Adaptive Machining Method and Apparatus for Precision Copper Tubes of Various Specifications

Abstract

This invention discloses an adaptive machining device and method for precision copper tubes of various specifications. The device includes an articulated robot, an end effector, a vision system, and a control system. The method includes planning the insertion depth and gripping posture of the copper tube based on its outer diameter and effective gripping length; when the copper tube reaches the insertion depth, controlling the articulated robot to perform a preset tilting motion so that the copper tube slides along a support rod, with axial limiting and centering achieved by a retaining ring; monitoring the relative posture of the end effector and the spindle hole in real time, dynamically adjusting the motion trajectory of the articulated robot; and feeding the copper tube into the spindle hole and clamping it by the machine tool when the end effector and the spindle hole are aligned. After machining is completed, controlling the end effector to align again and insert into the spindle hole, activating the air-blowing chip removal component to blow away copper chips, and transferring the copper tube to the unloading station. This invention can dynamically adjust the robot's motion trajectory to ensure precise alignment between the copper tube and the spindle hole, improving machining accuracy and production efficiency.

Description

Technical Field

[0001] This invention relates to the field of precision machining technology, and in particular to an adaptive machining method and apparatus for multi-specification precision copper tubes. Background Technology

[0002] In the field of precision copper tube machining, traditional loading and unloading methods mainly rely on manual operation or simple robotic grippers. Operators need to manually adjust the fixture position and gripping depth according to the copper tube specifications, which is not only inefficient but also prone to misalignment between the copper tube and the machine tool spindle hole due to inaccurate positioning, resulting in decreased machining accuracy or even equipment damage. Furthermore, switching between different copper tube specifications requires readjusting equipment parameters, leading to poor production flexibility and difficulty in adapting to the needs of multi-variety, small-batch production. Currently, although there are solutions using vision recognition systems for copper tube positioning, they typically only identify the basic positional information of the copper tube and cannot accurately obtain the specification parameters, resulting in a single gripping strategy and an inability to achieve adaptive gripping. Simultaneously, the copper tube is prone to slippage due to gravity during gripping, lacking effective axial limiting and alignment mechanisms, affecting machining accuracy and efficiency. Summary of the Invention

[0003] To address at least one of the aforementioned technical problems, the present invention provides an adaptive processing method and processing apparatus for multi-specification precision copper tubes.

[0004] In a first aspect, the present invention provides an adaptive processing method for multi-specification precision copper tubes, applied to an adaptive processing apparatus for multi-specification precision copper tubes. The processing apparatus includes an articulated robot, an end effector, a vision system, and a control system. The end effector includes loading and unloading stations, a support rod, a retaining ring, and an air-blowing chip removal assembly. The method includes:

[0005] The spatial pose of the copper tube located at the loading station is identified by the vision system. Based on the spatial pose, the outer diameter and effective gripping length of the copper tube are identified, and the insertion depth and gripping pose of the copper tube are adaptively planned.

[0006] The articulated robot is controlled to drive the end effector to insert the copper tube into the insertion depth according to the grasping posture. The articulated robot is controlled to perform a preset tilting action so that the copper tube slides along the support rod under the action of gravity, and the retaining ring realizes axial limitation and centering.

[0007] During the process of the articulated robot carrying the aligned copper tube moving towards the spindle hole of the machine tool, the relative pose of the end effector and the spindle hole is monitored in real time by the vision system, and the movement trajectory of the articulated robot is dynamically adjusted to control the alignment of the end effector and the spindle hole.

[0008] In the aligned state, the copper tube is fed into the spindle hole and clamped by the machine tool. After the processing is completed, the end effector is re-aligned and inserted into the spindle hole. The air blowing chip removal assembly is started to blow away the copper chips, and then the copper tube is transferred to the unloading station.

[0009] Preferably, the step of identifying the spatial pose of the copper tube located at the loading station using a vision system includes:

[0010] The vision control system scans the copper tubes at the loading and unloading stations to acquire depth point cloud data and two-dimensional image data, including the end areas of the copper tubes.

[0011] Anti-reflection processing is performed on the two-dimensional image data to extract the preliminary contour of the copper tube end face;

[0012] The spatial pose of the copper tube is calculated based on the preliminary contour and depth point cloud data.

[0013] Preferably, the step of performing anti-reflective processing on the two-dimensional image data and extracting the preliminary contour of the copper tube end face includes:

[0014] The vision system is controlled to acquire multi-exposure image sequences of the copper tube end under at least two orthogonal polarization states;

[0015] Calculate the values ​​of integrated polarization contrast and local gradient activity based on multi-exposure images;

[0016] The anti-reflective fusion weight is calculated based on the values ​​of comprehensive polarization contrast and local gradient activity.

[0017] An enhanced image is generated by fusing multiple exposure images according to anti-reflective fusion weights, and a preliminary contour of the copper tube end face is extracted based on the enhanced image.

[0018] Preferably, the step of identifying the outer diameter and effective gripping length of the copper tube based on its spatial pose includes:

[0019] Based on the depth point cloud data and spatial pose, the main axis direction of the copper tube is determined, and the depth point cloud data is sliced ​​at equal intervals along the main axis direction to generate a series of cross-sectional point cloud sets.

[0020] For each cross-sectional point cloud set, perform a ring fitting, record all the fitted outer circle radii, and take the most frequent value as the outer diameter of the copper tube.

[0021] Based on the effective distribution length of the deep point cloud data along the main axis, and combined with the preset position of the retaining ring, the effective gripping length of the copper tube is calculated.

[0022] Preferably, the adaptive planning of the copper tube insertion depth and grasping pose includes:

[0023] For length of The copper tube is used to calculate the insertion depth of the copper tube controlled by the end effector. ,satisfy:

[0024] ;

[0025] In the formula, The effective gripping length for copper tubes, This is a preset safety redundancy.

[0026] The initial gripping pose of the end effector is determined based on the spatial pose of the copper tube. The initial gripping pose is used to control the axis of the end effector to coincide with the axis of the copper tube.

[0027] Based on the difference between the outer diameter of the copper tube and the diameter of the support rod of the end effector, the radial offset compensation amount of the end effector during the gripping process is calculated, and the final gripping posture of the end effector is determined according to the initial gripping posture and the radial offset compensation amount.

[0028] Preferably, the step of monitoring the relative pose of the end effector and the spindle hole in real time through a vision system and dynamically adjusting the motion trajectory of the articulated robot includes:

[0029] The system acquires images in real time from the vision system. The images include visual markers fixed to the front end of the support rod and visual targets fixed to the end face of the spindle hole.

[0030] Establish a principal axis coordinate system based on the visual target, and establish an end coordinate system based on the visual marker points;

[0031] The real-time pose of the end-effector coordinate system relative to the principal axis coordinate system is calculated and compared with the pre-calibrated ideal insertion pose to obtain the pose error of the six degrees of freedom.

[0032] The pose error is used as a feedforward compensation to correct the motion trajectory of the articulated robot online.

[0033] Preferably, the activation of the air-blowing cleaning assembly to blow away copper shavings includes:

[0034] After the end effector carries the copper tube into the spindle hole and aligns it with the spindle hole, and before the machine tool clamping device releases the copper tube, the air blowing chip removal component is started.

[0035] The air-blowing chip removal assembly's spraying process includes spraying a high-pressure pulsed airflow to loosen stubborn copper chips adhering to the inner wall of the copper tube and inside the machine tool spindle bore; then switching to a low-pressure fan-shaped airflow to blow the loosened and scattered copper chips towards the opening of the copper tube; wherein,

[0036] During the chip removal process, an acoustic emission sensor installed on the end effector monitors the sound characteristics of the airflow to determine whether the copper chips have been completely removed.

[0037] In a second aspect, the present invention also provides an adaptive processing apparatus for multi-specification precision copper tubes, for implementing the processing method as described in any of the first aspects, the processing apparatus comprising:

[0038] Articulated robots;

[0039] An end effector, fixed to the end of an articulated robot, is equipped with loading and unloading stations for controlling the loading and unloading operations of copper tubes, including:

[0040] Support rod, used to insert into the inner hole of the copper tube;

[0041] A retaining ring and an air-blowing chip removal assembly are provided at the end of the support rod. The retaining ring is used to axially limit the sliding copper tube. The air-blowing chip removal assembly includes a fan-shaped nozzle that communicates with the internal air passage of the support rod.

[0042] Visual aid markers are placed at the front end of the support rod;

[0043] The vision system includes a depth camera mounted on the base of the articulated robot for scanning copper pipes at the loading and unloading station; and a monitoring camera mounted on the CNC machine tool for tracking visual aid markers and machine tool spindle targets.

[0044] The control system is electrically connected to the articulated robot, vision system, and CNC machine tool.

[0045] Preferably, the device further includes a reversing platform located between the articulated robot and the CNC machine tool, and a safety fence surrounding the work area; the CNC machine tool consists of at least two units arranged in a mirrored side-by-side configuration.

[0046] Preferably, the support rod is a stainless steel tube with a diameter of 30 mm and a length of not less than 1500 mm, and has a through air passage inside. The air passage is connected to an external air source through an air pipe connector located at the front end of the support rod, and leads to a fan-shaped nozzle at the end of the support rod.

[0047] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0048] This invention uses a vision system to identify the spatial pose and specifications of copper tubes. The control system adaptively plans the insertion depth and gripping pose based on the identification results, adapting to the processing needs of copper tubes of different specifications without manual intervention, significantly improving production flexibility. The vision system, composed of a depth camera and a monitoring camera, combined with visual auxiliary markers and a spindle target, can monitor the relative pose of the end effector and the spindle hole in real time, dynamically adjusting the robot's trajectory to ensure precise alignment of the copper tube and the spindle hole, improving processing accuracy. Through the structural design of the support rod and retaining ring, when the articulated robot performs tilting movements, the copper tube slides along the support rod under gravity and is axially limited and centered by the retaining ring, ensuring the stability and alignment of the copper tube during processing. The air-blowing chip removal component connects to a fan-shaped nozzle through the internal air passage of the support rod, combining high-pressure pulsed airflow and low-pressure fan-shaped airflow to effectively loosen and remove copper chips from the inner wall of the copper tube and the spindle hole. Combined with acoustic emission sensors to monitor the chip removal effect, it ensures thorough removal of copper chips, improving processing quality and equipment lifespan. The entire processing is completed automatically by the control system, including copper tube identification, gripping, centering, processing, chip removal and unloading, which reduces manual intervention, improves production efficiency and automation level, and reduces labor intensity.

[0049] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention or the background art, the accompanying drawings used in the embodiments of the present invention or the background art will be described below.

[0051] Figure 1 A flowchart illustrating an adaptive processing method for multi-specification precision copper tubes provided in an embodiment of the present invention;

[0052] Figure 2 for Figure 1 A flowchart illustrating the sub-steps of step S1;

[0053] Figure 3 for Figure 2 A flowchart illustrating the sub-steps of step S11.

[0054] Figure 4 for Figure 1 A flowchart illustrating the sub-steps of step S1;

[0055] Figure 5 for Figure 1 A flowchart illustrating the sub-steps of step S3 in the middle section;

[0056] Figure 6This is a schematic diagram of an adaptive processing device for multi-specification precision copper tubes provided in an embodiment of the present invention. Detailed Implementation

[0057] To enable those skilled in the art to better understand the present invention, 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.

[0058] Please see Figure 1 , Figure 1 This is a flowchart illustrating an adaptive processing method for multi-specification precision copper tubes, provided as an embodiment of the present invention.

[0059] First, this method is applied to an adaptive processing device for multi-specification precision copper tubes. The processing device includes an articulated robot, an end effector, a vision system, and a control system. The end effector includes loading / unloading stations, a support rod, a retaining ring, and an air-blowing chip removal assembly. The method includes:

[0060] S1. The spatial pose of the copper tube located at the loading station is identified through the vision system. The outer diameter and effective gripping length of the copper tube are identified based on the spatial pose, and the insertion depth and gripping pose of the copper tube are adaptively planned.

[0061] S2. Control the articulated robot to drive the end effector to insert the copper tube into the insertion depth according to the grasping posture, control the articulated robot to perform the preset tilting action so that the copper tube slides along the support rod under the action of gravity, and the retaining ring realizes axial limitation and centering;

[0062] S3. During the process of the articulated robot carrying the aligned copper tube moving towards the spindle hole of the machine tool, the relative pose of the end effector and the spindle hole is monitored in real time through the vision system, and the movement trajectory of the articulated robot is dynamically adjusted to control the alignment of the end effector and the spindle hole.

[0063] S4. With the alignment in place, the copper tube is fed into the spindle hole and clamped by the machine tool. After the processing is completed, the end effector is re-aligned and inserted into the spindle hole. The air blowing chip removal assembly is started to blow away the copper chips, and then the copper tube is transferred to the unloading station.

[0064] In this embodiment, the processing device mainly includes an articulated robot, an end effector, a vision system, and a control system. The end effector is usually fixed to the end of the articulated robot and has loading and unloading stations for controlling the loading and unloading of copper tubes. The end effector also includes a support rod, a retaining ring at the end of the support rod, and an air-blowing chip removal assembly. The retaining ring is used to axially limit the sliding copper tube. The air-blowing chip removal assembly includes a fan-shaped nozzle connected to the internal air passage of the support rod and a visual auxiliary marking point at the front end of the support rod. The vision system mainly consists of a depth camera and a monitoring camera. The control system is electrically connected to the articulated robot, the vision system, and the CNC machine tool.

[0065] Based on the structure of the aforementioned processing device, the method provided in this embodiment is implemented through steps S1-S4. In the field of precision copper tube processing, traditional loading and unloading methods suffer from the following core problems: poor specification adaptability; insufficient positioning accuracy, resulting in large misalignment between the copper tube and the spindle hole; easy slippage of the copper tube leading to poor gripping stability and unsatisfactory chip removal effect. These problems result in low production efficiency, difficulty in guaranteeing processing accuracy, and frequent equipment maintenance. Therefore, this embodiment aims to construct a complete adaptive processing closed-loop system through the collaborative control of visual recognition + adaptive planning + gravity alignment + visual servo + intelligent chip removal. In step S1, the depth camera acquires point cloud and image data of the copper tube end. Through anti-reflective processing and ring fitting algorithm, the outer diameter and effective gripping length are accurately identified. The control system automatically calculates the insertion depth and gripping posture accordingly. In step S2, after the support rod is inserted into the inner hole of the copper tube, the articulated robot performs a preset tilting action, using gravity to allow the copper tube to slide naturally. The retaining ring provides reliable axial limiting, achieving the effect of "one insertion and one tilt for alignment". In step S3, the monitoring camera tracks the visual markers and the spindle target in real time, establishes a coordinate system and calculates the pose error, which is used as a feedforward compensation to correct the robot trajectory online. After processing is completed in step S4, the robot is re-aligned with the spindle hole and the chip removal effect is monitored by a combination of high-pressure pulse and low-pressure fan-shaped airflow, combined with acoustic emission sensors, to ensure that copper chips are completely removed and achieve intelligent protection of "processing is cleaning".

[0066] See Figure 2 In one embodiment, the step of identifying the spatial pose of the copper tube located at the loading station using a vision system includes:

[0067] S10. Control the vision system to scan the copper tubes at the loading and unloading stations and acquire depth point cloud data and two-dimensional image data including the end area of ​​the copper tubes.

[0068] S11. Perform anti-reflection processing on the two-dimensional image data and extract the preliminary contour of the copper tube end face;

[0069] S12. Calculate the spatial pose of the copper tube based on the preliminary contour and depth point cloud data.

[0070] A depth camera in the vision system scans copper tubes at the loading and unloading stations. The camera acquires depth point cloud data and 2D image data, including the end region of the copper tubes. The depth point cloud data provides 3D coordinate information of the copper tube surface, while the 2D image data contains rich texture and edge information. To ensure data quality, uniform lighting must be ensured to avoid strong reflections, and camera parameters must be adjusted to suit the surface characteristics of the copper tubes. Since the copper tube surface is prone to reflection, directly acquired images may contain bright areas, interfering with contour extraction. Polarization imaging technology or multi-exposure image sequences acquired under multiple orthogonal polarization states can be used. By calculating the comprehensive polarization contrast and local gradient activity, anti-reflection fusion weights are generated, and finally, an enhanced image is synthesized to suppress reflections. Edge detection, such as using the Canny operator or Hough transform, is performed on the enhanced 2D image to identify the preliminary contour of the copper tube end face. Further, a sub-pixel-level algorithm is used to accurately fit the contour edges, obtaining pixel-level precision geometry of the copper tube end face.

[0071] Further, the two-dimensional contour extracted in step S11 is registered with the depth point cloud data, and the two-dimensional contour is mapped to the three-dimensional point cloud space through coordinate transformation to form a 3D contour point cloud with semantic information. Based on the fused point cloud, the spatial plane and central axis of the copper tube end face are fitted using the least squares method or the random sampling consensus algorithm. By calculating the rotation matrix and translation vector of the copper tube axis relative to the camera coordinate system, its spatial pose, including position and orientation, is determined. To improve robustness, Kalman filtering can be combined to optimize the pose estimation results.

[0072] This embodiment effectively overcomes the limitations of a single sensor by fusing depth point clouds and two-dimensional images, greatly improving the accuracy of spatial pose measurement and providing reliable input for subsequent grasping; the anti-reflection processing technology significantly reduces the impact of reflections on the copper tube surface on image quality, ensuring the stability of contour extraction, and is suitable for complex lighting environments in factories.

[0073] See Figure 3 In one embodiment, the step of performing anti-reflective processing on the two-dimensional image data and extracting the preliminary contour of the copper tube end face includes:

[0074] S110. Control the vision system to acquire multi-exposure image sequences of the copper tube end under at least two orthogonal polarization states;

[0075] S111. Calculate the values ​​of integrated polarization contrast and local gradient activity based on multi-exposure images;

[0076] S112. Calculate the anti-reflective fusion weights based on the values ​​of comprehensive polarization contrast and local gradient activity.

[0077] S113. Based on the anti-reflection fusion weight, fuse the multi-exposure images to generate an enhanced image, and extract the preliminary contour of the copper tube end face based on the enhanced image.

[0078] Specifically, this embodiment is implemented through the following steps:

[0079] Step 1) Acquire a multi-exposure image sequence and calculate the overall polarization contrast. The formula is as follows:

[0080] ;

[0081] ;

[0082] ;

[0083] ;

[0084] , , Separate spatial positions Pixel grayscale values ​​in low-exposure, medium-exposure, and high-exposure images;

[0085] Among the three types of exposure images, the low-exposure image is to avoid highlight saturation, the medium-exposure image is within the normal dynamic range, and the high-exposure image can preserve shadow details.

[0086] , , These represent the pixel grayscale values ​​of the three exposed images under orthogonal polarization (p-polarization);

[0087] , , The polarization contrast is shown for three different exposure images.

[0088] Small constants (e.g.) ), used for numerical stability.

[0089] in, The larger the value, the stronger the specular reflection in the region (due to the large difference in p-polarization response); the smaller the value, the more diffuse reflection dominates the region (suitable for contour extraction).

[0090] Step 2) Calculate the local gradient activity The formula is as follows:

[0091] ;

[0092] , , These are gradient factors, representing the direction and rate of maximum change in image brightness locally. They are typically represented by a two-dimensional vector, as follows:

[0093] ;

[0094] ;

[0095] ;

[0096] This represents the Euclidean norm, also known as the L² norm.

[0097] Step 3) Calculate the anti-reflective fusion weights ,include:

[0098] ;

[0099] : Reflection suppression intensity parameter, recommended initial value ;

[0100] Edge enhancement coefficient, suggested initial value ;

[0101] It is a natural constant;

[0102] In the formula, This is used to suppress contributions from highly reflective areas. This term is used to enhance the weight of edge regions and suppress smooth noise areas;

[0103] Step 4) Construct the fused image The formula is as follows:

[0104] ;

[0105] ;

[0106] In the formula, This is the result of standard multi-exposure HDR fusion. This is the result of fusing three exposure images.

[0107] In the above embodiments, the response differences of specular reflection on the copper tube surface under different polarization directions are utilized to accurately identify highly reflective areas. In the multi-exposure strategy, low exposure avoids high light saturation, medium exposure preserves the normal dynamic range, and high exposure preserves dark details, ensuring complete image information can be acquired under different lighting conditions. Gradients are calculated for each of the three exposure images to comprehensively evaluate the texture activity of local areas. In the edge region of the copper tube end face, the gradient value is larger, and the edge enhancement term approaches 1, significantly increasing the fusion weight in this region. In smooth regions, the gradient value is smaller, and the edge enhancement term approaches 0, effectively suppressing noise and areas with indistinct textures, avoiding false detections. The design of the anti-reflection fusion weight achieves adaptive image fusion. Through the above anti-reflection processing, the fusion strategy is adaptively adjusted to ensure a clear copper tube end face contour is obtained under different lighting conditions. Compared to single exposure or non-polarized imaging, this scheme significantly reduces problems such as contour breakage and false detections caused by reflection, improves the stability and repeatability of contour extraction, and this method can adapt to the surface characteristics of copper tubes of the same specification without requiring individual parameter adjustments for each specification, making it highly applicable.

[0108] See Figure 4 In one embodiment, identifying the outer diameter and effective gripping length of the copper tube based on its spatial pose includes:

[0109] S13. Determine the main axis direction of the copper tube based on the depth point cloud data and spatial pose, and slice the depth point cloud data at equal intervals along the main axis direction to generate a series of cross-sectional point cloud sets.

[0110] S14. Perform a ring fitting for each cross-sectional point cloud set, record all the fitted outer circle radii, and take the most frequent value as the outer diameter of the copper tube.

[0111] S15. Based on the effective distribution length of the depth point cloud data in the main axis direction, and combined with the preset position of the retaining ring, calculate the effective gripping length of the copper tube.

[0112] In step S13, the eigenvectors of the point cloud covariance matrix are analyzed using PCA, and the direction of the vector corresponding to the largest eigenvalue is the direction of the main axis. If the copper tube is bent, the RANSAC algorithm can be used to fit a straight line or B-spline curve piecewise to approximate the actual axis. Parallel planes are created at equal intervals along the main axis, and the point cloud within the threshold on both sides of each plane is extracted to form a continuous cross-sectional sequence. The slice spacing needs to be adjusted according to the straightness of the copper tube; when the bend is large, the spacing needs to be reduced to improve accuracy. In step S14, for each slice point cloud, the RANSAC algorithm is used to iteratively fit a circular model. Three points are randomly selected to calculate the initial circle parameters, and the number of inner points is counted (distance from point to circle < threshold). After iterative optimization, the model with the most inner points is selected as the final fitting result, effectively resisting the influence of cutting burrs or missing point clouds. After calculating the outer diameter values ​​of all slices, the interval with the highest frequency is analyzed by histogram, and the median is taken as the final outer diameter. If there is a multi-peak distribution, outliers can be eliminated by combining the copper tube process tolerance range. Finally, in step S15, the point cloud is projected along the main axis, and the point density distribution curve is calculated. Exclude areas with densities below the threshold to determine the actual length range of the copper tube. Calculate the maximum safe insertion depth dynamically based on the position of the retaining ring and the length of the support rod.

[0113] In one embodiment, the adaptive planning of the copper tube insertion depth and grasping pose includes:

[0114] For length of The copper tube is used to calculate the insertion depth of the copper tube controlled by the end effector. ,satisfy:

[0115] ;

[0116] In the formula, The effective gripping length for copper tubes, This is a preset safety redundancy.

[0117] The initial gripping pose of the end effector is determined based on the spatial pose of the copper tube. The initial gripping pose is used to control the axis of the end effector to coincide with the axis of the copper tube.

[0118] Based on the difference between the outer diameter of the copper tube and the diameter of the support rod of the end effector, the radial offset compensation amount of the end effector during the gripping process is calculated, and the final gripping posture of the end effector is determined according to the initial gripping posture and the radial offset compensation amount.

[0119] When calculating the insertion depth, the system first obtains the total length and effective gripping length of the copper tube from the vision recognition module. The effective gripping length typically refers to the length of the copper tube protruding from the clamping point, allowing for safe insertion of the support rod. A preset safety redundancy is a key parameter, which needs to be set based on the structure of the support rod end (such as the thickness of the retaining ring) and process requirements, usually between a few millimeters and tens of millimeters. Then, the control system directly applies the formula... The theoretical minimum insertion depth is calculated. To ensure absolute reliability, the calculated value will be used in actual control. .

[0120] The core of this step is achieving precise axial positioning. Through quantitative calculations, it ensures that after the copper tube is inserted, there is a gap between its end and the retaining ring on the support rod. The guaranteed safety clearance prevents the copper tube from becoming loose during movement or processing due to shallow insertion, and also avoids mechanical collisions or interference that may occur due to excessive insertion, greatly improving the safety and reliability of the operation.

[0121] The vision system acquires point cloud data of the copper tube and uses an algorithm to fit the tube's central axis in space. This axis is defined by a direction vector and the coordinates of a center point. The control system then performs a coordinate transformation, converting the direction vector and center point coordinates of the copper tube axis to the robot's base coordinate system. Next, the robot's inverse kinematics solver calculates the angles that each joint of the robot needs to achieve to align the axis of the support rod with the axis of the copper tube, based on the transformed axis information. This pose serves as the initial grasping pose. This step ensures that the support rod can approach and insert into the copper tube in an optimal, centered posture, minimizing the risk of jamming, scratching, or damage to the inner wall of the copper tube due to angular deviations.

[0122] Finally, the system calculates the radial gap between the known outer diameter of the copper tube and the diameter of the support rod: Radial gap = (outer diameter of copper tube - diameter of support rod) / 2. Under ideal alignment, this gap is uniform. However, to compensate for potential minor bends in the copper tube, fixture errors, or visual measurement errors, the control system further calculates a radial offset compensation. This compensation can be adaptively adjusted based on historical data or real-time sensor feedback. Finally, the calculated radial offset compensation is superimposed on the initial grasping pose to generate a corrected, more fault-tolerant final grasping pose command, which is then sent to the articulated robot for execution. By actively compensating for potential radial deviations, the system can adapt to unavoidable geometric errors in real-world production environments, significantly improving the success rate and adaptability of grasping copper tubes in non-ideal conditions.

[0123] See Figure 5 In one embodiment, the step of monitoring the relative pose of the end effector and the spindle hole in real time through a vision system and dynamically adjusting the motion trajectory of the articulated robot includes:

[0124] S30. Real-time acquisition of images collected by the vision system, the images including visual markers fixed to the front end of the support rod and visual targets fixed to the end face of the spindle hole;

[0125] S31. Establish the principal axis coordinate system based on the visual target, and establish the end coordinate system based on the visual marker points;

[0126] S32. Calculate the real-time pose of the end coordinate system relative to the principal axis coordinate system, compare it with the pre-calibrated ideal insertion pose, and obtain the pose error of the six degrees of freedom.

[0127] S33. Use the pose error as a feedforward compensation amount to correct the motion trajectory of the articulated robot online.

[0128] In this embodiment, the vision system acquires images containing visual markers and a visual target in real time. The visual markers are fixed to the front end of the support rod and typically employ high-contrast circular or square patterns; the visual target is fixed to the end face of the main shaft hole and uses a calibration plate or feature point array with known geometric dimensions. Image processing algorithms accurately identify the pixel coordinates of both. Based on the known geometric dimensions and spatial position of the visual target, the algorithm calculates the pose relationship between the camera and the target, establishing a principal axis coordinate system (with the center of the main shaft hole as the origin and the axis direction as the Z-axis). Similarly, based on the spatial distribution of the visual markers, an end coordinate system (with the support rod axis as the Z-axis) is established. Using camera calibration parameters, the two coordinate systems are unified to the same world coordinate system. The transformation matrix of the end coordinate system relative to the principal axis coordinate system is calculated and decomposed into translation vectors and rotation vectors (around the axis). The rotation angle of the axis is compared with the pre-calibrated ideal insertion pose (usually where the end effector coordinate system and the master axis coordinate system are completely coincident), resulting in a six-degree-of-freedom pose error: three translational errors and three rotational errors. This pose error is used as a feedforward compensation amount and input to the trajectory planning module of the robot controller. Based on the current motion state (velocity, acceleration) and the magnitude of the error, the controller generates compensation motion commands in real time, adjusting the angles of each joint of the robot to make the end effector move along the corrected trajectory, gradually reducing the pose error and ultimately achieving high-precision alignment.

[0129] This embodiment achieves sub-millimeter-level position compensation and milliradian-level attitude compensation through visual feedback, ensuring precise alignment between the copper tube and the spindle hole and avoiding collisions or jamming caused by mechanical errors. It can compensate in real time for interference factors such as thermal deformation, mechanical vibration, and load changes during robot movement, improving the system's stability and reliability under complex working conditions. It reduces the requirements for absolute positioning accuracy and mechanical repeatability of the robot, achieving high-precision docking through software compensation, thus reducing equipment manufacturing costs and maintenance difficulty. It reduces downtime and scrap rates caused by alignment failures, enabling continuous and stable automated production and improving overall processing efficiency. Through visual calibration and adaptive compensation, it can adapt to different specifications of copper tubes and machine tool spindles, enhancing the system's flexibility and versatility.

[0130] In one embodiment, the activation of the air-blowing cleaning assembly to blow away copper shavings includes:

[0131] After the end effector carries the copper tube into the spindle hole and aligns it with the spindle hole, and before the machine tool clamping device releases the copper tube, the air blowing chip removal component is started.

[0132] The air-blowing chip removal assembly's spraying process includes spraying a high-pressure pulsed airflow to loosen stubborn copper chips adhering to the inner wall of the copper tube and inside the machine tool spindle bore; then switching to a low-pressure fan-shaped airflow to blow the loosened and scattered copper chips towards the opening of the copper tube; wherein,

[0133] During the chip removal process, an acoustic emission sensor installed on the end effector monitors the sound characteristics of the airflow to determine whether the copper chips have been completely removed.

[0134] This embodiment achieves intelligent chip removal through timing control, airflow mode switching, and acoustic monitoring. The specific process is as follows: After the end effector carrying the copper tube is inserted into the spindle hole and aligned, and before the machine tool clamping device releases, the control system activates the air-blowing chip removal component. First, a high-pressure pulsed airflow, such as 0.5-0.8 MPa, is sprayed for 0.1-0.3 seconds, using the impact force of the airflow to loosen the stubborn copper chips adhering to the inner wall of the copper tube and inside the spindle hole. Then, it switches to a low-pressure fan-shaped airflow, such as 0.1-0.3 MPa, continuously blowing the loosened and scattered copper chips towards the opening of the copper tube. During the chip removal process, an acoustic emission sensor installed on the end effector monitors the sound characteristics of the airflow in real time. By analyzing the sound wave spectrum and energy changes, it determines whether the copper chips have been completely removed. When a chip removal completion signal is detected, the control system stops blowing air and executes subsequent operations. High-pressure pulsed airflow effectively loosens stubborn copper chips, while low-pressure fan-shaped airflow ensures directional discharge of the chips, preventing secondary adhesion and improving chip removal efficiency. Acoustic emission sensors monitor the airflow's acoustic characteristics, enabling online detection of the chip removal status and preventing over-blowing or incomplete chip removal, thus improving processing quality stability. Chip removal is performed before the machine tool clamping device releases, preventing copper chips from entering the clamping mechanism and causing jamming or damage, extending equipment lifespan.

[0135] Please see Figure 6 In one embodiment, a schematic diagram of an adaptive processing apparatus for multi-specification precision copper tubes is provided. Figure 6 As shown, the processing device specifically includes:

[0136] Articulated robots;

[0137] An end effector, fixed to the end of an articulated robot, is equipped with loading and unloading stations for controlling the loading and unloading operations of copper tubes, including:

[0138] Support rod, used to insert into the inner hole of the copper tube;

[0139] A retaining ring and an air-blowing chip removal assembly are provided at the end of the support rod. The retaining ring is used to axially limit the sliding copper tube. The air-blowing chip removal assembly includes a fan-shaped nozzle that communicates with the internal air passage of the support rod.

[0140] Visual aid markers are placed at the front end of the support rod;

[0141] The vision system includes a depth camera mounted on the base of the articulated robot for scanning copper pipes at the loading and unloading station; and a monitoring camera mounted on the CNC machine tool for tracking visual aid markers and machine tool spindle targets.

[0142] The control system is electrically connected to the articulated robot, vision system, and CNC machine tool.

[0143] Preferably, the device further includes a reversing platform located between the articulated robot and the CNC machine tool, and a safety fence surrounding the work area; the CNC machine tool consists of at least two units arranged in a mirrored side-by-side configuration.

[0144] Preferably, the support rod is a stainless steel tube with a diameter of 30 mm and a length of not less than 1500 mm, and has a through air passage inside. The air passage is connected to an external air source through an air pipe connector located at the front end of the support rod, and leads to a fan-shaped nozzle at the end of the support rod.

[0145] In this embodiment, the CNC machine tool includes at least a first CNC machine tool and a second CNC machine tool. The articulated robot is typically mounted on the front side of the first and second CNC machine tools, which are arranged in a mirror-image side-by-side configuration. The end effector is equipped with loading / unloading stations (carriers) and loading / unloading devices, which are mounted on the end flange of the articulated robot. Depth cameras are placed on the left and right sides of the articulated robot's base, and monitoring cameras are located near the CNC machine tools. The control system is electrically connected to the articulated robot, the vision system, and the CNC machine tools. A reversing platform is placed between the articulated robot and the CNC machine tools, and a safety fence surrounds the work area of ​​the articulated robot.

[0146] In one embodiment, the first and second CNC machine tools are preferably small-stroke machine tools. The loading / unloading device at the end of the robotic arm enters and exits through the spindle hole on the side of the CNC machine tool. The support rod is a stainless steel tube with a diameter of 30mm and a length of 1500mm. A buffer pad and an adapter are installed at the end of the support rod. The front end face of the support rod is provided with a flange connection surface for the articulated robot end. A cylindrical pin serves as an air pipe insertion port at the front end face of the support rod. An air pipe connector is installed inside the adapter. A fan-shaped nozzle is connected to the air pipe interface. A countersunk hexagonal screw is located at the end face of the adapter for fixing to the buffer pad and support rod. A countersunk hexagonal screw is located at the front end face of the support rod for connecting to the flange of the articulated robot end.

[0147] In one embodiment, based on the aforementioned processing apparatus, an automated loading and unloading process for precision copper tubes can be implemented. Taking a copper tube with an outer diameter of 80mm, a wall thickness of 10mm, and a length of 870mm as an example, the process can be similarly calculated for other sizes. Specifically, the process includes the following steps:

[0148] Scenario 1: Single-sided processing of copper tubing:

[0149] 1) Loading: Using a depth camera for visual measurement, recognition, and guided positioning, the articulated robot horizontally inserts the loading / unloading device completely into the copper tube, exceeding its length by 50mm to avoid errors during manual loading. The copper tube is then horizontally lifted off the support, and the first axis of the articulated robot rotates to a certain angle near the CNC machine tool's spindle hole. Simultaneously, the robotic arm slowly tilts downwards by 10 degrees, allowing the copper tube to slide to the support rod retaining ring. It is then slowly lifted back to a horizontal position, and its posture is adjusted to ensure the central axis of the loading / unloading device is parallel to the axis inside the spindle hole. The copper tube is then horizontally inserted into the spindle hole, with its central axis 1-2mm lower than the CNC machine tool's spindle axis. When the CNC machine tool's power chuck clamps the copper tube with its three jaws, the copper tube is centered upwards, and the loading / unloading device is completely separated from the copper tube. The articulated robot then moves vertically downwards by 15mm to ensure no interference between the outer edge of the retaining ring and the copper tube, before leaving the copper tube.

[0150] 2) CNC machine tools perform automated processes for face turning and thread turning;

[0151] 3) Unloading: Adjust the position to ensure the central axis of the loading / unloading device is parallel to the inner axis of the spindle hole. Insert the loading / unloading device horizontally into the spindle hole, ensuring the central axis of the device coincides with the machine tool spindle axis and the insertion depth exceeds the copper tube by 50mm. The cylinder in the articulated robot operates, and the gas passes through the cylindrical pin, the air pipe connector, and finally the fan-shaped nozzle to blow away excess copper chips generated during processing. The articulated robot then moves vertically upward by 14mm (maintaining a 1mm gap with the copper tube). After the three jaws of the CNC machine tool's power chuck release, the copper tube falls onto the loading / unloading device, and the articulated robot moves horizontally out of the copper tube. The first axis of the articulated robot rotates at a certain angle to the vicinity of the unloading bracket, while the robotic arm slowly tilts downward by 10 degrees, allowing the copper tube to slide to the retaining ring of the loading / unloading device. Then, it is slowly lifted until the loading / unloading device is in a horizontal position. Guided by a depth camera, the position is adjusted and the tube is neatly placed on the bracket. The articulated robot then moves vertically downward by 15mm to ensure that the outer edge of the retaining ring does not interfere with the copper tube, and then leaves the copper tube.

[0152] Scenario 2: Double-sided processing of copper tubing:

[0153] 1)-3): Refer to the single-sided processing flow of copper tubes;

[0154] 4) Reversal: Adjust the position so that the copper tube is placed horizontally on the reversal platform. Then, the articulated robot moves vertically downward by 15mm to ensure that the outer edge of the retaining ring does not interfere with the copper tube, and moves away from the copper tube. Adjust the position to the other end face of the copper tube, and insert the loading and unloading device horizontally into the copper tube. When inserting, the central axis of the loading and unloading device should coincide with the central axis of the copper tube and the insertion depth should exceed the copper tube by 50mm.

[0155] 5) Second loading: The copper tube is lifted horizontally away from the bracket. The first axis of the articulated robot rotates at a certain angle to the vicinity of the spindle hole of the CNC machine tool. At the same time, the robotic arm slowly tilts downward by 10 degrees, allowing the copper tube to slide to the retaining ring of the loading and unloading device. Then, it is slowly lifted to a horizontal position. The position is adjusted to ensure that the central axis of the loading and unloading device is parallel to the inner axis of the spindle hole. The copper tube is then horizontally inserted into the spindle hole. When inserted, the central axis of the copper tube is 1-2mm lower than the spindle axis of the CNC machine tool. When the three jaws of the CNC machine tool's power chuck clamp, the copper tube is centered upward. The loading and unloading device is completely separated from the copper tube. The articulated robot then moves vertically downward by 15mm to ensure that the outer edge of the retaining ring does not interfere with the copper tube, and then leaves the copper tube.

[0156] 6) CNC machine tools perform automated chamfering processes;

[0157] 7) Second loading / unloading: Adjust the position to ensure the central axis of the loading / unloading device is parallel to the axis inside the spindle hole. Insert the loading / unloading device horizontally into the spindle hole, ensuring the central axis of the device coincides with the machine tool spindle axis and the insertion depth exceeds the copper tube by 50mm. The cylinder in the articulated robot operates, and the gas passes through the cylindrical pin, the air pipe connector, and finally the fan-shaped nozzle to blow away excess copper shavings generated during processing. The articulated robot then moves vertically upward by 14mm (maintaining a 1mm gap with the copper tube). After the three jaws of the CNC machine tool's power chuck release, the copper tube falls onto the loading / unloading device, and the articulated robot moves horizontally out of the copper tube. The first axis of the articulated robot rotates at a certain angle to the vicinity of the unloading bracket, while the robotic arm slowly tilts downward by 10 degrees, allowing the copper tube to slide to the retaining ring of the loading / unloading device. Then, it is slowly lifted until the loading / unloading device is in a horizontal position. Guided by a depth camera, the position is adjusted and the tube is neatly placed on the bracket. The articulated robot then moves vertically downward by 15mm to ensure that the outer edge of the retaining ring does not interfere with the copper tube, and then leaves the copper tube.

[0158] In summary, the processing apparatus and method provided in this invention, through a vision system to identify the spatial pose and specifications of the copper tube, and the control system adaptively plans the insertion depth and gripping pose based on the identification results, can adapt to the processing needs of copper tubes of different specifications without manual intervention, significantly improving production flexibility. The vision system, composed of a depth camera and a monitoring camera, combined with visual auxiliary markers and a spindle target, can monitor the relative pose of the end effector and the spindle hole in real time, dynamically adjusting the robot's motion trajectory to ensure precise alignment of the copper tube and the spindle hole, improving processing accuracy. Through the structural design of the support rod and retaining ring, when the articulated robot performs tilting movements, the copper tube slides along the support rod under gravity and is axially limited and centered by the retaining ring, ensuring the stability and centering of the copper tube during processing. The air-blowing chip removal component connects to a fan-shaped nozzle through the internal air passage of the support rod, enabling the combined spraying of high-pressure pulsed airflow and low-pressure fan-shaped airflow to effectively loosen and remove copper chips from the inner wall of the copper tube and the spindle hole. Combined with acoustic emission sensors to monitor the chip removal effect, it ensures thorough removal of copper chips, improving processing quality and equipment lifespan. The entire processing is completed automatically by the control system, including copper tube identification, gripping, centering, processing, chip removal and unloading, which reduces manual intervention, improves production efficiency and automation level, and reduces labor intensity.

[0159] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

Claims

1. An adaptive machining method for multi-specification precision copper tubes, applied to an adaptive machining device for multi-specification precision copper tubes, the machining device comprising an articulated robot, an end effector, a vision system, and a control system, the end effector comprising loading and unloading stations, a support rod, a retaining ring, and an air-blowing chip removal assembly; characterized in that, The method includes: The spatial pose of the copper tube located at the loading station is identified by the vision system. Based on the spatial pose, the outer diameter and effective gripping length of the copper tube are identified, and the insertion depth and gripping pose of the copper tube are adaptively planned. The articulated robot is controlled to drive the end effector to insert the copper tube into the insertion depth according to the grasping posture. The articulated robot is controlled to perform a preset tilting action so that the copper tube slides along the support rod under the action of gravity, and the retaining ring realizes axial limitation and centering. During the process of the articulated robot carrying the aligned copper tube moving towards the spindle hole of the machine tool, the relative pose of the end effector and the spindle hole is monitored in real time by the vision system, and the movement trajectory of the articulated robot is dynamically adjusted to control the alignment of the end effector and the spindle hole. In the aligned state, the copper tube is fed into the spindle hole and clamped by the machine tool. After the processing is completed, the end effector is re-aligned and inserted into the spindle hole. The air blowing chip removal assembly is started to blow away the copper chips, and then the copper tube is transferred to the unloading station.

2. The adaptive processing method for multi-specification precision copper tubes according to claim 1, characterized in that, The process of identifying the spatial pose of copper tubes located at the loading station using a vision system includes: The vision control system scans the copper tubes at the loading and unloading stations to acquire depth point cloud data and two-dimensional image data, including the end areas of the copper tubes. Anti-reflection processing is performed on the two-dimensional image data to extract the preliminary contour of the copper tube end face; The spatial pose of the copper tube is calculated based on the preliminary contour and depth point cloud data.

3. The adaptive processing method for multi-specification precision copper tubes according to claim 2, characterized in that, The process of performing anti-reflective processing on the two-dimensional image data and extracting the preliminary contour of the copper tube end face includes: The vision system is controlled to acquire multi-exposure image sequences of the copper tube end under at least two orthogonal polarization states; Calculate the values ​​of integrated polarization contrast and local gradient activity based on multi-exposure images; The anti-reflective fusion weight is calculated based on the values ​​of comprehensive polarization contrast and local gradient activity. An enhanced image is generated by fusing multiple exposure images according to anti-reflective fusion weights, and a preliminary contour of the copper tube end face is extracted based on the enhanced image.

4. The adaptive processing method for multi-specification precision copper tubes according to claim 2, characterized in that, The process of identifying the outer diameter and effective gripping length of the copper tube based on its spatial pose includes: Based on the depth point cloud data and spatial pose, the main axis direction of the copper tube is determined, and the depth point cloud data is sliced ​​at equal intervals along the main axis direction to generate a series of cross-sectional point cloud sets. For each cross-sectional point cloud set, perform a ring fitting, record all the fitted outer circle radii, and take the most frequent value as the outer diameter of the copper tube. Based on the effective distribution length of the deep point cloud data along the main axis, and combined with the preset position of the retaining ring, the effective gripping length of the copper tube is calculated.

5. The adaptive processing method for multi-specification precision copper tubes according to claim 1, characterized in that, The adaptive planning of the copper tube insertion depth and grasping pose includes: For length of The copper tube is used to calculate the insertion depth of the copper tube controlled by the end effector. ,satisfy: ； In the formula, The effective gripping length for copper tubes, This is a preset safety redundancy. The initial gripping pose of the end effector is determined based on the spatial pose of the copper tube. The initial gripping pose is used to control the axis of the end effector to coincide with the axis of the copper tube. Based on the difference between the outer diameter of the copper tube and the diameter of the support rod of the end effector, the radial offset compensation amount of the end effector during the gripping process is calculated, and the final gripping posture of the end effector is determined according to the initial gripping posture and the radial offset compensation amount.

6. The adaptive processing method for multi-specification precision copper tubes according to claim 1, characterized in that, The method of dynamically adjusting the motion trajectory of the articulated robot by real-time monitoring of the relative pose between the end effector and the spindle hole using a vision system includes: The system acquires images in real time from the vision system. The images include visual markers fixed to the front end of the support rod and visual targets fixed to the end face of the spindle hole. Establish a principal axis coordinate system based on the visual target, and establish an end coordinate system based on the visual marker points; The real-time pose of the end-effector coordinate system relative to the principal axis coordinate system is calculated and compared with the pre-calibrated ideal insertion pose to obtain the pose error of the six degrees of freedom. The pose error is used as a feedforward compensation to correct the motion trajectory of the articulated robot online.

7. The adaptive processing method for multi-specification precision copper tubes according to claim 1, characterized in that, The activated air-blowing cleaning assembly removes copper shavings, including: After the end effector carries the copper tube into the spindle hole and aligns it with the spindle hole, and before the machine tool clamping device releases the copper tube, the air blowing chip removal component is started. The air-blowing chip removal assembly's spraying process includes spraying a high-pressure pulsed airflow to loosen stubborn copper chips adhering to the inner wall of the copper tube and inside the machine tool spindle bore; then switching to a low-pressure fan-shaped airflow to blow the loosened and scattered copper chips towards the opening of the copper tube; wherein, During the chip removal process, an acoustic emission sensor installed on the end effector monitors the sound characteristics of the airflow to determine whether the copper chips have been completely removed.

8. An adaptive machining apparatus for multi-specification precision copper tubes, used to implement the machining method as described in any one of claims 1-7, characterized in that, The processing apparatus includes: Articulated robots; An end effector, fixed to the end of an articulated robot, is equipped with loading and unloading stations for controlling the loading and unloading operations of copper tubes, including: Support rod, used to insert into the inner hole of the copper tube; A retaining ring and an air-blowing chip removal assembly are provided at the end of the support rod. The retaining ring is used to axially limit the sliding copper tube. The air-blowing chip removal assembly includes a fan-shaped nozzle that communicates with the internal air passage of the support rod. Visual aid markers are placed at the front end of the support rod; The vision system includes a depth camera mounted on the base of the articulated robot for scanning copper pipes at the loading and unloading station; and a monitoring camera mounted on the CNC machine tool for tracking visual aid markers and machine tool spindle targets. The control system is electrically connected to the articulated robot, vision system, and CNC machine tool.

9. The adaptive processing apparatus for multi-specification precision copper tubes according to claim 8, characterized in that, The device also includes a reversing platform located between the articulated robot and the CNC machine tool, and a safety fence surrounding the work area; the CNC machine tool consists of at least two units arranged in a mirrored side-by-side configuration.

10. The adaptive processing apparatus for multi-specification precision copper tubes according to claim 8, characterized in that, The support rod is a stainless steel tube with a diameter of 30mm and a length of not less than 1500mm. It has a through air passage inside, which is connected to an external air source through an air pipe connector at the front end of the support rod and leads to a fan-shaped nozzle at the end of the support rod.

Images