Complex curved surface normal follow-up mechanism used in high-pressure container and point cloud normal vector solving method of complex curved surface normal follow-up mechanism

Abstract

The invention belongs to the technical field of precision measurement in a high-pressure environment, and particularly relates to a complex curved surface normal follow-up mechanism used in a high-pressure container and a point cloud normal vector solving method of the complex curved surface normal follow-up mechanism. Power is transmitted to a movement mechanism in the container through the sealing transmission shaft; 2, the cable cabin-penetrating sealing isolation module adopts multi-channel combination to realize high-pressure cabin-penetrating sealing of a power line and a signal line; the mechanical arm operating system adopts a multi-axis movement mechanism and comprises a W axis, a B axis, a C axis, a Z axis, an X axis, a D axis and an F axis; the mechanism can ensure that the scanning probe is always perpendicular to a measured complex curved surface, and is a key functional unit for realizing high-precision and non-contact three-dimensional scanning and modeling in a high-pressure closed container.

Description

Technical Field

[0001] This invention belongs to the field of precision measurement technology under high pressure environments, specifically relating to a complex curved surface normal follower mechanism and its point cloud normal vector solution method for use in high-pressure containers. This mechanism ensures that the scanning probe is always perpendicular to the complex curved surface being measured, and is a key functional unit for realizing high-precision, non-contact 3D scanning and modeling within high-pressure sealed containers. Background Technology

[0002] High-pressure vessels are widely used in aerospace, military, and energy fields for structural performance testing under simulated high-pressure environments. For example, the deformation and sealing performance of complex curved surfaces under high pressure require precise measurement. However, performing scanning measurements under high-pressure environments presents several technical challenges: 1. High-voltage isolation issue: Servo motors, reducers and other power components need to work in high-voltage environments, and their sealing and reliability requirements are extremely high; 2. Cable penetration sealing: Signal lines and power lines need to pass through the wall of the high-pressure vessel. Traditional sealing methods are prone to leakage or signal interference. 3. Motion mechanism issues: Measurement of complex curved surfaces requires the scanning probe to not only reach any point in space, but also to accurately align the measurement direction with the surface normal at each point. This places extremely high demands on the degree of freedom and accuracy of the motion mechanism.

[0003] 4. Scanning path planning: This mechanism requires a multi-axis coordinated motion system with high-precision positioning and attitude adjustment capabilities. It needs to achieve automated and high-precision scanning paths under high-pressure environments to ensure data integrity and scanning accuracy.

[0004] Currently, most common scanning measurement equipment is used in normal pressure environments, lacking dedicated devices for scanning inside high-pressure vessels. Existing high-pressure scanning technologies have significant shortcomings: for example, some employ single-axis or three-axis motion structures, which cannot adapt to the normal tracking requirements of complex curved surfaces; others use hydraulic drives, which, while solving some sealing problems, suffer from slow response speed and low control accuracy. Regarding sealing, existing high-pressure chamber penetration technologies mostly use single O-rings or stuffing boxes for sealing, which are prone to aging and wear over long-term use, leading to seal failure. In terms of measurement methods, traditional methods often rely on preset paths or manual intervention, lacking the ability to adapt to curved surface shapes, making it difficult to achieve fully automatic, high-precision normal tracking measurements in high-pressure, confined environments. Summary of the Invention

[0005] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0006] In view of the aforementioned existing problems, the present invention is proposed.

[0007] Therefore, the technical problem solved by the present invention is to provide a complex curved surface normal follower mechanism for use in high-pressure containers, which solves the problems of poor sealing, low motion control accuracy and unreasonable path planning in the prior art when scanning under high pressure.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a complex curved surface normal follower mechanism for use in high-pressure containers, comprising: 1. a motor reducer sealed chamber, which adopts a mechanical seal structure, placing the motor and reducer inside the sealed chamber, and transmitting power to the internal motion mechanism of the container through a sealed transmission shaft; 2. a cable penetration sealed isolation module, which adopts a multi-channel combination to achieve high-voltage penetration sealing of power lines and signal lines; 3. a robotic arm operating system, which adopts a multi-axis motion mechanism, including W-axis, B-axis, C-axis, Z-axis, X-axis, D-axis, and F-axis. Driven by a servo motor, it achieves seven degrees of freedom motion control, enabling precise scanning of components within the high-pressure container and achieving full coverage of the scanning head within the high-pressure container; 4. a measurement system, based on a vision-guided path planning algorithm, which controls the measurement system to perform image acquisition and 3D reconstruction of complex toroidal surfaces. A "binocular camera + structured light" method is used to perform 3D scanning and reverse modeling of the surface of the workpiece under test, forming a 3D point cloud map; the point cloud map's normal vector is solved through local plane fitting.

[0009] Furthermore, the motor reducer sealing chamber is a separate enclosure in which the servo motor and reducer assembly are placed using a mechanical seal. The stationary ring is mounted on the sealing chamber housing and sealed with a sealing ring, while the rotating ring is similarly sealed to the shaft using a sealing ring and connected to the shaft with a set screw.

[0010] Furthermore, the cable penetration sealing module is designed as a cable penetration sealing and isolation module, using a low-carbon steel flange with an array of through holes inside, through which sealing joints are installed. Each cable through hole contains multiple layers of flexible seals (such as fluororubber), which are axially compressed by the tapered flange for sealing. The sealing module adopts a row-style design, allowing multiple sets of cables to be inserted side-by-side.

[0011] Furthermore, the W-axis of the robotic arm operating system connects the slide and slider on a linear guide rail to realize the in / out operation.

[0012] Furthermore, the B-axis of the robotic arm operating system is located on the W-axis slide and mainly consists of a worktable, a slewing bearing structure, and a drive mechanism. The worktable is located above the slewing bearing structure, thereby enabling the rotation of the worktable.

[0013] Furthermore, the robotic arm operating system comprises a C-axis, a Z-axis, and an X-axis. The C-axis adopts a ring structure, including a ring-shaped gantry, a ring-shaped sliding guide rail, a drive gear ring, and a drive system. The measuring system can move along the ring, achieving rotational alignment around the workpiece at any angle. The Z-axis includes a drive system, a guide rail, a slider, and a gear rack. The guide rail and rack are arranged on the W-axis slide and connected to the C-axis ring-shaped gantry, enabling the measuring mechanism to move axially along the pressure vessel. The X-axis is a radial feed axis, combined with the C-axis to form a polar coordinate system. Through polar coordinate calculations, the measuring sensor can be driven to any specified position within the axial cross-section of the workpiece.

[0014] Furthermore, regarding the D-axis and F-axis, considering the randomness of the normal vector on the surface of the workpiece, this measurement mechanism requires the testing system to have the ability to track the coordinates and normal vector of any point on the workpiece surface. The sensor, driven by the Z / X / C axes, can reach any position in the test space, but cannot guarantee that the measurement direction is aligned with the normal vector of the workpiece surface. Adjusting the rotation of the D / F axes allows the measurement direction of the system to achieve any configuration in the Cartesian coordinate system. The D-axis is used to adjust the pitch angle of the measurement system, controlled by a servo motor. The F-axis is used to adjust the rotation angle of the measurement system, controlled by a servo motor.

[0015] Furthermore, an efficient method for solving point cloud normal vectors is provided. The core of the measurement system lies in this efficient method, which is integrated into a complete sensor alignment system. This system mainly includes the following modules: a 3D data acquisition module; a point cloud preprocessing and modeling module; a normal vector solving module; and an alignment decision and control module. The normal vector solving module is based on the Principal Component Analysis (PCA) method, and its specific steps are as follows: Utilizing efficient spatial data structures such as KD-Tree, a point cloud normal vector is calculated... Find its k nearest neighbors to form a local neighborhood set. : The distance between points is usually calculated using Euclidean distance. The second step is covariance analysis: calculating the neighborhood point set. The covariance matrix C describes the distribution of neighborhood points in three-dimensional space. Calculate the neighborhood... The centroid (mean point) μ: ; Calculate the 3×3 covariance matrix C of this neighborhood: ; Step 3: Principal Component Analysis (PCA): Eigenvalue decomposition of the covariance matrix C. , ; Where the eigenvalues ​​satisfy , This is the corresponding feature vector.

[0016] Determine the normal vector: minimum eigenvalue Corresponding feature vector That is the point normal vector ; ; The alignment decision and control module described herein uses the solved average normal vector of the target area. The D-axis and F-axis are controlled for pose adjustment. The D-axis, driven by a servo motor, adjusts the pitch angle of the measurement system; the F-axis, also driven by a servo motor, adjusts the rotation angle. Through the coordinated movement of these two axes, the direction of the measurement optical axis is aligned with the normal vector of the workpiece surface. Precise alignment ensures that the measurement direction is always perpendicular to the surface being measured.

[0017] Beneficial effects of the invention: Compared with existing technologies, this invention realizes a fully automated, high-precision robotic arm system for measurement under high-pressure environments. Through seven-axis coordinated motion, combined with binocular vision-based reverse modeling technology, this system can automatically acquire and adapt to the complex geometry of free-form workpieces, ensuring that the measuring sensors are always precisely aligned with the measured point (e.g., normal alignment). This fundamentally solves the technical challenge of accurately measuring complex workpieces inside high-pressure containers. This invention not only significantly improves the accuracy, reliability, and efficiency of measurement, but also ensures long-term safe operation of the equipment under high-pressure environments through unique designs such as a sealed chamber and a system with multiple cable-through-the-chamber sealed isolation modules, combining high intelligence and engineering applicability. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall structure of a complex curved surface normal follower mechanism for use in a high-pressure vessel according to the present invention.

[0019] Figure 2 This is a schematic diagram of the operating system of a robotic arm for a complex curved surface normal follower mechanism in a high-pressure vessel according to the present invention.

[0020] Figure 3 This is a schematic diagram of the sealed chamber structure of a motor reducer for a complex curved surface normal follower mechanism in a high-pressure vessel according to the present invention.

[0021] Figure 4 This is a cross-sectional view of the sealing chamber of a motor reducer for a complex curved surface normal follower mechanism in a high-pressure vessel according to the present invention.

[0022] Figure 5This is a schematic diagram of the cable penetration sealing and isolation module for a complex curved surface normal follower mechanism in a high-pressure vessel according to the present invention.

[0023] Figure 6 This is a schematic diagram of the B-axis slewing bearing structure for a complex curved surface normal follower mechanism in a high-pressure vessel according to the present invention.

[0024] Figure 7 This is a schematic diagram of the F-axis and D-axis structure of a complex curved surface normal follower mechanism for use in a high-pressure vessel according to the present invention.

[0025] Figure 8 This is a schematic diagram of the X-axis structure of a complex curved surface normal follower mechanism for use in a high-pressure vessel according to the present invention.

[0026] Figure 9 This invention relates to a 3D camera scanning system for a complex curved surface normal follower mechanism inside a high-pressure vessel.

[0027] Figure 10 This is a schematic diagram of the normal vector flow of a complex curved surface normal follower mechanism for use in a high-pressure vessel according to the present invention.

[0028] The above figures include the following reference numerals: 1. Hydraulic auxiliary support platform; 2. Pressure vessel base platform; 3. Pressure vessel tank; 4. Cable penetration sealing and isolation module flange; 5. W-axis linear guide; 6. W-axis auxiliary support roller; 7. W-axis in / out slide; 8. Z-axis linear guide; 9. Z-axis slider; 10. C-axis annular gantry; 11. C-axis annular drive gear ring; 12. C-axis annular sliding guide; 13. X-axis 100×200 aluminum profile; 14. X-axis drive system; 15. C-axis drive system; 16. C-axis gear; 17. F-axis drive system; 18. Square block; 19. D-axis drive system; 20. Z-axis drive system; 21. Z-axis rack; 22. B-axis worktable; 23. W-axis rack; 24. W-axis drive system; 25. Square L 26. Shaped plate; 27. 3D camera scanning system; 28. Gear; 29. ​​Gear fixing plate; 30. Drive shaft; 31. Silicon carbide sealing ring (moving ring); 32. Nitrile rubber sealing ring (moving ring); 33. Silicon carbide sealing ring (stationary ring); 34. Nitrile rubber sealing ring (stationary ring); 35. Angular contact ball bearing; 36. Bearing cover; 37. 45 steel tempered sealing chamber shell; 48. Servo motor; 49. Planetary reducer; 40. Sealing cover; 41. Expansion connecting sleeve; 42. B-axis slewing bearing; 43. B-axis drive system; 44. B-axis gear; 45. Circular L-shaped plate; 46. Connecting plate; 47. Annular guide roller; 48. X-axis linear guide; 49. X-axis rack; 50. Through-chamber flange; 51. Flange base; 52. Proportional valve. Detailed Implementation

[0029] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0030] Example 1. A complex curved surface normal follower mechanism for use in high-pressure vessels 1. Implementation method of sealed chamber for motor reducer: like Figure 3 and Figure 4 As shown, the motor reducer sealing chamber structure includes a gear 27, a gear fixing plate 28, a transmission shaft 29, a silicon carbide sealing ring (moving ring) 30, a nitrile rubber sealing ring (moving ring) 31, a silicon carbide sealing ring (stationary ring) 32, a nitrile rubber sealing ring (stationary ring) 33, an angular contact ball bearing 34, a bearing cover 35, a 45 steel modified sealing chamber shell 36, a servo motor 37, a planetary reducer 38, a sealing cover 39, and a tightening connecting sleeve 40. The servo motor 37 and planetary reducer 38 are connected and placed inside the sealed chamber housing 36. The shaft of the reducer 38 is connected to the expansion connecting sleeve 40 and to the drive shaft 29. The angular contact ball bearing 34 is pressed and installed onto the drive shaft 29 with the bearing cover 35 and pressed against the top of the chamber. The drive shaft 29 passes through the sealed chamber housing 36. A silicon carbide sealing ring (stationary ring) 32 is installed at the top of the housing with an overfit. The stationary ring 32 and the housing 36 are sealed with a nitrile rubber sealing ring 33. Then, a silicon carbide sealing ring (moving ring) 30 is installed. The end faces of the two rings are tightly fitted and rotate relative to each other. The moving ring 30 is sealed with a nitrile rubber sealing ring 31 and connected to the shaft with a set pin.

[0031] 2. Implementation method of cable penetration sealing module: like Figure 1 and 5 As shown, the cable penetration sealing isolation module has regular through holes drilled in the pressure vessel wall according to the number and diameter of the cables, and a flange base 50 and a penetration flange 49 are welded together. A rubber gasket is placed between the penetration flange 49 on the pressure vessel wall and the cable penetration sealing isolation module flange 4. The cable penetration sealing isolation module flange 4 is aligned and bolted to the penetration flange 49, completing the first layer of static sealing of the pressure vessel wall. The cables are then passed through the pre-drilled through holes on the module in sequence. For each cable, a pressure adjusting nut, a two-piece clamping ring, and a tapered sealing flange are inserted in sequence. The tapered sealing flange is pushed into the cable through hole, and then the pressure adjusting nut is tightened with a wrench. The nut pushes the clamping ring, squeezing the tapered flange, causing it to axially compress and tightly wrap the cable, completing the second layer of dynamic sealing of the cable. A torque wrench is used to ensure that the tightening torque of all nuts is consistent.

[0032] 3. Implementation methods for each axis of the robotic arm operating system: W-axis: such as Figure 2The illustrated robotic arm operating system for a complex curved surface normal follower mechanism within a high-pressure vessel includes a W-axis drive system 24, two W-axis linear guides 5, six W-axis auxiliary support rollers 6, a W-axis rack 23, and a W-axis in / out slide 7. A base platform 2 is located at the bottom of the pressure vessel to support the entire experimental system. Two linear guides 5 are laid on the base platform and connected to the W-axis in / out slide 7 via guide rail sliders. The drive system 24 drives the gears to rotate, and the gears mesh with the rack 23, causing the W-axis in / out slide 7 to move axially and perform in / out operations along the linear guide system. Auxiliary support rollers 6 are located on the lower part of the in / out side of the W-axis in / out slide. After the slide 7 is removed from the pressure vessel tank 3, it is supported by these auxiliary support rollers on a hydraulic auxiliary support platform 1 outside the pressure vessel tank, ensuring the rigidity of the support after the slide is removed. The W-axis stroke is 3000 mm, its speed is 10 m / min, and its accuracy can reach ±0.1 mm.

[0033] B-axis: as shown Figure 2 and Figure 6 The B-axis worktable 22 shown is located on the W-axis slide 7. The worktable 22 is provided with M8 evenly distributed threaded holes for fixing the test piece fixture. The slewing bearing 41 is connected below it. The slewing bearing 41 is made of Fangyuan 013-30-710. The B-axis drive system 42 drives the pinion to rotate. The pinion meshes with the gear ring on the slewing bearing 41 and rotates around it, which can achieve ±30° rotational displacement. Its speed can reach 10 rpm and the accuracy is ±0.01°.

[0034] C-axis: as shown Figure 2 As shown, the C-axis has a ring structure. A dedicated ring sliding guide rail 12 and a ring drive gear ring 11 are installed on the C-axis ring gantry 10. The gear 16 is connected to the C-axis drive system and meshes with the drive gear ring 11, so that the measuring system rotates around the axis of the workpiece being measured to meet the testing requirements of different angles. Its working stroke is ±135°, the speed is 5rpm, and the accuracy is ±0.01°.

[0035] Z-axis: such as Figure 2 The Z-axis shown is driven by the Z-axis drive system 20, which drives the gear. The gear meshes with the Z-axis rack 21 on the W-axis slide 7. The Z-axis drive system 20 and the gear are installed on one side of the C-axis annular gantry 10. When the mechanism moves, it drives the C-axis annular mechanism to move axially along the Z-axis linear guide 8. The Z-axis stroke is 3500mm, the speed is 20m / min, and the accuracy is ±0.03°.

[0036] X-axis: such as Figure 8As shown, a 100×200 aluminum profile 13 is used. An X-axis linear guide 47 and an X-axis rack 48 are mounted on one side of the aluminum profile 13. An annular guide roller 46 is mounted on a connecting plate 45 and fixed to the C-axis annular sliding guide 12. Through the meshing of the X-axis rack and pinion 48, precise adjustment of the distance between the measuring system and the workpiece is achieved. X and C form a polar coordinate system. The polar coordinate calculation steps are: obtaining the Cartesian coordinates (x, y, z) of a point P on the surface of the workpiece, where the polar radius (r) is calculated. Here, r directly corresponds to the displacement along the X-axis. Calculate the polar angle (θ): The Z-axis independent control sensor positions itself longitudinally along the workpiece axis, directly acquiring the z-coordinate value of point P. The control system generates motion commands for each axis based on the calculated (r, θ, z) values. X-axis: Moved to a radial distance r (mm) C-axis: Rotate to polar angle θ (°) Z-axis: Move to axial position z (mm) The X-axis travel is 500mm, the speed is 10m / min, and the accuracy is ±0.03mm.

[0037] The D-axis and F-axis are as follows Figure 7 The F-axis and D-axis are connected to the bottom of the aluminum profile 13 of the X-axis via a square block 18. The F-axis drive system 17 is connected to the square block 18, and the drive shaft 29 passes through the square block 18 and connects to the circular L-shaped plate 44. The circular L-shaped plate 44 is connected to the D-axis drive system 19, which in turn is connected to the camera scanning system 26 via the square L-shaped plate 25. The F-axis controls the rotation angle adjustment of the camera scanning system, and the D-axis controls the pitch angle adjustment of the camera scanning system. Both the F-axis and D-axis have a travel of 360°, a speed of 50 rpm, and an accuracy of ±0.01°.

[0038] Example 2. A method for solving point cloud normal vectors. 4. Measurement System and Implementation Method for Normal Vector Calculation: like Figure 9 The specific steps of a camera scanning device for a complex curved surface normal follower mechanism inside a high-pressure vessel are as follows: 1. Using a high-resolution (6.3 million pixels × 2) binocular camera, combined with a structured light projector, scan the surface of the workpiece to be measured. Acquire dense three-dimensional point cloud data of the workpiece surface. , where each point Dense 3D point cloud data of the workpiece surface is acquired within a scanning distance of 500-800mm. A single scan can acquire approximately 1.5-2 million 3D points, achieving a point cloud density of 8-12 points / mm², with a positioning accuracy better than ±0.05mm. 2. [The text abruptly shifts to a different topic:] ...for the original point cloud... Preprocessing such as denoising and filtering is performed, and the point clouds from multiple scans are stitched together into a complete 3D model of the workpiece using a high-precision stitching algorithm. 3. For the point cloud within the target measurement area, its average normal vector is calculated to represent the surface orientation of that area, such as... Figure 10 The process for solving the normal vector shown begins with a neighborhood search: utilizing efficient spatial data structures such as KD-Tree, for points... Find its k=32 nearest neighbors to form a local neighborhood set. : The distance between points is usually calculated using Euclidean distance. The second step is covariance analysis: calculating the neighborhood point set. The covariance matrix C describes the distribution of neighborhood points in three-dimensional space. Calculate the neighborhood... The centroid (mean point) μ: ; Calculate the 3×3 covariance matrix C of this neighborhood: ; Step 3: Principal Component Analysis (PCA): Eigenvalue decomposition of the covariance matrix C. , ; Where the eigenvalues ​​satisfy , This is the corresponding feature vector.

[0039] Determine the normal vector: minimum eigenvalue Corresponding feature vector That is the point normal vector ; ; In actual testing, for gently curving surfaces with a radius of curvature greater than 50mm, the accuracy of normal vector calculation can reach ±0.5°; for complex surfaces with a radius of curvature of 10-50mm, the accuracy is ±1.5°. 4. The average normal vector calculated in the target measurement area represents the average orientation of the workpiece surface in that area. The system compares this normal vector direction with the ideal measurement optical axis direction of the sensor, calculates the pose parameters (rotation and translation) that need adjustment, and drives the actuator to complete precise alignment. The complete cycle time from normal vector calculation to final pose alignment is less than 2.5 seconds, meeting the requirements for efficient measurement within high-pressure vessels.

[0040] 5. Institutional Operation Process: The operational process of this facility is as follows: the pressure vessel is opened under safe conditions; the ground hydraulic auxiliary support platform 1 is lifted; the W-axis slide 7, driven by the W-axis drive system 24, moves out of the pressure vessel tank 3, removing 3000mm, and is then supported on the ground hydraulic auxiliary support; the test piece is hoisted onto the B-axis worktable 22; reflective markings are affixed to the surface of the test piece; the W-axis slide 7 moves into the designated position inside the pressure vessel tank 3; the robotic arm operating system is activated, and the Z / X / C three-axis adjustment 3D camera scanning system 26 performs reverse modeling of the test piece surface, and uses point cloud processing software to obtain the coordinates and normals of any point on the test piece surface; W-axis Slide 7 is removed from pressure vessel tank 3, reflective markings are manually removed, and W-axis slide 7 is then moved back into pressure vessel tank 3; the pressure vessel quick-opening door is closed, and the ground hydraulic auxiliary support platform 1 is lowered; the pressure vessel is filled with gas: first, compressed air is filled to 2.0 MPa, then a mixture of argon and helium is filled, and compressed air is continued to be filled to 3.2 MPa to ensure uniform gas mixing; the test is started, and the tank is filled with gas through 12 6mm through-chamber gas pipes via proportional valve 51 (filling speed is 6m / s), and the pressure fluctuation inside the tank is controlled to be less than 1% by proportional valve 51 at the top of tank 3; multiple tests can be performed based on the test results; after the test is completed, the compressed air inside the tank is discharged.

Claims

1. A complex curved surface normal follower mechanism for use in high-pressure vessels, characterized in that... include: The motor reducer sealed chamber adopts a mechanical seal structure, placing the motor and reducer inside the sealed chamber, and transmitting power to the internal moving mechanism of the container through a sealed transmission shaft; The cable penetration sealing and isolation module employs a multi-channel combination to achieve high-voltage sealing of power lines and signal lines through the high-pressure container. The robotic arm operating system utilizes a multi-axis motion mechanism, including W-axis, B-axis, C-axis, Z-axis, X-axis, D-axis, and F-axis. Driven by servo motors, it achieves seven degrees of freedom motion control, enabling precise scanning of components within the high-pressure container and achieving full coverage of the scanning head within the container. The measurement system, based on a vision-guided path planning algorithm, controls the measurement system to acquire images and perform 3D reconstruction of complex toroidal surfaces. A "binocular camera + structured light" approach is used to perform 3D scanning and reverse modeling of the workpiece surface, generating a 3D point cloud map. The point cloud map's normal vector is solved through local plane fitting.

2. The complex curved surface normal follower mechanism for use in a high-pressure vessel according to claim 1, characterized in that... The motor reducer sealing chamber is a separate chamber housing in which the servo motor and reducer assembly are placed using a mechanical seal. The stationary ring is installed on the sealing chamber housing and sealed with a sealing ring. The rotating ring is also isolated and sealed with a sealing ring between itself and the shaft, and is connected to the shaft with a set screw.

3. A complex curved surface normal follower mechanism for use in a high-pressure vessel according to claim 1, characterized in that... The cable penetration sealing module uses a low-carbon steel flange with an array of through holes inside, through which sealing joints are installed. Each cable through hole is equipped with multiple layers of flexible seals, which are axially compressed by the tapered flange for sealing. The sealing module adopts a row-style design, allowing multiple sets of cables to be inserted side by side.

4. A complex curved surface normal follower mechanism for use in a high-pressure vessel according to claim 1, characterized in that... The W-axis of the robotic arm operating system connects the slide table and slider and places them on a linear guide rail to realize the in / out operation.

5. A complex curved surface normal follower mechanism for use in a high-pressure vessel according to claim 1, characterized in that... The B-axis of the robotic arm operating system is located on the W-axis slide and mainly consists of a worktable, a slewing bearing structure, and a drive mechanism. The worktable is located above the slewing bearing structure, thereby enabling the rotation of the worktable.

6. A complex curved surface normal follower mechanism for use in a high-pressure vessel according to claim 1, characterized in that... The robotic arm operating system comprises a C-axis, a Z-axis, and an X-axis. The C-axis uses a ring structure, including a ring gantry, a ring sliding guide, a drive gear ring, and a drive system. The measuring system can move along the ring to achieve rotational alignment around the workpiece at any angle. The Z-axis includes a drive system, a guide rail, a slider, and a gear rack. The guide rail and rack are arranged on the W-axis slide and connected to the C-axis ring gantry, enabling the measuring mechanism to move axially along the pressure vessel. The X-axis is a radial feed axis, combined with the C-axis to form a polar coordinate system. Through polar coordinate calculations, the measuring sensor can be driven to any specified position within the axial cross-section of the workpiece.

7. A complex curved surface normal follower mechanism for use in a high-pressure vessel according to claim 1, characterized in that... Considering the randomness of the normal vector on the surface of the workpiece, the measurement mechanism requires the testing system to have the ability to track the coordinates and normal vector of any point on the workpiece surface. The sensor, driven by the Z / X / C axes, can reach any position in the test space, but cannot guarantee that the measurement direction is aligned with the normal vector on the workpiece surface. Adjusting the rotation of the D and F axes allows the measurement direction of the system to achieve any configuration in the Cartesian coordinate system. The D-axis is used to adjust the pitch angle of the measurement system, controlled by a servo motor. The F-axis is used to adjust the rotation angle of the measurement system, controlled by a servo motor.

8. An efficient method for solving point cloud normal vectors, and its integration into a complete sensor alignment system; the system mainly includes the following modules: a 3D data acquisition module; a point cloud preprocessing and modeling module; a normal vector solving module; and an alignment decision and control module; wherein the normal vector solving module is based on the principal component analysis (PCA) method, characterized in that... The specific steps are as follows: The first step is to utilize efficient spatial data structures such as KD-Tree to provide point... Find its k nearest neighbors to form a local neighborhood set. : The distance between points is usually calculated using Euclidean distance. ; Step 2: Covariance Analysis: Calculate the neighborhood point set The covariance matrix C describes the distribution of neighborhood points in three-dimensional space. Calculate the neighborhood... The centroid (mean point) μ: ； Calculate the 3×3 covariance matrix C of this neighborhood: ； Step 3, Principal Component Analysis (PCA): Perform eigenvalue decomposition on the covariance matrix C. ， ； Where the eigenvalues ​​satisfy , The corresponding feature vector; Determine the normal vector: minimum eigenvalue Corresponding feature vector That is the point normal vector ; ； The alignment decision and control module, based on the solved average normal vector of the target area... Control the D-axis and F-axis to adjust the pose; The D-axis is driven by a servo motor to adjust the pitch angle of the measurement system; the F-axis is also driven by a servo motor to adjust the rotation angle of the measurement system. Through the coordinated movement of these two axes, the direction of the measuring optical axis of the measuring system is aligned with the normal vector of the workpiece surface. Precise alignment ensures that the measurement direction is always perpendicular to the surface being measured.

Images