Paddle deformation measurement method based on rotation-following vision

Abstract

The invention belongs to the field of helicopter rotor blade motion parameter measurement, and relates to a blade deformation measurement method based on rotation-following vision. The method comprises the steps that LED mark points are installed on the upper surfaces of the paddles; a follow-up rotating binocular stereoscopic vision system used for shooting the upper surfaces of the blades obliquely and downwards is mounted above the propeller hub; images of the LED mark points are collected through a rotation-following binocular stereoscopic vision system according to the test state; identifying pixel coordinates of the LED mark points in the image; according to an internal reference matrix, a rotation matrix and a translation matrix, obtained through pre-calibration, of the following-rotation binocular stereoscopic vision system, the pixel coordinates of the LED mark points are converted into three-dimensional coordinates under a coordinate system of the following-rotation binocular stereoscopic vision system; and converting the three-dimensional coordinates of the LED mark points in the coordinate system of the follow-up rotation binocular stereoscopic vision system into the three-dimensional coordinates of the LED mark points in the coordinate system of the propeller hub, thereby calculating the deformation of the propeller blade.

Description

Technical Field

[0001] The invention belongs to the field of helicopter rotor blade motion parameter measurement, and relates to a blade deformation measurement method based on rotation vision. Background Art

[0002] The rotor is the main source of lift for helicopters. The deformation of rotor blades is a direct reflection of the aerodynamic shape, dynamic characteristics, and weight distribution of the blades, and is directly related to the dynamic stability and vibration level during flight. The monitoring and adjustment methods of the working status of helicopter rotors have always been the research focus of relevant scholars at home and abroad. Although the rotor undergoes strict production quality control of materials, processes, tooling, etc. during the manufacturing process, it is still difficult to ensure the consistency of the characteristics of each blade. This inconsistency will directly affect the flight performance of the helicopter, especially when the helicopter is flying at high speed. By monitoring the deformation of the blades, the dynamic changes of the blades during rotation can be known, which can provide an important reference for monitoring and adjusting the status of the helicopter rotor.

[0003] At present, the deformation of rotor blades is generally measured by pre-embedding resistance strain gauges in known locations. This method has the advantages of high accuracy and stable measurement results. However, this method is a contact measurement, which requires the object to be measured to have a certain rigidity. It can only be measured step by step, and full-field data cannot be obtained. The improved method is to use fiber grating sensors, which have the advantages of more measurement points and larger measurement information than resistance strain gauge measurement. However, fiber optic measurement causes measurement errors because the pasting position of the fiber optic sensor is difficult to completely match the reference position; and both fiber grating strain gauges and resistance strain gauges have lateral effects. If the fiber grating strain gauge does not coincide with the strain direction, there must be errors between the measured strain and the true strain. Summary of the invention

[0004] Purpose of the invention: To provide a blade deformation measurement method based on rotation vision to improve the measurement accuracy of helicopter high-speed rotating blade deformation.

[0005] Technical solution:

[0006] A blade deformation measurement method based on rotation vision is provided, comprising:

[0007] An LED marking point is installed on the upper surface of the blade; a rotating binocular stereo vision system is installed above the propeller hub for obliquely downward photographing the upper surface of the blade;

[0008] According to the test state, the image of the LED marking point is collected through the rotating binocular stereo vision system;

[0009] Identify the pixel coordinates of the LED marker points in the image;

[0010] According to the pre-calibrated intrinsic parameter matrix, rotation matrix and translation matrix of the rotating binocular stereo vision system, the pixel coordinates of the LED marker point are converted into three-dimensional coordinates in the coordinate system of the rotating binocular stereo vision system;

[0011] The three-dimensional coordinates of the LED marking point in the coordinate system of the rotating binocular stereo vision system are converted to the three-dimensional coordinates of the LED marking point in the hub coordinate system, so as to calculate the blade deformation.

[0012] Furthermore, LED marking points are installed on the upper surface of the propeller blade, including:

[0013] N light source marking points are pasted on the quarter chord line and three quarters chord line of the 0.5R~1.0R blades, and the center distance of each marking point is 0.045R, where R is the blade length.

[0014] Furthermore, the rotational vision system has two variants:

[0015] The first variant is a single-channel system SZSL-01, equipped with a CMOS camera with a resolution of 1920×1464 pixels and a maximum frame rate of 411 frames; the CMOS camera photographs the upper surface of the blade obliquely downward; the minimum aperture is selected;

[0016] The second variant is the SZSL-02 dual-channel system, which is equipped with two CMOS cameras side by side with a resolution of 1920×1464 pixels and a maximum frame rate of 411 frames. The upper surface of the blade is in the common field of view of the two CMOS cameras. The two CMOS cameras need to be installed as far apart as possible on the propeller hub, using the smallest aperture.

[0017] Furthermore, for the first variant, a calibration function is fitted using a binomial according to the height of the artificially lifted blade and the corresponding pixel displacement recorded, and the blade flapping amount can be calculated in real time according to the calibration function and the pixel displacement under different test conditions.

[0018] Furthermore, the method further comprises:

[0019] The camera is calibrated according to Zhang Zhengyou's calibration method to obtain the intrinsic parameter matrix M int ;M int Including intrinsic matrix, rotation matrix, and translation matrix.

[0020] Furthermore, the calculation method of the transfer matrix of the rotating binocular stereo vision system coordinate system-hub coordinate system includes:

[0021] When the same mark point on the blade rotates around the hub center for one circle at low speed and zero total pitch, a circular motion trajectory can be obtained. The center coordinates of the circular motion trajectory can be obtained by fitting the center of the circle. The n LED mark points can obtain n center coordinates. The n center coordinates are fitted to obtain the optimal center coordinates by using the least squares method. The x and y coordinates of the optimal center are used as the x and y coordinates of the hub center, that is, the x and y coordinates of the origin of the hub coordinate system.

[0022] Substitute the x and y coordinates of the origin of the hub coordinate system into the plane equation Ax+By+Cz+D=0 to obtain the z coordinate of the origin of the hub coordinate system, thereby obtaining the transfer matrix of the binocular stereo vision system coordinate system-hub coordinate system.

[0023] Furthermore, the rotating binocular stereo vision system is mounted on the propeller hub through a mounting frame;

[0024] The mounting frame includes: connector, fixing frame, 10G fiber optic converter, and counterweight;

[0025] The connector is a barrel-shaped structure with an opening at the lower end. An annular hole wall is arranged in the middle of the barrel along the axial direction. The annular hole wall and the barrel wall are divided into five spaces by five partition plates. The top plate of each space is provided with a weight-avoiding hole for avoiding the hub, and the inner hole of the annular hole wall is used for wiring. A fixing frame is arranged on the upper surface of the connector, and a camera is installed in the fixing frame. The data transmission line of the camera is connected to the optoelectronic slip ring installed on the hub shaft through a 10 Gigabit optical fiber converter. A counterweight block is installed on the upper end face of the connector to balance the center of gravity.

[0026] Furthermore, the method further comprises:

[0027] A photoelectric sensor is installed on the hub, which is connected to a photoelectric slip ring and then to a camera. The photoelectric sensor is used to measure the rotor speed.

[0028] When the photoelectric sensor senses the reflective sticker on the hub shaft, it triggers the generation of a rising edge signal, and the rotation direction of the blade is recorded based on the rising edge signal, starting multi-directional acquisition.

[0029] Beneficial effects:

[0030] The blade deformation measurement method of the present invention has the characteristics of high precision, good stability and strong anti-interference ability, and can be used for high-precision visual measurement engineering of the motion trajectory and deformation of high-speed rotating blades of helicopters. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a flow chart of a blade deformation measurement method based on rotational vision.

[0032] Figure 2 Paste the schematic for the rotor blade marking points.

[0033] Figure 3a This is the overall layout diagram of the monocular.

[0034] Figure 3b This is the overall layout diagram of the binocular system.

[0035] Figure 4a Schematic diagram of the binocular camera mounting bracket.

[0036] Figure 4b Schematic diagram of the propeller hub-camera mount connector.

[0037] Figure 5 Schematic diagram of coordinate transformation.

[0038] Figure 6 Create a plot of the results for the hub coordinate system.

[0039] Figure 7 Schematic diagram of the blade swinging motion.

[0040] Figure 8 Schematic diagram of the blade oscillation motion.

[0041] Fig. 9 Schematic diagram of the blade twisting motion. DETAILED DESCRIPTION

[0042] Video measurement technology based on computer vision is a new technology that has developed rapidly in the field of non-contact measurement in recent years. It is based on modern optics and integrates modern scientific technologies such as computer technology, image processing and analysis technology. The main feature of video measurement is that it can measure and interpret on the image without touching the object itself, and can capture the image of dynamic objects in an instant. This paper proposes a visual, non-contact helicopter rotor blade deformation measurement method based on rotational vision to improve the accuracy of blade deformation measurement.

[0043] Specifically include:

[0044] (1) First, paste ultra-thin LED point light source marking points on the upper surface of each blade. Paste N light source marking points on the quarter chord line and three-quarter chord line of the 0.5R to 1.0R blades, such as Figure 2 As shown, the center distance of each marking point is 0.045R, where R is the blade length.

[0045] (2) There are two variants of the rotational vision system:

[0046] SZSL-01 single channel system Figure 3a As shown, it is equipped with a CMOS camera with a resolution of 1920×1464 pixels and a maximum frame rate of 411 frames.

[0047] SZSL-02 dual channel system Figure 3b As shown, it is equipped with two CMOS cameras with a resolution of 1920×1464 pixels and a maximum frame rate of 411 frames.

[0048] The camera model selected is VLXT-28M.I.JP, the lens focal length is 8mm, the field of view of a single camera at a distance of 1.5m is 1.5m×1.125m, the aperture size is 16, and the baseline distance is selected as 210mm.

[0049] The focal length of the lens is 8mm: Theoretically, 8mm can provide a sufficient field of view to tolerate blade deformation.

[0050] The baseline distance is 210mm: In theory, the larger the baseline distance, the higher the accuracy and the larger the field of view required. In this project, the blade shape variables under the rotation vision are mainly measured. Therefore, many limiting factors need to be considered. After multiple experiments, it was verified that under the consideration of distance limit, center of gravity and parallax size, 210mm will neither introduce too large errors nor too extreme mechanical installation.

[0051] The height of the mounting frame is 380mm: The height of the mounting frame in this project is limited to less than 500mm. Taking into account the factors of aerodynamics, rotation and pixel compression, the height is set to 380mm, which can obtain sufficient blade imaging.

[0052] Aperture size is 16: Select the smallest aperture to minimize background interference.

[0053] Figure 4a The diameter of the mounting bracket is consistent with the diameter of the helicopter hub, and the rotating visual system is installed on the bracket; at the bottom is a device that can be connected to the helicopter hub ( Figure 4b ), the device can be mounted on the propeller hub; the data transmission line and power line are connected to the photoelectric slip ring under the blade through the hole of the mounting frame, where the electrical ring supplies power to the camera and the optical fiber transmits data to the camera. And the whole device can be rotated by loosening the fixing device, so that different blades can be photographed without removing the upper camera.

[0054] (3) Continuously collect blade images according to the test status.

[0055] (4) Using an edge detection algorithm to detect the blade image and extract the edge of the object contour in the image; in this embodiment, the EDPF algorithm is used to detect the blade image;

[0056] (5) During the movement of the blades, they will twist and deform, that is, flap, oscillate, and twist.

[0057] The SZSL-01 single-channel system is calibrated only by pixel equivalent, and the calibration function is fitted with binomial to obtain the relationship between pixel offset and blade flapping, and the blade flapping amount can be calculated in real time.

[0058] The SZSL-02 dual-channel system must first establish the conversion relationship from the image coordinate system to the camera coordinate system, and then convert the camera coordinate system to the coordinate system with the hub center as the origin. The specific conversion process is as follows: Figure 5 shown.

[0059] First, the camera internal parameters are calibrated according to Zhang Zhengyou's calibration method to obtain the intrinsic parameter matrix M int ; Then, according to the known marker positions m, n in the image, according to the calculation formula:

[0060]

[0061] Get the coordinate X of the camera coordinate system c , Y c , Z c Finally, the rotation center of the hub is calculated based on the 0 total pitch of 300 revolutions, and the conversion from the camera coordinate system to the hub coordinate system is performed according to the formula. Finally, the rotation matrix is ​​used to transfer the data in all states to the hub coordinate system.

[0062]

[0063] Specifically, the same marked point on the blade rotates around the hub center for one circle at low speed and zero total pitch to obtain a circular motion trajectory. The center coordinates are obtained by circle center fitting. n marked points can obtain n center coordinates. The n center coordinates are fitted to obtain the optimal center coordinates using the least squares method. The x and y coordinates of the optimal center are used as the x and y coordinates of the hub center, that is, the x and y coordinates of the origin of the hub coordinate system.

[0064] Substitute the x and y coordinates of the origin of the hub coordinate system into the plane equation Ax+By+Cz+D=0 to obtain the z coordinate of the origin of the hub coordinate system. The result is as follows: Figure 6 shown.

[0065] (2) Swing angle calculation

[0066] The blade swings like Figure 7 As shown:

[0067] O t is the center of the hub coordinate system, and the blade flapping angle is Figure 5-2 1. Beta angle, P r P is the mark point at one quarter of the rotor position from the blade tip when the rotor rotates at low speed. r The mark point is at O ​​after the blades flap under high-speed rotation.t Projection point in the yz plane.

[0068] Let P r With P r 'In O t The three-dimensional coordinates in the hub coordinate system with the origin as P are r (x r ,y r , z r )and By calculating P r , P r ′ and hub center O t The vector composed of and The included angle between them can be used to obtain the blade flapping angle, as shown in formula (5-36):

[0069]

[0070] in,

[0071]

[0072] (3) Calculation of swing angle

[0073] The blade vibration motion is as follows Figure 8 As shown:

[0074] O t is the center of the hub coordinate system, and the blade swing angle is Figure 5-2 2In the ξ angle, P r P is the mark point at one quarter of the rotor position from the blade tip when the rotor rotates at low speed. r ′ is the mark point at O ​​after the blades vibrate under high-speed rotation of the rotor t Projection point in the XY plane.

[0075] Let P r and P′ r In O t The three-dimensional coordinates in the hub coordinate system with the origin as P are r (x r ,y r , z r ) and P′ r (x′ r , y′ r , z′ r ). By calculating P r , P′ r With hub center O t The vector composed of and The included angle between the blades can be used to obtain the blade swing angle, as shown in formula (5-38):

[0076]

[0077] (4) Torsion angle calculation

[0078] The blade twisting motion is Fig. 9 As shown:

[0079] O t is the center of the hub coordinate system, the plane where the solid line is located is the plane where the blades are located when the rotor is rotating at low speed without twisting, the dotted line is the plane where the blades are located after twisting motion occurs when the rotor is rotating at high speed, and the angle θ is the blade twist angle. The angle θ can be calculated using the angle between the two plane normal vectors, and the calculation formula is shown in formula (5-39).

[0080]

[0081] in, are the normal vectors of the two planes, The three-dimensional coordinate information of the blade marking points can be used for calculation. Suppose the three-dimensional coordinates of any three non-collinear marking points on a blade before twisting in the hub coordinate system are p a (x a ,y a , z a ), p b (x b ,y b , z b ), p0(x c ,y c , z c ). At the same azimuth, the three-dimensional information of the marking point after the blade is twisted is Then we have:

[0082]

[0083]

Claims

1. A blade deformation measurement method based on rotation vision, characterized in that: include: An LED marking point is installed on the upper surface of the blade; a rotating binocular stereo vision system is installed above the propeller hub for obliquely downward photographing the upper surface of the blade; According to the test state, the image of the LED marking point is collected through the rotating binocular stereo vision system; Identify the pixel coordinates of the LED marker points in the image; According to the pre-calibrated intrinsic parameter matrix, rotation matrix and translation matrix of the rotating binocular stereo vision system, the pixel coordinates of the LED marker point are converted into three-dimensional coordinates in the coordinate system of the rotating binocular stereo vision system; The three-dimensional coordinates of the LED marking point in the coordinate system of the rotating binocular stereo vision system are converted to the three-dimensional coordinates of the LED marking point in the hub coordinate system, so as to calculate the blade deformation.

2. The method according to claim 1, characterized in that LED marking points are installed on the surface of the propeller, including: N light source marking points are pasted on the quarter chord line and three quarters chord line of the 0.5R~1.0R blades, and the center distance of each marking point is 0.045R, where R is the blade length.

3. The method according to claim 2, characterized in that The turn-by-turn vision system is available in two variants: The first variant is a single-channel system SZSL-01, equipped with a CMOS camera with a resolution of 1920×1464 pixels and a maximum frame rate of 411 frames; the CMOS camera photographs the upper surface of the blade obliquely downward; the minimum aperture is selected; The second variant is the SZSL-02 dual-channel system, which is equipped with two CMOS cameras side by side with a resolution of 1920×1464 pixels and a maximum frame rate of 411 frames. The upper surface of the blade is within the common field of view of the two CMOS cameras. The installation distance between the two CMOS cameras on the hub needs to be as large as possible, and the minimum aperture is selected.

4. The method according to claim 3, characterized in that: For the first variant, the calibration function is fitted using the binomial according to the height of the artificially lifted blade and the corresponding pixel displacement recorded. The blade flapping amount can be calculated in real time based on the calibration function and the pixel displacement under different test conditions.

5. The method according to claim 1, characterized in that The method further comprises: The camera is calibrated according to Zhang Zhengyou's calibration method to obtain the intrinsic parameter matrix M int ;M int Including intrinsic matrix, rotation matrix, and translation matrix.

6. The method according to claim 5, characterized in that The calculation method of the transfer matrix of the rotating binocular stereo vision system coordinate system-hub coordinate system includes: When the same mark point on the blade rotates around the hub center for one circle at low speed and zero total pitch, a circular motion trajectory can be obtained. The center coordinates of the circular motion trajectory can be obtained by fitting the center of the circle. The n LED mark points can obtain n center coordinates. The n center coordinates are fitted to obtain the optimal center coordinates by using the least squares method. The x and y coordinates of the optimal center are used as the x and y coordinates of the hub center, that is, the x and y coordinates of the origin of the hub coordinate system. Substitute the x and y coordinates of the origin of the hub coordinate system into the plane equation Ax+By+Cz+D=0 to obtain the z coordinate of the origin of the hub coordinate system, thereby obtaining the transfer matrix of the binocular stereo vision system coordinate system-hub coordinate system.

7. The method according to claim 1, characterized in that The rotating binocular stereo vision system is installed on the propeller hub through a mounting frame; The mounting frame includes: connector, fixing frame, 10G fiber optic converter, and counterweight; The connector is a barrel-shaped structure with an opening at the lower end. An annular hole wall is arranged in the middle of the barrel along the axial direction. The annular hole wall and the barrel wall are divided into five spaces by five partition plates. The top plate of each space is provided with a weight-avoiding hole for avoiding the hub, and the inner hole of the annular hole wall is used for wiring. A fixing frame is arranged on the upper surface of the connector, and a camera is installed in the fixing frame. The data transmission line of the camera is connected to the optoelectronic slip ring installed on the hub shaft through a 10 Gigabit optical fiber converter. A counterweight block is installed on the upper end face of the connector to balance the center of gravity.

8. The method according to claim 7, characterized in that The method further comprises: A photoelectric sensor is installed on the hub, which is connected to a photoelectric slip ring and then to a camera. The photoelectric sensor is used to measure the rotor speed. When the photoelectric sensor senses the reflective sticker on the hub shaft, it triggers the generation of a rising edge signal, and the rotation direction of the blade is recorded based on the rising edge signal, starting multi-directional acquisition.

Images