Full-domain ultra-sensitive rotating body vibration measurement method under polar coordinate system

Abstract

The invention discloses a global ultra-sensitive rotating body vibration measurement method under a polar coordinate system, and relates to the technical field of structural health monitoring. High-precision vibration monitoring of a rotating structure is achieved through non-contact optical measurement, and the method comprises the steps that scale parameters are calibrated through known physical dimensions, feature points are recognized through an SIFT + RANSAC algorithm, a homography matrix is calculated, and the rotating angle and the center are obtained through decomposition. Descartes coordinate pixel intensity is mapped to polar coordinates through bilinear interpolation, singular value decomposition is carried out after a space-time pixel intensity matrix is constructed, and front k-order modal reconstruction is selected according to the fact that the cumulative energy proportion is larger than 90% so as to separate noise. And calculating full-field vibration displacement under polar coordinates by adopting an optical flow algorithm, and finally converting the full-field vibration displacement into a physical space displacement field through scale parameters. According to the method, rotation and vibration are effectively separated, the vibration measurement precision reaches the sub-pixel level, the method is suitable for high-precision vibration analysis of complex rotating bodies such as fan blades, and a new way is provided for nondestructive testing of rotating machinery.

Description

Technical Field

[0001] The present invention belongs to the technical field of structural health monitoring, and particularly relates to a method for measuring the vibration of a rotating body with ultra-high sensitivity in the polar coordinate system over the entire domain. Background Art

[0002] The vibration measurement of a rotating body is crucial in modern engineering, especially in the structural monitoring and maintenance of wind turbines. During the operation of a wind turbine, the rotational vibrations of the blades and the generator directly affect its efficiency and safety. Through precise vibration measurement, potential faults such as imbalance, resonance, or wear can be identified in a timely manner, avoiding equipment damage and downtime, thereby extending the service life of the wind turbine and reducing maintenance costs. In addition, real-time monitoring of vibration data helps optimize the operating performance of the wind turbine, improve the energy output efficiency, and support the sustainable development of renewable energy. Therefore, the vibration measurement of a rotating body is not only the key to ensuring equipment safety but also an important means to improve the efficiency of wind energy utilization. Traditional methods for measuring the vibration of a rotating body (such as accelerometers) achieve vibration measurement by deploying sensors on the object to be measured and obtaining real-time acceleration data. This technology can quickly respond to dynamic vibration changes and provide reliable long-term detection data. However, this method has low measurement accuracy under low-frequency and small vibrations, and its spatial resolution depends on the number of sensors deployed, usually unable to achieve full-domain measurement. In addition, the installation, calibration, and maintenance of multiple sensors also increase the usage cost and operation difficulty.

[0003] For the full-domain measurement technology based on a camera, only a consumer-grade camera is required to capture the object to be measured; each pixel of the camera can be regarded as a sensor, so full-domain dense vibration measurement can be achieved; and the quantity measured by the camera is the displacement change, so it has good measurement accuracy at low frequencies. However, the existing camera-based measurement technology is rarely used for the vibration measurement of rotating bodies because the vibration measurement of rotating bodies involves complex motion coupling, and traditional vision algorithms cannot directly decouple the vibration components from the images. At the same time, since the camera is usually deployed at a relatively far position from the object to be measured (such as the wind turbine blade), the quantization rounding error of the camera itself will be amplified in the actual physical quantity, resulting in insufficient measurement accuracy.

[0004] Combined with the actual engineering requirements, when using a camera for measurement, how to extract the vibration components from the response of rotation and vibration coupling and break through the measurement accuracy limitation of the camera itself is the key to achieving high-precision full-field measurement of the deformation vibration of a rotating structure.

[0005] The present invention overcomes the disadvantages of traditional measurement technologies, such as low measurement accuracy, insufficient spatial resolution, and high cost in the low-frequency state; at the same time, it improves the deficiencies that the existing machine vision algorithms cannot achieve the decoupling of the vibration of a rotating body and the measurement accuracy is affected by the quantization rounding error. Summary of the Invention

[0006] In order to overcome the disadvantages and deficiencies of the above-mentioned prior art, the present invention adopts the following technical solutions: A method for measuring the vibration of a global hypersensitive rotating body in a polar coordinate system has the following process: S11. Image acquisition: Adjust the camera parameters so that the imaging plane coincides with the rotating plane of the rotating body, and continuously acquire images of the rotating body; S12. Calibrate the scale parameter: Determine the scale parameter of millimeters / pixel based on the physical size of the rotating body when it is not rotating and the number of image pixels, and dynamically update it through fixed marker points; S13. Feature recognition and rotation parameter estimation: Apply the SIFT algorithm to each frame of image to extract feature points, match the feature points through BFMatcher and RANSAC, and decompose the homography matrix to obtain the rotation angle and the rotation center; S14. Polar coordinate mapping: Based on the rotation center and the rotation angle, map the Cartesian coordinate pixel intensity to the polar coordinate system through bilinear interpolation; S15. Construct a spatio-temporal pixel intensity matrix: Compose the time-series pixel intensities after polar coordinate mapping into a spatio-temporal matrix; S16. Singular value decomposition and modal reconstruction: Perform SVD decomposition on the spatio-temporal matrix, and select the first k-order singular modes to reconstruct the signal; S17. Optical flow displacement calculation: Use the optical flow algorithm to calculate the vibration displacement for the reconstructed polar coordinate image; S18. Coordinate transformation: Combine the scale parameter to convert the displacement field in the polar coordinate system to the physical displacement field in the Cartesian coordinate system.

[0007] Preferably, calibrating the scale parameter includes: Calculate the average value as the final scale parameter by measuring the ratio of the physical size of the rotating body when it is not rotating to the corresponding number of pixels multiple times; if the deviation of a certain measurement value from the average value exceeds 10%, then reject it and re-measure; when the rotating body is occluded, update the scale parameter in real time through the fixed marker points near the rotation axis.

[0008] Preferably, decomposing the homography matrix includes: Extract the first two column vectors of the homography matrix H , ; Calculate the rotation matrix , where , , are the column vectors constituting the rotation matrix , , , ; The rotation angle , in radians, and are respectively and the first elements of the vector, used to determine the rotation angle of the rotating body; the rotation center is obtained by solving ; the rotation angle is obtained by calculating the eigenvalues in the homography matrix using the arctangent function ; the rotation center is estimated by the eigenvalues in the homography matrix , .

[0009] Preferably, the conversion formula from Cartesian coordinates to polar coordinates for polar coordinate mapping is: ; where is the rotation angle of the t-th frame, obtained by decomposing the homography matrix, used to reset the direction of the polar coordinates; the pixel intensity in polar coordinates is calculated using bilinear interpolation : ; where are the weight coefficients of the adjacent four pixel points; the calculation of the weight coefficients is as follows: Let the Cartesian coordinates corresponding to the polar coordinate point be , and its adjacent four pixel points be , , , , then the weight coefficients are: , , , ; where , .

[0010] Preferably, the condition for selecting the first k singular modes is: the cumulative energy ratio of the singular values is greater than 90%.

[0011] Preferably, the displacement calculation in the optical flow algorithm satisfies: , indicating the optical flow displacement of the measurement point with coordinates at time t in the polar coordinate system, reflecting the position change of the measurement point over time in polar coordinates; where is the pixel intensity value; is the pixel intensity gradient of the measurement point; and are the coordinates of any measurement point in the polar coordinate system; is the time; is the measurement point with coordinates The pixel intensity value of the measurement point at the initial moment (t = 0); is the pixel intensity value of the measurement point with coordinates at time t.

[0012] Preferably, the dynamic update of the scale parameter further includes: when the rotating body blade is blocked, the scale parameter is adjusted in real time through the pixel position change of the fixed marking point.

[0013] Preferably, the frame rate and lens focal length in the camera parameters are determined according to the motion characteristics of the measurement object.

[0014] In summary, due to the adoption of the above technical solutions, the beneficial effects of the present invention are as follows: 1. By combining SIFT feature matching, RANSAC outlier rejection, and homography matrix decomposition, the present invention accurately extracts the rotation angle and rotation center, reducing image noise and occlusion interference; at the same time, when calibrating the scale parameter, multiple measurements are taken and outliers (re-measure when the deviation exceeds 10%) are removed, and combined with a high-precision accelerometer (such as PCB352C22) for comparative verification, improving the measurement accuracy and reliability of the vibration displacement.

[0015] 2. The present invention adopts polar coordinate mapping (Cartesian coordinates → polar coordinates), and resets the 0 direction of the polar coordinates through the rotation angle, making the vibration analysis of the rotating body more in line with its motion characteristics and avoiding the complex deformation calculation caused by rotation in the Cartesian coordinate system. The bilinear interpolation weight optimization and polar coordinate subset sampling strategy further reduce image distortion, ensuring the continuity and stability of the displacement field reconstruction.

[0016] 3. The present invention uses singular value decomposition (SVD) to extract the dominant vibration mode (cumulative energy > 90%), and directly estimates the displacement field in polar coordinates by combining the optical flow algorithm, realizing an efficient conversion from the spatio-temporal pixel intensity matrix to the physical displacement. This method can not only filter out high-frequency noise, but also obtain the global vibration information of the rotating body through one measurement, avoiding the limitations of single-point layout of traditional contact sensors (such as accelerometers), and greatly improving the analysis efficiency and coverage. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 Shows a flowchart of a global ultra-sensitive rotating body vibration measurement method in polar coordinates of the present invention; Figure 2 This is the vibration measurement result diagram of the fan blade in Embodiment 1 of the present invention. Detailed implementation manners

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0020] In addition, the described features, structures, or characteristics may be combined in any suitable manner in one or more example embodiments. In the following description, numerous specific details are provided to give a thorough understanding of the example embodiments of the present disclosure. However, those skilled in the art will realize that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or other methods, components, steps, etc. may be used. In other cases, well-known structures, methods, implementations, or operations are not shown or described in detail to avoid obscuring aspects of the present disclosure.

[0021] Embodiment 1:

[0022] Refer to Figure 1 As shown, a method for measuring the vibration of a global ultra-sensitive rotating body in a polar coordinate system in this embodiment is as follows: S11. Image acquisition: Adjust the camera's perspective and focal length so that the rotation plane of the structure overlaps with its imaging plane, start the camera, and continuously collect the structure images; In this embodiment, in order to measure the blade deformation vibration of the fan structure during rotation, the camera is fixed in front of the fan structure with a tripod, and the center of the lens is aligned with the rotation center of the fan. The camera model is SONY NEX-FS700RH, the resolution of the CMOS imaging chip is 1920×1080 pixels, the lens focal length is 24 mm, the acquisition frame rate is 100 frames per second, and the image signal bit depth is 8 bits. Refer to Figure 2 As shown, in order to compare and evaluate the measurement results of the method of the present invention, an accelerometer is arranged on the back of the blade to measure the blade deformation vibration. The accelerometer model is PCB352C22, which is a high-precision contact measurement method, and its measurement accuracy and acquisition frequency are much higher than those of general vibration measurement sensors.

[0023] S12. Calibrate the scale parameter (Unit: mm / pixel); In the non-rotating case, using the known physical size of the fan blade and the number of pixels it occupies in the image, according to the formula Determine the scale parameter; where Represents the physical size of the wind turbine blade; Indicates the number of pixels occupied by the wind turbine blade in the image.

[0024] Through multiple measurements and the corresponding number of pixels , taking the average value as the final scale parameter value. In this embodiment mm / pixel. If occlusion occurs during blade rotation, fixed marking points are set on the wind turbine structure (such as near the rotation axis), and the scale parameter is updated in real time through the marking points. The method of multiple measurements is as follows: Measure the same blade 5 times. Manually select the blade edge each time and record the number of pixels, and calculate the average value , and the final scale parameter . is the number of pixels for the i-th measurement; if the deviation of a certain measurement from the average value exceeds 10%, then this data is excluded and re-measured.

[0025] S13. Apply the Scale-Invariant Feature Transform algorithm (SIFT) to each frame to identify features and determine the rotation center , and then estimate the rotation angle of each frame relative to the reference frame ; The process of identifying features is as follows: Use the SIFT algorithm to identify feature points in the image; use the BFMatcher algorithm to match the feature points of the current image and the feature points of the reference image (the first frame); according to the matched feature points, use the RANSAC algorithm to exclude false matches and calculate the best homography matrix ; This matrix contains rotation information.

[0026] Decompose the homography matrix H into a rotation matrix R and a translation vector t through the following steps: Extract the first two column vectors of H , ; Calculate the rotation matrix , where , , ; The rotation angle , in radians; the rotation center is obtained by solving . Calculate the rotation angle using the arctangent function for the eigenvalues in the homography matrix.

[0027] Estimate the rotation center through the eigenvalues in the homography matrix , .

[0028] S14. Perform interpolation sampling and map the pixel intensity from Cartesian coordinates to polar coordinates based on the rotation center and the estimated rotation angle by solving the equation The polar coordinate subset uses sampling points for each integer radius and resets the 0 - direction of the polar coordinates by subtracting the rotation angle.

[0029] Mapping formula from Cartesian coordinates to polar coordinates: ; where is the rotation angle (in radians) of the \(t\) - th frame, obtained by homography matrix decomposition, and is used to reset the direction of the polar coordinates.

[0030] Calculate the pixel intensity in polar coordinates using bilinear interpolation : ; where are the weight coefficients of the four neighboring pixels.

[0031] The calculation of the weight coefficients is as follows: Let the Cartesian coordinates corresponding to the polar coordinate point be , and its four neighboring pixels be , , , , then the weight coefficients are: , , , ; where , .

[0032] S15. The time - series of the sampled pixel intensity in the transformed polar coordinates can form a spatio - temporal pixel intensity matrix , whose size is , where and are the spatial sampling and temporal sampling numbers of the spatio - temporal pixel intensity matrix respectively; S16. Perform singular value decomposition on the spatio - temporal pixel intensity matrix to obtain the left singular matrix , the singular value diagonal matrix and the right singular matrix ; S17. Select the first singular modes according to the cumulative energy ratio of singular values being greater than 90%; and perform singular value reconstruction by using the singular value modes ; S18. Apply the optical flow algorithm to each frame of the reconstructed (filtered) polar coordinate space image to estimate the full-field vibration displacement of the structure in the polar coordinate system.

[0033] Specifically, the displacement is obtained by interpolating the pixel intensity of the t-th frame of the image in the polar coordinate system and the pixel intensity of the original frame (the 0-th frame) and dividing by the gradient at that position. , where is the pixel intensity value; is the pixel intensity gradient of the measurement point; and are the coordinates of any measurement point in the polar coordinate system; is the time.

[0034] S19. Calculate the spatio-temporal displacement field in the physical space of the Cartesian coordinate system through the scale parameter and the coordinate transformation relationship .

[0035] The beneficial effects of this embodiment are as follows: non-contact full-field measurement, avoiding the interference of the additional mass of the sensor, and realizing non-destructive monitoring of the rotating body; polar coordinate transformation combined with bilinear interpolation effectively avoids the inaccuracy of large displacement measurement caused by rotation and improves the sensitivity of micro-vibration detection; SIFT+RANSAC feature matching and homography matrix decomposition accurately estimate the rotation parameters and have strong anti-occlusion ability; spatio-temporal matrix SVD noise reduction and modal reconstruction effectively separate the noise and the real vibration signal; the optical flow algorithm realizes the full-field displacement solution in the polar coordinate system, and the spatial resolution reaches the pixel level; the calibration parameter dynamic update mechanism ensures long-term measurement accuracy; comparison and verification with a high-precision accelerometer ensure the credibility of the results, and it is applicable to the ultra-sensitive vibration analysis of complex rotating structures such as wind turbine blades.

[0036] The weight coefficients of the present invention are used to measure the influence degree of different factors or variables on a certain result or decision. The definition of the weight coefficient refers to the numerical value assigned to each factor when comparing and evaluating multiple factors to reflect its importance or priority. These weight coefficients can be determined according to specific situations and requirements, and are usually formulated and confirmed by professionals or relevant interested parties. By reasonably setting the weight coefficients, it can help the program or system make more accurate decisions or predictions.

[0037] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0038] Those of ordinary skill in the art will appreciate that the units and algorithm steps of the examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present application.

[0039] Those skilled in the art can clearly understand that for the sake of convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be described herein again.

[0040] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only for some logical function divisions. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of the devices or units can be in electrical, mechanical, or other forms. The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0041] As described above, the foregoing is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for measuring the vibration of a rotating body with ultra-high sensitivity in the entire domain under a polar coordinate system, characterized in that The method process is as follows: S11. Image acquisition: Adjust the camera parameters to make the imaging plane coincide with the rotation plane of the rotating body, and continuously acquire images of the rotating body; S12. Calibrate the scale parameter: Determine the scale parameter of millimeters / pixel based on the physical size of the rotating body when it is not rotating and the number of image pixels, and dynamically update it through fixed marker points; millimeters / pixel represents the actual physical length corresponding to each pixel; S13. Feature recognition and rotation parameter estimation: Apply the SIFT algorithm to each frame of image to extract feature points, match the feature points through BFMatcher and RANSAC, and decompose the homography matrix to obtain the rotation angle and rotation center; S14. Polar coordinate mapping: Based on the rotation center and rotation angle, map the Cartesian coordinate pixel intensity to the polar coordinate system through bilinear interpolation; S15. Construct a spatio-temporal pixel intensity matrix: The time-series pixel intensities after polar coordinate mapping form a spatio-temporal matrix; S16. Singular value decomposition and modal reconstruction: Perform SVD decomposition on the spatio-temporal matrix, and select the first k-order singular modes to reconstruct the signal; S17. Optical flow displacement calculation: Use the optical flow algorithm to calculate the vibration displacement for the reconstructed polar coordinate image; S18. Coordinate transformation: Combine the scale parameter to convert the displacement field in the polar coordinate system to the physical displacement field in the Cartesian coordinate system.

2. The all-region ultra-sensitive rotating body vibration measurement method in a polar coordinate system according to claim 1, wherein The scale parameter calibration includes: Calculate the average value as the final scale parameter by measuring the ratio of the physical size of the rotating body when it is not rotating to the corresponding number of pixels multiple times; if the deviation of a certain measurement value from the average value exceeds 10%, then eliminate it and re-measure; when the rotating body is occluded, update the scale parameter in real time through the fixed marker points near the rotation axis.

3. A method for measuring the vibration of a global ultra-sensitive rotating body in a polar coordinate system according to claim 1, characterized in that The homography matrix decomposition includes: Extract the first two column vectors of the homography matrix H , ; Calculate the rotation matrix , where , , are the column vectors that make up the rotation matrix ; , , ; The rotation angle , in radians, and are respectively and the first elements of the vectors, used to determine the rotation angle of the rotating body; The rotation center is obtained by solving ; The rotation angle is obtained by calculating the eigenvalues in the homography matrix using the arctangent function; Estimating the rotation center through the eigenvalues in the homography matrix , .

4. A method for measuring the vibration of a global ultrasensitive rotating body in a polar coordinate system according to claim 1, characterized in that The conversion formula from Cartesian coordinates to polar coordinates for the polar coordinate mapping is: ; wherein, is the rotation angle of the t-th frame, obtained by homography matrix decomposition, and is used to reset the direction of the polar coordinates; bilinear interpolation is used to calculate the pixel intensity in polar coordinates : ; wherein, are the weight coefficients of the four neighboring pixel points; the calculation of the weight coefficients is as follows: Let the Cartesian coordinates corresponding to the polar coordinate points be , and its four neighboring pixel points are , , , , then the weight coefficients are as follows: , , , ; among them, , .

5. A method for measuring the vibration of an all-region ultrasensitive rotating body in a polar coordinate system according to claim 1, characterized in that, The condition for selecting the first k-order singular modes is: The cumulative energy ratio of the singular values is greater than 90%.

6. The all-region ultra-sensitive rotating body vibration measurement method in a polar coordinate system according to claim 1, wherein In the optical flow algorithm, the displacement calculation satisfies: , which represents the optical flow displacement of the measurement point with coordinates at time t in the polar coordinate system, reflecting the position change of the measurement point over time in the polar coordinates; where is the pixel intensity value; is the pixel intensity gradient of the measurement point; and are the coordinates of any measurement point in the polar coordinate system; is the time; is the pixel intensity value of the measurement point with coordinates at the initial time (t = 0); is the pixel intensity value of the measurement point with coordinates at time t.

7. A method for measuring the vibration of a global ultra-sensitive rotating body in a polar coordinate system according to claim 1, characterized in that, The dynamic update of the scale parameter further includes: When the blades of the rotating body are occluded, adjust the scale parameter in real time through the change in the pixel positions of the fixed marker points.

8. A method for measuring the vibration of an all-region ultrasensitive rotating body in a polar coordinate system according to claim 1, characterized in that In the camera parameters, the frame rate and lens focal length are determined according to the motion characteristics of the measurement object.

Images