Slender shaft machining detection and self-adaptive control method

Abstract

The invention relates to a slender shaft processing detection and self-adaptive control method, which is characterized by comprising the following steps of: 1, preprocessing an industrial robot (7): determining the pose of the industrial robot (7) during processing by an operator through teaching, and performing gravity compensation on a six-dimensional force sensor (11) through calculation; 2, an operator uses a binocular vision sensor to calibrate a base coordinate system of the industrial robot (7), a coordinate system of the tail end of the supporting claw and a coordinate system of the line laser scanner (5), and the pose relation of the coordinate system of the tail end (13) of the supporting claw and the coordinate system of the line laser scanner (5) relative to the base coordinate system of the industrial robot (7) is obtained through visual tracking; according to the invention, the workpiece can be adaptively adjusted in real time according to the detected numerical value, the deformation and vibration of the slender shaft in the machining process are effectively reduced, and the machining quality and the machining precision are improved.

Description

Technical Field

[0001] The invention belongs to the field of slender shaft turning, and in particular relates to a slender shaft machining detection and adaptive control method. Background Art

[0002] Shaft parts are widely used in various mechanical equipment, mainly playing the role of support and transmission. Most of them are used in high temperature and high speed environments. Therefore, the processing of such parts requires very high technical requirements to ensure that the assembled workpieces meet the requirements of rotation accuracy. Slender shaft parts have the characteristics of large aspect ratio, poor rigidity, and difficult processing. They are prone to deformation and vibration during traditional turning, which affects the final processing quality and processing accuracy, and easily makes the shaft parts scrapped. Therefore, a slender shaft processing detection and adaptive control method is needed. Industrial robots have the characteristics of high flexibility, high precision, and high efficiency. They are widely used in assembly fields such as aerospace, automobiles, and electronics, replacing many repetitive, high-intensity, and high-complexity manual tasks, reducing processing and manufacturing costs, and improving product quality. The six-dimensional force sensor is a special sensor that can measure force and torque at the same time. Applying the six-dimensional force sensor to the industrial robot can obtain the vibration information of the object; the line laser scanner can obtain the point cloud information of the object to be measured. Combining it with the industrial robot to detect and adaptively control the slender shaft in processing can effectively reduce the deformation and vibration during the processing of the slender shaft. Summary of the invention

[0003] The object of the present invention is to provide a slender shaft processing detection and adaptive control method to solve the problems mentioned in the background technology.

[0004] A slender shaft processing detection and adaptive control method, characterized in that detection and adaptive control are used, the detection and adaptive control includes a robot guide rail 8, a small industrial robot body 7, a six-dimensional force sensor connector 10, a six-dimensional force sensor 11, a support claw connector 12, a support claw, a line laser scanner connector 4, and a line laser scanner 5; the robot guide rail 8 is placed parallel to the lathe 1 and fixed on the ground, the small industrial robot body 7 is fixed on the robot guide rail 8, an industrial robot 7 flange 9 is provided at one end of the small industrial robot body 7, the six-dimensional force sensor 11 is connected to the industrial robot flange 9 through the six-dimensional force sensor connector 10, and the support claw is connected to the six-dimensional force sensor 11 through the support claw connector 12; the line laser scanner 5 is fixed to one side of the tool shaft 2 of the lathe 1 through the line laser scanner connector 4; the detection and adaptive control method includes the following steps:

[0005] Step 1: pre-processing the industrial robot 7: the operator determines the position and posture of the industrial robot 7 during processing through teaching, and performs gravity compensation on the six-dimensional force sensor 11 through calculation;

[0006] Step 2: The operator uses a binocular vision sensor to calibrate the industrial robot 7 base coordinate system, the support claw end coordinate system, and the line laser scanner 5 coordinate system, and obtains the position and posture relationship of the support claw end 13 coordinate system and the line laser scanner 5 coordinate system relative to the industrial robot 7 base coordinate system through visual tracking;

[0007] Step 3: During the lathe machining of the slender shaft, the line laser scanner 5 can detect the deformation displacement of the slender shaft 6 in real time through the emitted laser and transmit the information to the robot 7, and the robot 7 controls the movement of the end support claw to compensate and reduce the deformation;

[0008] Step 4: During lathe machining of the slender shaft, the six-dimensional force sensor 11 collects the force generated by the vibration of the workpiece 6 on the support claw during the machining process;

[0009] Step 5: The operator adjusts the processing parameters according to the vibration data measured by the force sensor 11 to reduce the vibration.

[0010] The specific content of step one is:

[0011] S11: The debugging personnel establish a gravity compensation calculation model, analyze the relationship between the six-dimensional force sensor 11 and the robot end effector 3, the six-dimensional force sensor 11 and the robot 7, and the six-dimensional force sensor 11 and the flange 9 in the model, and record the zero point values ​​of the three force components and torque components of the six-dimensional force sensor 11 as (F 0x , F 0y , F 0z )、(M 0x , M 0y , M 0z ), the coordinates of the center of mass of the robot end effector 3 in the six-dimensional force sensor 11 coordinate system are (x, y, z);

[0012] S12: Compensation calculation is performed on the six-dimensional force sensor 11 by the least square method, and the torque equation is set as:

[0013]

[0014] The force and moment components in the x, y, and z directions directly measured by the six-dimensional force sensor 11 are recorded as (F x , F y , F z )、(M x , M y , M z ), we can get:

[0015]

[0016] S13: Control the position of the robot's end support claw and obtain the force sensor 11 data under different postures.

[0017] M=F·A,A=[x,y,z,k1,k2,k3] T

[0018] Then the matrix least squares solution is A=(F T F) -1 ·F T M;

[0019] S14: The six-dimensional force sensor eliminates the end load gravity through the above calculation results, and the vibration measurement results are more accurate.

[0020] The method for detecting and adaptively controlling slender shaft machining according to claim 1 is characterized in that the specific contents of step 2 are:

[0021] S21: The operator uses a binocular vision sensor to calibrate the base coordinate system of the industrial robot 7, pastes at least 5 target points on the base of the robot 7 and the robot end effector 3 respectively, controls the industrial robot 7 to move to the initial origin position, drives the A1 axis to move, and the binocular vision sensor tracks the target point position of the robot end effector 3. Then the operator returns the robot 7 to the initial origin position again, drives the A2 axis to move, and the binocular vision sensor tracks the target point position of the robot end effector 3. According to the target point position sequence tracked by the binocular vision sensor during the two movements, fit the A1 axis and the A2 axis for a total of two circles, thereby realizing the calibration of the base coordinate system of the industrial robot 7;

[0022] S22: The operator calibrates the end support claw coordinate system using a binocular vision sensor. On the basis of step S21, the base coordinate system of the industrial robot 7 is set as the working coordinate system. A light pen is used to perform contact measurement on the robot end effector 3, and a set of cylinders and points, lines and surfaces are fitted respectively, thereby establishing a tool coordinate system.

[0023] S23: The operator establishes a visual dynamic tracking model in the software, and according to the two groups of target point groups pasted in steps S21 and S22, the target point group of the base of the industrial robot 7 is fixedly connected to the working coordinate system, and the target point group of the robot end effector 3 is fixedly connected to the tool coordinate system; when tracking the target point group, the binocular vision sensor measures the position and posture of the coordinate system to which the target point group is fixed in its own coordinate system, thereby realizing visual dynamic tracking.

[0024] The specific contents of step three are:

[0025] S31: The line laser scanner 5 is installed beside the knife shaft 2 and can move with the knife shaft 2. When starting to work, the line laser scanner 5 emits a laser to the slender shaft 6 to scan out an elliptical contour point;

[0026] S32: Fit the point cloud by the least squares method to obtain the center of the ellipse; the general equation of the ellipse is:

[0027] x 2 +axy+by 2 +cx+dy+e=0

[0028] According to the collected point cloud p i (x i ,y i ), the fitted objective function is:

[0029]

[0030] By making all partial derivatives of f equal to 0, we can obtain:

[0031]

[0032] Solving the values ​​of a, b, c, d, and e will give you the relevant parameters of the ellipse;

[0033] S33: The upper computer software calculates the ellipse center offset displacement value caused by the processing deformation in real time, and sends the value to the robot 7, and the robot 7 controls the end support claw to perform compensation movement.

[0034] The specific content of step four is: the six-dimensional force sensor 11 collects the magnitude of the force applied to the end support claw to calculate the vibration during the processing of the slender shaft 6. The operator adjusts the spindle speed, feed speed and other parameters according to the calculation results to reduce the vibration until the force is stabilized within the threshold.

[0035] The debugging personnel integrated the algorithm of the line laser scanner 5 into the upper computer of the robot 7 to realize communication with the robot 7 and real-time online detection.

[0036] The beneficial effects of the present invention are:

[0037] The present invention aims at the deformation and vibration problems generated during the processing of slender shafts, and proposes a detection and control method, that is, a line laser scanner is used to scan the deformation displacement, the offset value is sent to the robot for compensation, and a six-dimensional force sensor is installed at the robot flange to detect the vibration, so as to adjust the process parameters. The robot of the present invention can adaptively adjust the workpiece in real time through the detected values, effectively reduce the deformation and vibration of the slender shaft during the processing, and improve the processing quality and processing accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 , is a schematic diagram of the device structure of the present invention;

[0039] Figure 2 ,for Figure 1Schematic diagram of the structure at point A.

[0040] Names marked in the figure: lathe 1; tool 2; robot end effector 3; line laser scanner connector 4; line laser scanner 5; slender shaft 6; industrial robot 7; robot ground rail 8; robot flange 9; six-dimensional force sensor connector 10; six-dimensional force sensor 11. DETAILED DESCRIPTION

[0041] In order to achieve the above technical objectives, the technical solution adopted by the present invention is:

[0042] A slender shaft processing detection and adaptive control method includes a lathe for fixing and processing the slender shaft, a line laser scanner connecting piece, a line laser scanner, a small industrial robot body, a six-dimensional force sensor connecting piece, a six-dimensional force sensor, a support claw connecting piece, and a support claw; the small industrial robot body is fixed opposite to the lathe tool axis, an industrial robot flange is provided at one end of the small industrial robot body, the six-dimensional force sensor is connected to the industrial robot flange via the six-dimensional force sensor connecting piece, and the support claw is connected to the six-dimensional force sensor via the support claw connecting piece; the line laser scanner is fixed to one side of the lathe tool axis via the line laser scanner connecting piece.

[0043] Furthermore, a slender shaft processing detection and adaptive control method comprises the following steps:

[0044] S1: Preprocessing of the industrial robot: The operator determines the position and posture of the industrial robot during processing through teaching, and performs gravity compensation on the six-dimensional force sensor through calculation;

[0045] S2: The operator uses a binocular vision sensor to calibrate the industrial robot base coordinate system, the support claw end coordinate system, and the line laser scanner coordinate system, and obtains the position and posture relationship of the support claw end coordinate system and the line laser scanner coordinate system relative to the industrial robot base coordinate system through visual tracking;

[0046] S3: During the lathe processing of the slender shaft, the line laser scanner can detect the deformation displacement of the slender shaft in real time through the emitted laser and transmit the information to the robot. The robot controls the movement of the end support claw to compensate and reduce the deformation;

[0047] S4: During the lathe machining of the slender shaft, the six-dimensional force sensor collects the force generated by the workpiece vibration on the support claw during the machining process;

[0048] S5: The operator adjusts the processing parameters based on the vibration data measured by the force sensor to reduce vibration.

[0049] Furthermore, the specific content of step S1 is:

[0050] S11: The debugging personnel establish a gravity compensation calculation model, analyze the relationship between the force sensor and the robot end sensor, the force sensor and the robot, and the force sensor and the flange in the model, and record the zero point values ​​of the three force components and torque components of the six-dimensional force sensor as (F 0x , F 0y , F 0z )、(M 0x , M 0y , M 0z ), the coordinates of the end mass center in the six-dimensional force sensor coordinate system are (x, y, z),

[0051] S12: Compensation calculation of the six-dimensional force sensor is performed by the least square method, and the torque equation is set as:

[0052]

[0053] The force and moment components in the x, y, and z directions directly measured by the six-dimensional force sensor are recorded as (M x , F y , F z )、(M x , M y , M z ), we can get:

[0054]

[0055] S13: Control the position of the robot's end support claw and obtain force sensor data under different postures.

[0056] M=F·A,A=[x,y,z,k1,k2,k3] T

[0057] Then the matrix least squares solution is A=(F T F) -1 ·F T ·M.

[0058] S14: The six-dimensional force sensor eliminates the end load gravity through the above calculation results, and the vibration measurement results are more accurate.

[0059] Furthermore, the specific content of step S2 is:

[0060] S21: The operator uses a binocular vision sensor to calibrate the base coordinate system of the industrial robot, pastes at least 5 target points on the base and the end of the robot, controls the industrial robot to move to the initial origin position, drives the A1 axis to move, and the binocular vision sensor tracks the position of the end target point. Then the operator returns the robot to the initial origin position again, drives the A2 axis to move, and the binocular vision sensor tracks the position of the end target point. According to the target point position sequence tracked by the binocular vision sensor during the two movements, fit the A1 axis and the A2 axis for a total of two circles, thereby realizing the calibration of the base coordinate system of the industrial robot;

[0061] S22: The operator calibrates the coordinate system of the end support claw using a binocular vision sensor. On the basis of step S21, the base coordinate system of the industrial robot is set as the working coordinate system. A light pen is used to perform contact measurement on the end, and a set of cylinders and points, lines and surfaces are fitted respectively, thereby establishing a tool coordinate system.

[0062] S23: The operator establishes a visual dynamic tracking model in the software, and based on the two groups of target point groups pasted in steps S21 and S22, fixes the target point group of the industrial robot base with the working coordinate system, and fixes the terminal target point group with the tool coordinate system; when tracking the target point group, the binocular vision sensor measures the position and posture of the coordinate system to which the target point group is fixed in its own coordinate system, thereby realizing visual dynamic tracking.

[0063] Furthermore, the specific content of step S3 is:

[0064] S31: The line laser scanner is installed next to the tool shaft and can move with the tool shaft. When it starts working, the line laser scanner emits laser to the slender shaft and scans out elliptical contour points;

[0065] S32: Fit the point cloud by the least squares method to obtain the center of the ellipse; the general equation of the ellipse is:

[0066] x 2 +axy+by 2 +cx+dy+e=0

[0067] According to the collected point cloud p i (x i ,y i ), the fitted objective function is:

[0068]

[0069] By making all partial derivatives of f equal to 0, we can obtain:

[0070]

[0071] By solving the values ​​of a, b, c, d, and e, we can get the relevant parameters of the ellipse.

[0072] S33: The upper computer software calculates the ellipse center offset displacement value caused by machining deformation in real time and sends the value to the robot. The robot controls the end support claw to perform compensation movement.

[0073] Furthermore, the specific content of step S4 is: the six-dimensional force sensor collects the magnitude of the force applied to the end support claw to calculate the vibration during the processing of the slender shaft, and adjusts the spindle speed, feed speed and other parameters according to the calculation results to reduce the vibration until the force is stabilized within the threshold.

[0074] Furthermore, in step S33, the debugging personnel integrate the line laser scanner algorithm into the robot host computer to achieve communication with the robot and real-time online detection.

[0075] The present invention is further described in detail below with reference to the accompanying drawings.

[0076] The main technical solutions of this application are as follows:

[0077] A slender shaft processing detection and adaptive control method, comprising:

[0078] S1: Preprocessing of the industrial robot: The operator determines the position and posture of the industrial robot during processing through teaching, and performs gravity compensation on the six-dimensional force sensor through calculation;

[0079] S2: The operator uses a binocular vision sensor to calibrate the industrial robot base coordinate system, the support claw end coordinate system, and the line laser scanner coordinate system, and obtains the position and posture relationship of the support claw end coordinate system and the line laser scanner coordinate system relative to the industrial robot base coordinate system through visual tracking;

[0080] S3: During the lathe processing of the slender shaft, the line laser scanner can detect the deformation displacement of the slender shaft in real time through the emitted laser and transmit the information to the robot. The robot controls the movement of the end support claw to compensate and reduce the deformation;

[0081] S4: During the lathe machining of the slender shaft, the six-dimensional force sensor collects the force generated by the workpiece vibration on the support claw during the machining process;

[0082] S5: The operator adjusts the processing parameters based on the vibration data measured by the force sensor to reduce vibration.

[0083] Furthermore, the specific content of step S1 is:

[0084] S11: The debugging personnel establish a gravity compensation calculation model, analyze the relationship between the force sensor and the robot end sensor, the force sensor and the robot, and the force sensor and the flange in the model, and record the zero point values ​​of the three force components and torque components of the six-dimensional force sensor as (F 0x , F 0y , F 0z )、(M 0x , M 0y , M 0z ), the coordinates of the end mass center in the six-dimensional force sensor coordinate system are (x, y, z),

[0085] S12: Compensation calculation of the six-dimensional force sensor is performed by the least square method, and the torque equation is set as:

[0086]

[0087] The force and moment components in the x, y, and z directions directly measured by the six-dimensional force sensor are recorded as (F x , F y , F z )、(M x , M y , M z ), we can get:

[0088]

[0089] S13: Control the position of the robot's end support claw and obtain force sensor data under different postures.

[0090] M=F·A,A=[x,y,z,k1,k2,k3] T

[0091] Then the matrix least squares solution is A=(F T F) -1 ·F T ·M.

[0092] S14: The six-dimensional force sensor eliminates the end load gravity through the above calculation results, and the vibration measurement results are more accurate.

[0093] Furthermore, the specific content of step S2 is:

[0094] S21: The operator uses a binocular vision sensor to calibrate the base coordinate system of the industrial robot, pastes at least 5 target points on the base and the end of the robot, controls the industrial robot to move to the initial origin position, drives the A1 axis to move, and the binocular vision sensor tracks the position of the end target point. Then the operator returns the robot to the initial origin position again, drives the A2 axis to move, and the binocular vision sensor tracks the position of the end target point. According to the target point position sequence tracked by the binocular vision sensor during the two movements, fit the A1 axis and the A2 axis for a total of two circles, thereby realizing the calibration of the base coordinate system of the industrial robot;

[0095] S22: The operator calibrates the coordinate system of the end support claw using a binocular vision sensor. On the basis of step S21, the base coordinate system of the industrial robot is set as the working coordinate system. A light pen is used to perform contact measurement on the end, and a set of cylinders and points, lines and surfaces are fitted respectively, thereby establishing a tool coordinate system.

[0096] S23: The operator establishes a visual dynamic tracking model in the software, and based on the two groups of target point groups pasted in steps S21 and S22, fixes the target point group of the industrial robot base with the working coordinate system, and fixes the terminal target point group with the tool coordinate system; when tracking the target point group, the binocular vision sensor measures the position and posture of the coordinate system to which the target point group is fixed in its own coordinate system, thereby realizing visual dynamic tracking.

[0097] Furthermore, the specific content of step S3 is:

[0098] S31: The line laser scanner is installed next to the tool shaft and can move with the tool shaft. When it starts working, the line laser scanner emits laser to the slender shaft and scans out elliptical contour points;

[0099] S32: Fit the point cloud by the least square method to obtain the center of the ellipse; the general equation of the ellipse is:

[0100] x 2 +axy+by 2 +cx+dy+e=0

[0101] According to the collected point cloud p i (x i ,y i ), the fitted objective function is:

[0102]

[0103] By making all partial derivatives of f equal to 0, we can obtain:

[0104]

[0105] By solving the values ​​of a, b, c, d, and e, we can get the relevant parameters of the ellipse.

[0106] S33: The upper computer software calculates the ellipse center offset displacement value caused by machining deformation in real time and sends the value to the robot, which controls the end support claw to perform compensation movement.

[0107] Furthermore, the specific content of step S4 is: the six-dimensional force sensor collects the magnitude of the force applied to the end support claw to calculate the vibration during the processing of the slender shaft, and adjusts the spindle speed, feed speed and other parameters according to the calculation results to reduce the vibration until the force is stabilized within the threshold.

[0108] Furthermore, in step S33, the debugging personnel integrate the line laser scanner algorithm into the robot host computer to achieve communication with the robot and real-time online detection.

[0109] The present invention discloses a slender shaft processing detection and adaptive control method, and relates to the field of slender shaft turning processing. The device of the present invention comprises a robot guide rail, a small industrial robot body, a six-dimensional force sensor connector, a six-dimensional force sensor, a support claw connector, a support claw, a line laser scanner connector, and a line laser scanner. The robot guide rail is placed parallel to the lathe and fixed on the ground, the small industrial robot body is fixed on the robot guide rail, the six-dimensional force sensor connector, the six-dimensional force sensor, the support claw connector, and the support claw are sequentially installed on the robot flange, and the line laser scanner is installed next to the lathe tool shaft through the line laser scanner connector. The present invention detects the deformation displacement of the slender shaft during the processing process through the line laser scanner and transmits the information to the robot. The robot controls the movement of the end support claw through the compensation algorithm, wherein the six-dimensional force sensor collects the force generated by the workpiece vibration on the support claw during the processing process. In this process, the robot end can move with the lathe tool, and has strong adaptability, which reduces the vibration and deformation during the processing of the slender shaft, thereby improving the processing efficiency and quality of the slender shaft product.

Claims

1. A slender shaft processing detection and adaptive control method, characterized in that: A detection and adaptive control device is used to perform industrial detection and adaptive control. The detection and adaptive control device comprises a robot guide rail (8), a small industrial robot body (7), a six-dimensional force sensor connector (10), a six-dimensional force sensor (11), a support claw connector (12), a support claw, a line laser scanner connector (4), and a line laser scanner (5); the robot guide rail (8) is placed parallel to a lathe (1) and fixed on the ground; the small industrial robot body (7) is fixed on the robot guide rail (8); an industrial robot (7) flange (9) is provided at one end of the small industrial robot body (7); the six-dimensional force sensor (11) is connected to the industrial robot flange (9) via the six-dimensional force sensor connector (10); the support claw is connected to the six-dimensional force sensor (11) via the support claw connector (12); the line laser scanner (5) is fixed to one side of a tool shaft (2) of the lathe (1) via the line laser scanner connector (4); the detection and adaptive control method comprises the following steps: Step 1: pre-processing the industrial robot (7): the operator determines the position and posture of the industrial robot (7) during processing through teaching, and performs gravity compensation on the six-dimensional force sensor (11) through calculation; Step 2: The operator uses a binocular vision sensor to calibrate the base coordinate system of the industrial robot (7), the support claw end coordinate system, and the line laser scanner (5) coordinate system, and obtains the position and posture relationship between the support claw end (13) coordinate system and the line laser scanner (5) coordinate system relative to the base coordinate system of the industrial robot (7) through visual tracking; Step 3: During the lathe processing of the slender shaft, the line laser scanner (5) can detect the deformation displacement of the slender shaft (6) in real time through the emitted laser and transmit the information to the robot (7), and the robot (7) controls the movement of the end support claw to compensate and reduce the deformation; Step 4: During lathe processing of the slender shaft, the six-dimensional force sensor (11) collects the force generated by the vibration of the workpiece (6) exerted on the supporting claw during the processing; Step 5: The operator adjusts the processing parameters according to the vibration data measured by the force sensor (11) to reduce the vibration.

2. A slender shaft processing detection and adaptive control method according to claim 1, characterized in that: The specific contents of step one are: S11: The commissioning personnel establish a gravity compensation calculation model, analyze the relationship between the six-dimensional force sensor (11) and the robot end effector (3), the six-dimensional force sensor (11) and the robot (7), and the six-dimensional force sensor (11) and the flange (9) in the model, and record the zero point values ​​of the three force components and torque components of the six-dimensional force sensor (11) as (F 0x , F 0y , F 0z )、(M 0x , M 0y , M 0z ), the coordinates of the center of mass of the robot end effector (3) in the coordinate system of the six-dimensional force sensor (11) are (x, y, z); S12: Compensation calculation is performed on the six-dimensional force sensor (11) by the least square method, and the torque equation is set as: The force and moment components in the x, y, and z directions directly measured by the six-dimensional force sensor (11) are recorded as (F x , F y , F z )、(M x , M y , M z ), we can get: S13: Control the position of the robot's end support claw and obtain the force sensor (11) data under different postures. M=F·A,A=[x,y,z,k1,k2,k3] T Then the matrix least squares solution is A=(F - F) -1 ·F T M; S14: The six-dimensional force sensor (11) eliminates the end load gravity through the above calculation results, and the vibration measurement result is more accurate.

3. A slender shaft processing detection and adaptive control method according to claim 1, characterized in that: The specific contents of step 2 are: S21: The operator uses a binocular vision sensor to calibrate the base coordinate system of the industrial robot (7), affixes at least five target points on the base of the robot (7) and the robot end effector (3), controls the industrial robot (7) to move to the initial origin position, drives the A1 axis to move, and the binocular vision sensor tracks the position of the target point of the robot end effector (3). Then, the operator returns the robot (7) to the initial origin position again, drives the A2 axis to move, and the binocular vision sensor tracks the position of the target point of the robot end effector (3). According to the target point position sequence tracked by the binocular vision sensor during the two movements, two circles of the A1 axis and the A2 axis are fitted, thereby realizing the calibration of the base coordinate system of the industrial robot (7); S22: The operator calibrates the coordinate system of the end support claw using a binocular vision sensor. On the basis of step S21, the base coordinate system of the industrial robot (7) is set as the working coordinate system. A light pen is used to perform contact measurement on the robot end effector (3), and a set of cylinders and points, lines and surfaces are fitted respectively, thereby establishing a tool coordinate system. S23: The operator establishes a visual dynamic tracking model in the software, and based on the two groups of target point groups pasted in steps S21 and S22, the base target point group of the industrial robot (7) is fixedly connected to the working coordinate system, and the target point group of the robot end effector (3) is fixedly connected to the tool coordinate system; when the binocular vision sensor tracks the target point group, it measures the position of the coordinate system to which the target point group is fixed in its own coordinate system, thereby realizing visual dynamic tracking.

4. A slender shaft processing detection and adaptive control method according to claim 1, characterized in that: The specific contents of step three are: S31: The line laser scanner (5) is installed beside the knife shaft (2) and can move with the knife shaft (2). When starting to work, the line laser scanner (5) emits laser light to the slender shaft (6) to scan out elliptical contour points; S32: Fit the point cloud by the least squares method to obtain the center of the ellipse; the general equation of the ellipse is: x 2 +axy+by 2 +cx+dy+e=0 According to the collected point cloud p i (x i ,y i ), the fitted objective function is: By making all partial derivatives of f equal to 0, we can obtain: Solving the values ​​of a, b, c, d, and e will give you the relevant parameters of the ellipse; S33: The upper computer software calculates the ellipse center offset displacement value caused by the processing deformation in real time, and sends the value to the robot (7), and the robot (7) controls the end support claw to perform compensation movement.

5. A slender shaft processing detection and adaptive control method according to claim 1, characterized in that: The specific content of step four is: the six-dimensional force sensor (11) collects the magnitude of the force applied to the end support claw, and uses this to calculate the vibration of the slender shaft (6) during processing. The operator adjusts the spindle speed, feed speed and other parameters according to the calculation results to reduce the vibration until the force is stabilized within the threshold.

6. A slender shaft processing detection and adaptive control method according to claim 4, characterized in that: The debugging personnel integrate the algorithm of the line laser scanner (5) into the upper computer of the robot (7) to realize communication with the robot (7) and real-time online detection.

Images