Large rotary equipment blade measurement system calibration method based on LM algorithm optimization

Abstract

The invention provides a large-scale rotary equipment blade measurement system calibration method based on LM algorithm optimization, and belongs to the technical field of blade three-dimensional measurement. Points in a world coordinate system are converted to a rotary table coordinate system, are rotated in the rotary table coordinate system, and then are converted to the world coordinate system; and projecting the rotated world coordinate to a pixel coordinate system to obtain a theoretical pixel coordinate, comparing the theoretical pixel coordinate with an actual pixel coordinate, and calculating a re-projection error. Constructing a target function by using the re-projection error, and iteratively optimizing the axis parameters of the rotary table through an L-M algorithm until convergence; and outputting the optimized rotary table axis for blade point cloud splicing. According to the method, the problem of linear axis calculation errors on the basis of mark point position tracking is solved, nonlinear optimization of the rotary table rotating shaft is achieved, and technical support is provided for follow-up rotary splicing and all-topography measurement of blade point clouds.

Description

Technical Field

[0001] This invention belongs to the field of blade three-dimensional measurement technology, specifically, it relates to a calibration method for a large rotating equipment blade measurement system based on LM algorithm optimization. Background Technology

[0002] As the heart of an aircraft, the aero-engine plays a decisive role in its performance. In recent years, with the development of aircraft technology, the thrust-to-weight ratio of aero-engines has increased, and the compressor pressure ratio has been further improved, exacerbating the harshness of the working environment for aero-engine blades and imposing stringent performance requirements. A very common and dangerous mode of aero-engine failure is the formation of cracks at the blade edge, root, and blade itself, eventually leading to fracture. This threatens the engine's performance and reliability. Therefore, the precision requirements for blade manufacturing are very high, making the measurement of aero-engine blade characteristic parameters and morphology extremely important.

[0003] Currently, the demand for aero-engine blade morphology inspection is increasingly moving towards higher resolution and higher efficiency. To achieve accurate 3D reconstruction of the blades, this research aims to optimize the rotation axis spatial pose nonlinearly based on a binocular camera system, after successfully building a blade measurement system and realizing linear calculation of the axis pose. This provides technical and accuracy support for subsequent rotation of point clouds from multiple perspectives around the turntable axis, ultimately achieving full-body 3D measurement of the blade morphology.

[0004] Turntable axis calibration consists of two steps: a linear calculation step based on marker trajectory tracking and a nonlinear optimization step based on the LM algorithm. Although linear calculation of the turntable axis has been achieved in existing research, its pose accuracy cannot be guaranteed. To assess the pose accuracy of the turntable axis, the markers on the calibration module placed on the turntable are rotated around the axis. When rotated to a specific angle, if the theoretical position of the marker at that angle highly coincides with its actual position, it indicates that the pose calibration of the turntable axis is sufficiently accurate. Summary of the Invention

[0005] To address the above, this invention proposes a calibration method for a large rotary equipment blade measurement system based on the LM algorithm optimization, building upon the prediction of marker point motion trajectories in the calibration module. First, based on the transformation relationship between the turntable coordinate system and the binocular camera coordinate system, the three-dimensional positions of each marker point are predicted when the turntable moves to a specific angular position. Based on these prediction results, the position prediction error of the marker points is incorporated into the LM algorithm for optimization, enabling nonlinear optimization of the turntable axis pose. This provides a solid technical foundation for subsequent accurate measurement of the blade's three-dimensional point cloud and accurate evaluation of blade parameters.

[0006] This invention is achieved through the following technical solution: A calibration method for a large rotating equipment blade measurement system based on the LM algorithm optimization: The method specifically includes the following steps: Step 1: Coordinate system transformation. Transform the world coordinate system to the turntable coordinate system to restore the trajectory of the marker point to a standard circle. Step 2: Rotate within the turntable coordinate system. Based on the Rodriguez formula, generate the theoretical position of the marker point for the next frame to ensure that the rotation trajectory conforms to the actual circular motion law. Step 3: Perform inverse coordinate transformation and projection to calculate the theoretical position and output the theoretical pixel coordinates; Step 4: Calculate the reprojection error, calculate the reprojection error between the theoretical position and the actual position, and quantify the turntable axis calibration error; Step 5: Optimize the objective function value using the LM algorithm to obtain the optimal solution for the turntable axis; Step 6, blade point cloud stitching: Using the optimized axis as the rotation reference, the three-dimensional point cloud data of the blade collected from multiple perspectives are fused to achieve three-dimensional reconstruction of the blade surface with sub-pixel accuracy, thus solving the problem of blade parameter measurement error.

[0007] Furthermore, in step 1, let the initial axis direction of the turntable in the world coordinate system be ( x 0, y 0, z The direction vector of this axis in the turntable coordinate system is (0,0,1). x shaft and y The axes form the plane being measured; the transformation relationship between the two coordinate systems can be expressed by a rotation matrix. Translation vector The relationship is as follows;

[0008] in The coordinates of a three-dimensional point in the world coordinate system. Coordinates in the turntable coordinate system; rotation matrix The solution is based on two coordinate systems z The unit vector of the axis and the coordinates of the axis center can be used to achieve this. The solved rotation matrix and translation vector will be used in the subsequent mutual transformation between the turntable coordinate system and the world coordinate system.

[0009] Furthermore, in step 2, after transferring the centers of each feature point from the world coordinate system to the turntable coordinate system, each feature point is rotated within the turntable coordinate system. According to the Rodriguez formula, the transformation matrix of the feature point under a certain rotation angle α in the turntable coordinate system is as follows:

[0010] The transformed coordinates are: .

[0011] Furthermore, in step 3, based on solving the coordinates of the feature points in the turntable coordinate system for the next frame, the coordinates can be transferred to the world coordinate system using the rotation matrix and translation vector between the turntable and the world coordinate system. These coordinates are the theoretical 3D world coordinates of the feature points when the turntable rotates to the next angle. The coordinate transformation formula is as follows:

[0012] in It is the inverse of the rotation matrix.

[0013] Furthermore, in step 4, by left-multiplying the epipolar-corrected intrinsic parameter matrix based on the world coordinate system coordinates, the theoretical coordinates of the marker point in the epipolar-corrected image at the next rotation angle can be obtained. This coordinate quantity can then be used to construct the loss function in the LM algorithm, calculated as follows:

[0014] in, p theory =[ u , v ,1] T These are the pixel coordinates after theoretical epipolar correction. K rect This represents the intrinsic parameter matrix after epipolar correction, used to project three-dimensional coordinates in the world coordinate system into pixel coordinates; The objective function of the LM algorithm is obtained by calculating the reprojection error with the actual pixel coordinates at the next angle, as follows:

[0015] In the above objective function, p il , p ir These represent the actual pixel coordinates of the left and right cameras when the turntable rotates to the corresponding angles. p theoryl , p theoryr This indicates the theoretical position of the turntable at this angle. The objective function can be derived based on the deviation between the theoretical and actual values.

[0016] Furthermore, in step 5, based on the objective function obtained above, an approximate Hessian matrix is ​​constructed as ( J T J+λI ),in IThe identity matrix is ​​used; the update amount Δ of the parameter vector can be solved by solving the linear system equations based on the approximate Hessian matrix. d ;

[0017] Update parameter vector:

[0018] By setting appropriate iteration thresholds and using the LM algorithm to iterate continuously, the precise calculation of the turntable axis equation can be completed, thereby reducing the blade parameter error caused by axis deviation during actual measurement.

[0019] Furthermore, in step 6, the turntable axis calculated by the LM algorithm is used as the rotation center to transform the point clouds at different angles to the same coordinate system, align overlapping areas, remove redundant points, and generate a complete blade surface model.

[0020] A calibration device for a large rotary equipment blade measurement system optimized based on the LM algorithm; The device includes a coordinate system transformation module, a trajectory correction module, a projection module, an error quantization module, an axis optimization module, and a stitching module; The coordinate system transformation module transforms the world coordinate system coordinates to the turntable coordinate system, restoring the trajectory of the marker point to a standard circle; The trajectory correction module is used for rotation within the turntable coordinate system. Based on the Rodriguez formula, it generates the theoretical position of the marker point in the next frame, ensuring that the rotation trajectory conforms to the actual circular motion law. The projection module is used for coordinate inverse transformation and projection, to calculate the theoretical position, and to output the theoretical pixel coordinates. The error quantization module is used to calculate the reprojection error, calculate the reprojection error between the theoretical position and the actual position, and quantify the turntable axis calibration error. The axis optimization module optimizes the objective function value according to the LM algorithm to obtain the optimal solution for the axis of the turntable; The stitching module is used for stitching blade point clouds; using the optimized axis as the rotation reference, it integrates multi-view acquired three-dimensional point cloud data of the blade to achieve sub-pixel-level precision three-dimensional reconstruction of the blade surface, thus solving the problem of blade parameter measurement error.

[0021] An electronic device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the above method.

[0022] A computer-readable storage medium for storing computer instructions that, when executed by a processor, implement the steps of the above-described method.

[0023] Beneficial effects of the present invention Because the camera's optical axis is not perpendicular to the turntable plane, the trajectory of the marker points is elliptical (rather than an ideal circle), which causes model errors and affects the accuracy of subsequent blade point cloud stitching. Therefore, it is necessary to optimize the turntable axis parameters by using coordinate system transformation and the LM algorithm to improve the accuracy of blade measurement. This invention transforms points in the world coordinate system to the turntable coordinate system, rotates them in the turntable coordinate system, and then transforms them back to the world coordinate system. The rotated world coordinates are projected onto the pixel coordinate system to obtain theoretical pixel coordinates, which are then compared with the actual pixel coordinates to calculate the reprojection error. The reprojection error is used to construct an objective function, and the turntable axis parameters are iteratively optimized using the LM algorithm until convergence. The optimized turntable axis is output for blade point cloud stitching.

[0024] This invention addresses the linear axis calculation error based on marker point position tracking by predicting the 3D coordinates of marker points through coordinate system transformation, thereby achieving 2D pixel coordinate prediction for each marker point. An objective function is constructed based on the deviation between the predicted and actual values ​​and then optimized using the LM algorithm. Ultimately, nonlinear optimization of the turntable axis is achieved, providing technical support for subsequent rotational stitching and full-topography measurement of blade point clouds. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating the construction of the LM algorithm based on coordinate system transformation in this invention.

[0026] Figure 2 An image optimized for axis fitting based on the LM fitting algorithm. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Unless otherwise specified, the experimental methods used in the following examples are conventional methods. Unless otherwise specified, the materials, reagents, methods, and instruments used are all conventional materials, reagents, methods, and instruments in the art, and can be obtained commercially by those skilled in the art.

[0029] To ultimately achieve multi-view full-morphology measurement of the blade, turntable axis pose calibration is required. During turntable axis calibration, a disk with multiple marker points is placed on the turntable. The drive software controls the turntable to rotate a certain angle each time, and a binocular camera captures an image at that angle. By tracking the marker point trajectory, a linear calculation result of the turntable axis is obtained. Based on this, the objective function can be successfully constructed using the coordinate system transformation method of this invention. Substituting this into the LM algorithm allows for nonlinear optimization of the axis, resulting in the final pose calibration result.

[0030] This invention proposes a calibration method for a large rotating equipment blade measurement system based on the LM algorithm optimization: Step 1: Coordinate system transformation. Transform the world coordinate system to the turntable coordinate system to restore the marker point's trajectory to a standard circle; eliminate trajectory errors, solve the elliptical trajectory problem caused by camera tilt, and restore the marker point's standard circular motion in the turntable coordinate system. Step 2: Rotate within the turntable coordinate system to correct the trajectory. Based on the Rodriguez formula, generate the theoretical position of the marker point for the next frame to ensure that the rotation trajectory conforms to the actual circular motion law. Step 3: Inverse coordinate transformation and projection, perform theoretical position calculation, and output theoretical pixel coordinates as error comparison benchmark; Step 4: Calculate the reprojection error, calculate the reprojection error between the theoretical position and the actual position, and quantify the turntable axis calibration error; Step 5: Optimize the objective function value using the LM algorithm to obtain the optimal solution for the turntable axis; Step 6, blade point cloud stitching: Using the turntable axis calculated by the LM algorithm, the optimized axis is used as the rotation reference to transform point clouds at different angles to the same coordinate system, align overlapping areas, remove redundant points, and generate a complete blade surface model.

[0031] By fusing 3D point cloud data of the blade acquired from multiple perspectives, accurate 3D reconstruction of the blade surface is achieved, solving the problem of blade parameter measurement error.

[0032] The method of this invention predicts the theoretical positions of each marker point in the next frame of the image by rotating each feature point along the rotation axis based on the three-dimensional coordinates of each feature point in the previous frame. The reprojection error between the theoretical and actual positions is then calculated and used as an important indicator to measure the accuracy of the turntable axis fitting. This indicator can also be used to construct the objective function in the LM algorithm. Based on this, the optimal solution for the turntable axis can be obtained by optimizing the objective function value according to the LM algorithm.

[0033] In the process of measuring marker points on a turntable surface using binocular vision, because the camera's optical axis is not perpendicular to the turntable plane, the trajectories of each marker point are elliptical during the actual rotation of the turntable. This means that when rotating each marker point directly around the rotation axis of the world coordinate system, it is essentially rotating around a circular trajectory, which does not match the actual motion trajectory of the marker point. The result after rotation will deviate significantly from the actual position of the marker point in the next frame. To avoid this model error, a reprojection error optimization method based on the transformation between the world coordinate system and the turntable coordinate system is proposed. According to the translation and rotation transformation relationship between the world coordinate system and the turntable coordinate system, the coordinates of the 3D feature points in the world coordinate system are first transformed to the turntable coordinate system. The feature points of the previous frame are rotated in the turntable coordinate system. At this time, each feature point moves along a circular trajectory in the turntable coordinate system, which conforms to the actual motion. Based on the rotation in the turntable coordinate system, the coordinates of the rotated feature points are transformed back to the world coordinate system and projected onto the epipolar-corrected pixel coordinate system to calculate the reprojection error. Using this reprojection error to solve the LM optimization, the measurement accuracy problem under non-perpendicular viewpoints can be effectively solved. The construction process of the LM algorithm based on coordinate system transformation is as follows: Figure 1 As shown.

[0034] The above process allows for the linear calculation of the initial axis and the rotation of the marker point around the three-dimensional trajectory. The deviation between the theoretical and actual positions of the marker point is calculated to construct the objective function. Finally, LM nonlinear optimization is performed to obtain an accurate estimate of the final turntable axis pose.

[0035] The transformation relationship between the stereo camera coordinate system and the world coordinate system should be analyzed first. Let the initial axis direction of the turntable in the world coordinate system be ( x 0, y 0, z The direction vector of this axis in the turntable coordinate system is (0,0,1). x shaft and y The axes form the plane being measured. The transformation relationship between the two coordinate systems can be expressed by a rotation matrix. Translation vector The relationship is as follows.

[0036]

[0037] in The coordinates of a three-dimensional point in the world coordinate system. These are coordinates in the turntable coordinate system. Rotation matrix. The solution is based on two coordinate systems z The unit vector of the axis and the coordinates of the axis center can be used to achieve this. The solved rotation matrix and translation vector will be used in the subsequent mutual transformation between the turntable coordinate system and the world coordinate system.

[0038] Based on transferring the center of each feature point from the world coordinate system to the turntable coordinate system, each feature point is rotated within the turntable coordinate system. According to the Rodriguez formula, the transformation matrix of the feature point rotating by a certain angle in the turntable coordinate system is shown below.

[0039]

[0040] The transformed coordinates are:

[0041] After solving for the coordinates of the feature points in the turntable coordinate system for the next frame, they can be transferred to the world coordinate system using the rotation matrix and translation vector between the turntable and the world coordinate system. These coordinates are the theoretical 3D world coordinates of the feature points when the turntable rotates to the next angle. The coordinate transformation formula is as follows.

[0042]

[0043] in This is the inverse of the rotation matrix (the inverse of an orthogonal matrix is ​​equal to its transpose). Based on the ideal 3D coordinates in the next frame, the attitude optimization of the turntable axis can be achieved by constructing an objective function through reprojection error calculation.

[0044] Since the binocular camera system of this invention uses the optical center of the left camera after stereo correction as the starting point of the world coordinate system, the theoretical coordinates of the marker point in the epipolar-corrected image at the next rotation angle can be obtained by multiplying the coordinates of the world coordinate system by the epipolar-corrected intrinsic parameter matrix. The loss function in the LM algorithm can be constructed using this coordinate quantity, and the calculation formula is as follows.

[0045]

[0046] in, p theory =[ u , v ,1] T These are the pixel coordinates after theoretical epipolar correction. K rect This represents the intrinsic parameter matrix after epipolar correction, used to project 3D coordinates in the world coordinate system into pixel coordinates. The reprojection error, calculated from this theoretical coordinate and the actual pixel coordinates at the next angle, yields the objective function for the LM algorithm, as follows.

[0047]

[0048] In the above objective function, p il , p irThese represent the actual pixel coordinates of the left and right cameras when the turntable rotates to the corresponding angles. p theoryl , p theoryr This represents the theoretical position of the turntable at this angle. The objective function can be derived from the deviation between the theoretical and actual values. Based on the objective function obtained above, an approximate Hessian matrix is ​​constructed as ( J T J+λI ),in I The identity matrix is ​​used. The update Δ of the parameter vector can be obtained by solving the linear system equations based on the approximate Hessian matrix. d .

[0049]

[0050] Update parameter vector:

[0051] By setting appropriate iteration thresholds and iteratively applying the LM algorithm, the precise calculation of the turntable axis equation can be completed, thereby reducing blade parameter errors caused by axis deviation during actual measurement. The final turntable axis calculated using the LM algorithm is shown below. Figure 2 As shown.

[0052] Figure 2 The red axis in the image represents the initial axis obtained through linear calculation, while the green axis represents the optimized turntable axis. The image shows that the positions of the two axes are basically located at the center of each trajectory, consistent with theoretical expectations. In the axis fitting process of this invention, the objective function value is optimized to 0.057 pixels, meeting the accuracy requirements. Based on this axis fitting, multi-view blade point cloud stitching can be achieved.

[0053] Accordingly, this invention proposes a calibration device for a large-scale rotating equipment blade measurement system based on LM algorithm optimization: The device includes a coordinate system transformation module, a trajectory correction module, a projection module, an error quantization module, an axis optimization module, and a stitching module; The coordinate system transformation module transforms the world coordinate system coordinates to the turntable coordinate system, restoring the trajectory of the marker point to a standard circle; The trajectory correction module is used for rotation within the turntable coordinate system. Based on the Rodriguez formula, it generates the theoretical position of the marker point in the next frame, ensuring that the rotation trajectory conforms to the actual circular motion law. The projection module is used for coordinate inverse transformation and projection, to perform theoretical position calculation, and to output theoretical pixel coordinates as an error comparison benchmark. The error quantization module is used to calculate the reprojection error, calculate the reprojection error between the theoretical position and the actual position, and quantify the turntable axis calibration error. The axis optimization module optimizes the objective function value according to the LM algorithm to obtain the optimal solution for the axis of the turntable; The stitching module is used for stitching blade point clouds; using the optimized axis as the rotation reference, it integrates multi-view acquired three-dimensional point cloud data of the blade to achieve sub-pixel-level precision three-dimensional reconstruction of the blade surface, thus solving the problem of blade parameter measurement error.

[0054] An electronic device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the above method.

[0055] A computer-readable storage medium for storing computer instructions that, when executed by a processor, implement the steps of the above-described method.

[0056] The memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM). It should be noted that the memory of the methods described in this invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0057] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. 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 via wired means such as coaxial cable, optical fiber, digital subscriber line, DSL, or wireless means such as infrared, wireless, microwave, etc. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium such as a floppy disk, hard disk, magnetic tape; an optical medium such as a high-density digital video disc, DVD; or a semiconductor medium such as a solid-state disk, SSD, etc.

[0058] During implementation, each step of the above method can be completed by an integrated logic circuit of the hardware in the processor or by instructions in the form of software. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as being executed by a hardware processor, or can be executed by a combination of hardware and software modules in the processor. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in conjunction with its hardware. To avoid repetition, it will not be described in detail here.

[0059] It should be noted that the processor in the embodiments of this application can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method embodiments can be completed by the integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied as execution by a hardware decoding processor, or as execution by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory; the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above methods.

[0060] The calibration method for a large rotary equipment blade measurement system based on the LM algorithm optimization proposed in this invention has been described in detail above. The principles and implementation methods of this invention have been explained. The above description of the embodiments is only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. An optimization method for an aero-engine blade measurement system based on the LM algorithm, characterized in that: The method specifically includes the following steps: Step 1: Coordinate system transformation. Transform the world coordinate system to the turntable coordinate system to restore the trajectory of the marker point to a standard circle. Step 2: Rotate within the turntable coordinate system. Based on the Rodriguez formula, generate the theoretical position of the marker point for the next frame to ensure that the rotation trajectory conforms to the actual circular motion law. Step 3: Perform inverse coordinate transformation and projection to calculate the theoretical position and output the theoretical pixel coordinates; Step 4: Calculate the reprojection error, calculate the reprojection error between the theoretical position and the actual position, and quantify the turntable axis calibration error; Step 5: Optimize the objective function value using the LM algorithm to obtain the optimal solution for the turntable axis; Step 6, blade point cloud stitching: Using the optimized axis as the rotation reference, the three-dimensional point cloud data of the blade collected from multiple perspectives are fused to achieve three-dimensional reconstruction of the blade surface with sub-pixel accuracy, thus solving the problem of blade parameter measurement error.

2. The optimization method according to claim 1, characterized in that: In step 1, let the initial axis direction of the turntable in the world coordinate system be ( x 0, y 0, z The direction vector of this axis in the turntable coordinate system is (0,0,1). x shaft and y The axes form the plane being measured; the transformation relationship between the two coordinate systems can be expressed by a rotation matrix. Translation vector The relationship is as follows; in The coordinates of a three-dimensional point in the world coordinate system. Coordinates in the turntable coordinate system; rotation matrix The solution is based on two coordinate systems z The unit vector of the axis and the coordinates of the axis center can be used to achieve this. The solved rotation matrix and translation vector will be used in the subsequent mutual transformation between the turntable coordinate system and the world coordinate system.

3. The optimization method according to claim 2, characterized in that: In step 2, after transferring the centers of each feature point from the world coordinate system to the turntable coordinate system, each feature point is rotated within the turntable coordinate system. According to the Rodriguez formula, the transformation matrix of the feature point under a certain rotation angle α in the turntable coordinate system is shown below: The transformed coordinates are: 。 4. The optimization method according to claim 3, characterized in that: In step 3, after solving for the coordinates of the feature points in the turntable coordinate system for the next frame, the coordinates can be transferred to the world coordinate system based on the rotation matrix and translation vector between the turntable and the world coordinate system. These coordinates are the theoretical 3D world coordinates of the feature points when the turntable rotates to the next angle. The coordinate transformation formula is as follows: in It is the inverse of the rotation matrix.

5. The optimization method according to claim 4, characterized in that: In step 4, by left-multiplying the epipolar-corrected intrinsic parameter matrix based on the world coordinate system coordinates, the theoretical coordinates of the marker point in the epipolar-corrected image at the next rotation angle can be obtained. This coordinate system can then be used to construct the loss function in the LM algorithm, calculated as follows: in, p theory =[ u , v ,1] T These are the pixel coordinates after theoretical epipolar correction. K rect This represents the intrinsic parameter matrix after epipolar correction, used to project three-dimensional coordinates in the world coordinate system into pixel coordinates; The objective function of the LM algorithm is obtained by calculating the reprojection error with the actual pixel coordinates at the next angle, as follows: In the above objective function, p il , p ir These represent the actual pixel coordinates of the left and right cameras when the turntable rotates to the corresponding angles. p theoryl , p theoryr This indicates the theoretical position of the turntable at this angle. The objective function can be derived based on the deviation between the theoretical and actual values.

6. The optimization method according to claim 5, characterized in that: In step 5, based on the objective function obtained above, an approximate Hessian matrix is ​​constructed as ( J T J+λI ),in I It is the identity matrix; The update Δ of the parameter vector can be solved by solving the linear system equations using the approximate Hessian matrix. d ; Update parameter vector: By setting appropriate iteration thresholds and using the LM algorithm to iterate continuously, the precise calculation of the turntable axis equation can be completed, thereby reducing the blade parameter error caused by axis deviation during actual measurement.

7. The optimization method according to claim 6, characterized in that: In step 6, the turntable axis calculated by the LM algorithm is used as the rotation center to transform the point clouds at different angles to the same coordinate system, align overlapping areas, remove redundant points, and generate a complete blade surface model.

8. A calibration device for a large rotary equipment blade measurement system based on LM algorithm optimization, characterized in that: The optimization device is used to perform the optimization method according to any one of claims 1 to 7; The device includes a coordinate system transformation module, a trajectory correction module, a projection module, an error quantization module, an axis optimization module, and a stitching module; The coordinate system transformation module transforms the world coordinate system coordinates to the turntable coordinate system, restoring the trajectory of the marker point to a standard circle; The trajectory correction module is used for rotation within the turntable coordinate system. Based on the Rodriguez formula, it generates the theoretical position of the marker point in the next frame, ensuring that the rotation trajectory conforms to the actual circular motion law. The projection module is used for coordinate inverse transformation and projection, to calculate the theoretical position, and to output the theoretical pixel coordinates. The error quantization module is used to calculate the reprojection error, calculate the reprojection error between the theoretical position and the actual position, and quantify the turntable axis calibration error. The axis optimization module optimizes the objective function value according to the LM algorithm to obtain the optimal solution for the axis of the turntable; The splicing module is used for splicing leaf point clouds; Using the optimized axis as the rotation reference, and integrating multi-view acquired 3D point cloud data of the blade, subpixel-level precision 3D reconstruction of the blade surface is achieved, solving the problem of blade parameter measurement error.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium for storing computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the method according to any one of claims 1 to 7.

Images