[0070] Where, d i Indicates cabin displacement, y i y0 represents the initial calibration value of the equivalent point of the reference object at the bottom of the tower during operation, r represents the spatial distance represented by a unit pixel in the image, h represents the camera installation height, and θ represents the coordinates of the equivalent point of the reference object at the bottom of the tower. i θ0 represents the pitch angle during cabin operation, and θ0 represents the initial calibration value of the cabin pitch angle.

[0071] After the above process, the initial equivalent point coordinates are (732, 333), and the equivalent point coordinates during operation are (736, 311). The displacement in the y-direction is -22 pixels. The spatial distance r represented by a unit pixel in the ground image is 0.0569 m / pixel. The measured change in the forward and backward tilt angle is 0.029 rad (approximately 1.66°). The camera installation height is h = 100 m. The displacement value d is calculated using the displacement formula. i =22*0.0569-100*0.029=-1.648m, meaning the cabin moved backward by 1.648m, of which the displacement of the ground reference point was -1.252m.

[0072] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0073] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for monitoring cabin displacement based on cabin video and a dynamic tilt sensor, characterized in that, The method includes: Step 1: Use the camera installed at the front of the cabin to capture real-time ground images, and use the two-axis dynamic tilt sensor installed inside the cabin to collect the tilt angle in real time. Step 2: Perform static data calibration while the wind turbine is stationary. The parameters for static data calibration include: the initial pitch angle of the nacelle and the initial calibration values ​​of the coordinates of the equivalent point of the tower base reference. The equivalent point of the tower base reference is obtained through the following process: S2.1, Image Acquisition: Capture an image containing the reference object using a camera; S2.2 Region Selection: Preprocess the image to determine the region of interest; S2.3 Edge Recognition: Applying edge detection methods to identify the contour edges of reference objects within the region of interest; S2.4 Reference object contour point picking: For the identified edges, extract the contour points containing all pixel coordinate information of the object edges; S2.5 Curve Fitting: Perform shape analysis on the extracted contour point set, perform shape fitting, and obtain the fitted curve; S2.6, Locating the equivalent point: Select a point on the fitted curve as the equivalent point of the reference object at the bottom of the tower; Step 3: Perform operational data calculations while the wind turbine is running; Step 4: Calculate the cabin displacement using the following formula: ， in, d i Indicates cabin displacement. y i This represents the coordinates of the equivalent point of the reference point at the bottom of the tower during operation. y 0 represents the initial calibration value of the coordinates of the equivalent point of the reference object at the bottom of the tower. r This represents the spatial distance represented by a unit pixel in an image. h Indicates the camera installation height. θ i Indicates the pitch angle during cabin operation. θ 0 indicates the initial calibration value of the cabin pitch angle.

2. The cabin displacement monitoring method based on cabin video and dynamic tilt sensor according to claim 1, characterized in that: The preprocessing described in step S2.2 includes forming regions of interest through threshold segmentation, edge detection, and identification of specific color and / or texture features.

3. The cabin displacement monitoring method based on cabin video and dynamic tilt sensor according to claim 1, characterized in that: The edge recognition described in step S2.3 includes using an edge detection method to identify the contour edges of a reference object within the region of interest.

4. The cabin displacement monitoring method based on cabin video and dynamic tilt sensor according to claim 1, characterized in that: The reference object contour point picking in step S2.4 is achieved using a contour detection method.

5. The cabin displacement monitoring method based on cabin video and dynamic tilt sensor according to claim 1, characterized in that: The curve fitting described in step S2.5 uses the least squares method to fit the tower profile into an elliptical curve.

6. The cabin displacement monitoring method based on cabin video and dynamic tilt sensor according to claim 1, characterized in that: Step S2.6 Select the point on the fitted curve that is farthest from the tower as the equivalent point of the reference object at the bottom of the tower.

7. The cabin displacement monitoring method based on cabin video and dynamic tilt sensor according to claim 1, characterized in that: Step 3, operational data calculation, includes analyzing real-time images captured during wind turbine operation to calculate the forward and backward direction pixels of the equivalent point of the tower base reference, as well as the pitch angle of the nacelle during operation.

Images