High-precision force-position control method with dual-stage coordination of "robot-end effector"

Abstract

The present invention belongs to, but is not limited to, the field of grinding and polishing processing technology, and particularly relates to a high-precision force-position control method, system and terminal with "robot-end effector" dual-stage coordination, including: S1, establishing a kinematic model of the robot and a position error model of the robot end; S2, identifying the geometric parameters of the robot body; S3, calculating the relative displacement error of "tool-workpiece" in the workpiece coordinate system through the actual joint information of each axis of the robot feedback during the grinding and polishing process and the identified geometric parameters of the robot body; S4, constructing a decoupling model of the tracking error from the end displacement error to the tangential and binormal directions of the grinding and polishing trajectory, and converting the workpiece error into tangential displacement error and binormal displacement error; S5, the three-degree-of-freedom end effector controls the normal contact force of grinding and polishing through the generalized predictive decoupling algorithm to compensate for the tangential displacement error and binormal displacement error.

Description

Technical Field

[0001] The present invention belongs to, but is not limited to, the field of grinding and polishing processing technology, and particularly relates to a high-precision force-position control method for dual-stage collaboration of a "robot-end effector". Background Art

[0002] Complex curved surface parts are increasingly widely used in fields such as aerospace, automotive, and shipbuilding. To ensure the profile and roughness of the part surface, complex curved surface parts generally require grinding and polishing after milling.

[0003] Currently, the grinding and polishing of complex curved surface parts mainly include manual grinding and polishing and numerical control special machine grinding and polishing. Among them, manual grinding and polishing has high labor intensity, low processing efficiency, and poor consistency, seriously restricting the surface quality of parts. Numerical control special machine grinding and polishing also has the disadvantages of high price and poor versatility. Compared with manual grinding and polishing and numerical control special machine grinding and polishing, robot grinding and polishing has the advantages of good flexibility, strong versatility, and easy expansion. In addition, since the grinding and polishing method with a force control actuator installed at the end of the robot has a relatively high contact force control accuracy, it is necessary to study the method of clamping the end effector of the robot to achieve the grinding and polishing of complex curved surfaces.

[0004] The grinding and polishing of complex curved surfaces requires very high positioning accuracy of the robot. However, due to manufacturing errors of robot components, joint clearances, equipment errors, etc., there is a deviation between the actual kinematic parameter values of the robot and the theoretical values preset by the robot control system. The resulting end positioning error can reach up to about 2 mm. Therefore, it is necessary to effectively compensate for the robot end positioning error caused by geometric errors.

[0005] Currently, the robot end positioning error compensation methods are usually divided into online compensation and offline compensation. Online error compensation uses a high-precision measuring device or sensor to obtain the real-time pose information of the robot end, and the control system corrects the pose in real time. Although the online compensation method can improve the positioning accuracy of the robot, it is highly dependent on external measuring devices and requires establishing communication between the robot and external measuring devices, which is difficult to achieve in a complex industrial site.

[0006] The offline compensation method can be divided into a kinematic model method and a non-kinematic model method. The kinematic calibration method identifies the actual kinematic parameter values through steps such as pose measurement, error modeling, and parameter identification, and corrects the robot parameters set in the robot controller to compensate for the errors caused by geometric factors of the robot. This method requires opening the underlying permissions of the controller and making corresponding modifications to the theoretical parameters of the robot model in the controller. However, the permissions of the robot controller are generally not open to users, so this method is not universal. The non-kinematic model method mainly uses machine learning to predict the robot end pose error, thereby compensating for the robot positioning error offline. However, using the non-motion model method requires collecting a large amount of sample data for model training.

[0007] In view of the above analysis, the technical problems urgently to be solved in the prior art are as follows: Position error compensation is crucial for the high-precision trajectory tracking of robots. The online error compensation method has a high compensation cost and strict requirements for measuring devices. The offline compensation method based on the kinematic model requires the robot to have an open control system. The offline compensation method based on machine learning requires a large amount of measurement data. Summary of the Invention

[0008] Aiming at the problems existing in the prior art, the present invention provides a high-precision force-position control method, system and terminal with "robot-end effector" dual-stage cooperation, which utilizes the high control precision and high responsiveness of the end effector to compensate the positioning error of the robot end in real time while controlling the normal contact force, thereby improving the grinding and polishing trajectory tracking accuracy and enabling high-quality grinding and polishing of parts.

[0009] The present invention is implemented as follows. A high-precision force-position control method with "robot-end effector" dual-stage cooperation includes:

[0010] S1, establishing a robot kinematic model and a robot end position error model;

[0011] S2, identifying the geometric parameters of the robot body;

[0012] S3, calculating the relative displacement error between the "tool-workpiece" in the workpiece coordinate system through the actual joint information of each axis of the robot fed back during the grinding and polishing process and the geometric parameters of the robot body obtained by identification;

[0013] S4, constructing a decoupling model for tracking errors of the end displacement error to the tangential and binormal directions of the grinding and polishing trajectory, and converting the workpiece error into tangential displacement error and binormal displacement error;

[0014] S5, the three-degree-of-freedom end effector controls the normal contact force of grinding and polishing through the generalized predictive decoupling algorithm to compensate the tangential displacement error and binormal displacement error.

[0015] Further, S1 specifically includes: The forward kinematic expression of the robot based on POE is:

[0016]

[0017] where g is the transformation from the base coordinate system to the tool coordinate system, is the robot joint screw, q i (i = 1, 2,..., 6) is the joint angle, is the initial transformation screw related to the selection of the base coordinate system and the tool coordinate system; the robot position error with respect to and has the following model:

[0018]

[0019] According to the error model, the following linear iterative equation is obtained:

[0020] y = Jx

[0021] where,

[0022] y = [δgg -1 ∨

[0023] J = [J 1 , J 2 , J 3 , J 4 , J 5 , J 6 , J st

[0024] x = [δξ 1 , δξ 2 , δξ 3 , δξ 4 , δξ 5 , δξ 6 , δξ st T

[0025] where, y represents the difference between the actually measured pose of the end-effector tool center point and the pose calculated using the theoretical robot geometric parameters, J is the identification Jacobian matrix, and x is the parameter to be identified.

[0026] Furthermore, S2 specifically includes: the position P of the fixed point on the end-effector in the tool coordinate system 0 and the position P in the robot base coordinate system e The relationship is

[0027]

[0028] If P 0 can be measured, then the above equation can be written as

[0029]

[0030] where, δP e = P e a - P e n , P e a and P e n ​​​They are respectively the actual position and the theoretical position of the fixed point on the end effector measured relative to the base coordinate system;

[0031] Therefore, we can obtain

[0032]

[0033] where is the skew-symmetric matrix of P e ;

[0034] A new linear iterative equation is obtained as follows

[0035] z = Kx

[0036] where z = δP e , By measuring the positions of three non-collinear center points at the end of the robot in m different poses, the following equations can be obtained:

[0037]

[0038] Therefore, the parameter x to be identified is calculated by the least squares method as

[0039]

[0040] Furthermore, S3 specifically includes: after obtaining the true geometric parameters of the robot, substituting the actual position information of each joint of the robot and the identified actual geometric parameters of the robot body during the grinding and polishing process into the forward kinematic model, the relative displacement error between the tool and the workpiece in the workpiece coordinate system can be calculated as

[0041] ΔP = g(ξ act , θ act ) - P a

[0042] where P a represents the position of the tool center point in the workpiece coordinate system calculated by the theoretical geometric parameters of the robot.

[0043] Furthermore, S4 specifically includes: by means of the transformation matrix transform the displacement error ΔP in the workpiece coordinate system into the tangential position error Δp t and the secondary normal position error Δp s ;

[0044] In order to obtain the transformation matrix Establish a tool coordinate system (X t Y t Z t ) at the center of the grinding head, and establish a workpiece local coordinate system (X p at the grinding and polishing point on the workpiece.Y p Z p ), where Z p is the normal vector of the surface point, X p , Y p are respectively the maximum and minimum principal curvature directions of the grinding and polishing point, and this is the grinding and polishing attitude for path planning at this point. A tool contact point coordinate system (X pt Y pt Z pt ) is established at the contact point between the grinding and polishing tool and the workpiece, where X pt has the same direction as X t , Y pt has the opposite direction to Z t , and Z pt has the opposite direction to Y t ;

[0045] During the grinding and polishing process, the direction of the grinding head tool axis is parallel to the minimum principal curvature direction. Therefore, Y pt , Y p have the same direction.

[0046] Therefore, the homogeneous coordinate transformation matrix between the tool contact point coordinate system (X pt Y pt Z pt ) and the workpiece local coordinate system (X p Y p Z p ) is

[0047]

[0048] where α is the contact angle, which can be obtained by offline calculation of the workpiece model and the planned grinding and polishing path points;

[0049] The homogeneous coordinate transformation matrix between the tool coordinate system (X t Y t Z t ) and the tool contact point coordinate system (X pt Y pt Z pt ) is

[0050]

[0051] where R is the radius of the grinding head, K e is the stiffness of the grinding head, and F n is the normal contact force of grinding and polishing;

[0052] Therefore, the homogeneous coordinate transformation matrix between the tool coordinate system (X t Y t Z t ) and the workpiece local coordinate system (X p Y p Z pThe homogeneous coordinate transformation matrix between

[0053]

[0054] The workpiece coordinate system (X b Y b Z b ) and the local workpiece coordinate system (X p Y p Z p ) is

[0055]

[0056] Since the end effector controller obtains the pose of the tool coordinate system (X t Y t Z t ) relative to the workpiece coordinate system (X b Y b Z b ) through real-time communication with the ABB robot controller via the EGM and ADS protocols, thus it can be calculated;

[0057] Therefore, the tangential position error Δp t and the binormal position error Δp s in the local workpiece coordinate system of the grinding and polishing contact point are

[0058]

[0059] Furthermore, S5 specifically includes: The end effector uses a Y-axis servo motor to control the normal contact force F n , the X-axis servo motor compensates for the tangential displacement error Δp t , and the Z-axis servo motor compensates for the binormal displacement error Δp s . Therefore, this system has three inputs and three outputs, and the normal force and tangential position control are coupled with each other;

[0060] The control model of the system is as follows:

[0061]

[0062] Among them, A 11 (z -1 ), A 22 (z -1 ), A 33 (z -1 ), B 11 (z -1 ), B 12 (z -1 ), B 21 (z -1 ), B22 (z -1 ) and B 33 (z -1 ) polynomial, y 1 (k), y 2 (k) and y 3 (k) system output, respectively representing the normal contact force, tangential displacement error, and binormal displacement error measured by the force sensor and the servo motor encoder. u 1 (k), u 2 (k) and u 3 (k) system input, ζ 1 (k), ζ 2 (k) and ζ 3 (k) is white noise;

[0063] Decomposing the above equation into 3 subsystems, we can obtain

[0064] A 11 (z -1 )y 1 (k) = B 11 (z -1 )u 1 (k - 1) + B 12 (z -1 )u 2 (k - 1) + ζ 1 (k) / Δ

[0065] A 22 (z -1 )y 2 (k) = B 21 (z -1 )u 1 (k - 1) + B 22 (z -1 )u 2 (k - 1) + ζ 2 (k) / Δ

[0066] A 33 (z -1 )y 3 (k) = B 33 (z -1 )u 3 (k - 1) + ζ 3 (k) / Δ

[0067] Combining the Diophantine equations, the optimal output prediction value can be obtained as

[0068] Y 1 = G 11 ΔU 1 + G 12 ΔU 2 + H11 Δu 1 (k - j)+H 12 Δu 2 (k - j)+F 1 y 1 (k)

[0069] Y 2 = G 21 ΔU 1 + G 22 ΔU 2 + H 21 Δu 1 (k - j)+H 22 Δu 2 (k - j)+F 2 y 2 (k)

[0070] Y 3 = G 3 ΔU 3 + H 3 Δu 3 (k - j)+F 3 y 3 (k)

[0071] wherein, G 11 , G 12 , H 11 , H 12 , F 1 , G 21 , G 22 , H 21 , H 22 , F 2 , G 3 , H 3 , F 3 are matrix polynomials calculated according to Diophantine equations and three subsystem equations;

[0072] For the three - input three - output model control system represented above, the following objective function is adopted

[0073]

[0074] wherein, n is the prediction length, m is the control length, and λ is the control weight, In predictive control, in order to make the output y(k + j) smoothly transition to the set value y at a certain response speed r , the reference trajectory is generated by the following formula

[0075]

[0076] Then, the performance indices of the three-input and three-output system are decomposed into those of three subsystems, and we can obtain

[0077] J = J 1 + J 2 + J 3

[0078] where

[0079]

[0080] Replacing y i (k + j) with Y i (i = 1, 2, 3), we can obtain

[0081] J 1 = (Y 1 - W 1 ) T (Y 1 - W 1 ) + λΔU 1 T ΔU 1

[0082] J 2 = (Y 2 - W 2 ) T (Y 2 - W 2 ) + λΔU 2 T ΔU 2

[0083] J 3 = (Y 3 - W 3 ) T (Y 3 - W 3 ) + λΔU 3 T ΔU 3

[0084] To minimize J i , by making we can obtain

[0085] ΔU 1 = (G 11 T G 11 + λI) -1 G 11 T (W 1 - G 12 ΔU 2 - f 1 )

[0086] ΔU 2 = (G 22 T G 22 + λI) -1 G 22 T (W 2 - G 21 ΔU 1 - f 2 )

[0087] ΔU 3 = (G 3 T G 3 + λI) -1 G 3 T (W 3 - f 3 )

[0088] In the above formula, ΔU on the right side of the equation 1 and ΔU 2 are regarded as disturbances and eliminated based on feedforward decoupling control. Since ΔU on the right side of the equation 1 and ΔU 2 are unknown at the k-th step, ΔU at the (k - 1)-th step 1 and ΔU 2 are used instead. After the above formula is solved by rolling optimization, the first element Δu of the matrix ΔU i (k) (i = 1, 2, 3) is applied to the system and repeated at each sampling period; then, the control signal is calculated as i (k)(i = 1, 2, 3) is applied to the system and repeated at each sampling period; then, the control signal is calculated as

[0089] u i (k) = u i (k - 1)+Δu i (k).

[0090] Another object of the present invention is to provide a high-precision force-position control system for realizing the two-stage coordination of the "robot-end effector", including a flange, an x-direction servo motor, an x- and y-direction ball screw platform, a y-direction spring, a y-direction servo motor, an x-direction spring, a z-direction servo motor, a z-direction ball screw platform, a three-dimensional force sensor, a grinding and polishing electric spindle, and a grinding head; the entire grinding and polishing force control device is connected to the robot end flange through the flange, the three-dimensional force sensor is installed on the z-direction servo motor, the grinding and polishing electric spindle is connected to the three-dimensional force sensor, the grinding head is installed on the grinding and polishing electric spindle, and the y-direction spring and the x-direction spring improve the compliance.

[0091] Another object of the present invention is to provide a computer device, which includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the high-precision force-position control method for "robot-end effector" two-stage collaboration.

[0092] Another object of the present invention is to provide a computer-readable storage medium storing a computer program. When the computer program is executed by the processor, the processor executes the steps of the high-precision force-position control method for "robot-end effector" two-stage collaboration.

[0093] Another object of the present invention is to provide an information data processing terminal, which includes the high-precision force-position control system for "robot-end effector" two-stage collaboration.

[0094] Combined with the above technical solutions and solved technical problems, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0095] First, the present invention designs a grinding and polishing actuator that can move in three degrees of freedom directions, and utilizes the characteristics of high control precision and fast response speed of the end effector to perform high-speed dynamic compensation for the position error of the robot end, breaking through the restriction of low trajectory accuracy in the large working space of the robot on the grinding and polishing profile accuracy.

[0096] 1. A grinding and polishing force control actuator that can move in three degrees of freedom directions is designed. The three degrees of freedom respectively control the normal contact force of grinding and polishing, compensate for the tangential position error and the secondary normal position error of the robot end.

[0097] 2. A decoupling model of the tracking error from the displacement error of the robot end to the tangential and secondary normal directions of the grinding and polishing trajectory is proposed to decouple the displacement error of the robot end, and the decoupling result is used as the reference displacement of the end effector.

[0098] 3. A generalized predictive decoupling control algorithm with three inputs and three outputs is proposed to achieve decoupling control of the normal force and the tangential position, improving the force-position control precision.

[0099] Second, as the creative auxiliary evidence of the present invention, it is also reflected in the following important aspects:

[0100] (1) Aero-engines are the pearls on the industrial crown. Complex curved surface parts such as blades and blisks are the core parts of aero-engines. These parts have the characteristics of complex surface shapes, high requirements for profile tolerance and surface roughness. The machining surface quality and geometric accuracy directly affect the working efficiency and service life of the engine. A high-precision force-position control method, system and terminal with "robot-end effector" two-stage coordination proposed by the present invention can compensate for the tangential position error and sub-normal position error of the robot grinding and polishing trajectory while controlling the normal contact force during grinding and polishing, and can be well applied to the grinding and polishing of complex curved surface parts such as blades and integral blades, and can achieve grinding and polishing with high profile accuracy.

[0101] (2) The technical solution of the present invention solves the technical problems that people have been eager to solve but have never succeeded in:

[0102] Position error compensation is crucial for the high-precision trajectory tracking of robots. Online error compensation methods have high compensation costs and strict requirements for measuring equipment. Offline compensation methods based on kinematic models require robots to have an open control system. Offline compensation methods based on machine learning require a large amount of measurement data. The present invention proposes an integrated method for grinding and polishing curved surface parts of "robot-end effector" to ensure high-precision force and path tracking. By using the high control precision and high responsiveness of the end effector, the end positioning error of the robot is compensated in real time while controlling the normal contact force, thereby improving the grinding trajectory tracking accuracy and further improving the profile accuracy of the complex curved surface parts after grinding and polishing. Therefore, the present invention can well solve the technical problem of large robot trajectory tracking error.

[0103] Third, the technical solution of the present invention solves the following key problems in the prior art in industrial applications and achieves remarkable technological progress:

[0104] 1. Insufficient force-position control accuracy: In the traditional robot grinding and polishing process, due to the limitation of a single control strategy and mechanical structure, the accuracy of force-position control is often insufficient, resulting in unstable machining quality. Especially when dealing with complex curved surfaces, it is difficult to meet the requirements of high precision and consistency. The present invention effectively improves the force-position control accuracy of the robot system through a two-stage coordination control method, can accurately adjust the contact force and displacement during grinding and polishing, and significantly improves the machining accuracy and surface quality.

[0105] 2. Weak error compensation ability: In the traditional robot system during the machining process, affected by various factors such as geometric error and force transmission error, there is a large deviation between the final machining effect and the design requirements. To solve this problem, the present invention greatly improves the error compensation ability of the system by establishing a detailed kinematic model and error model, combined with an online error identification and compensation mechanism, making the machining result closer to the design requirements and improving the consistency and qualification rate of parts.

[0106] 3. Limited ability to process complex curved surfaces: In the existing technology, when dealing with complex curved surface parts, there are often problems such as low force control accuracy and insufficient flexibility, which are difficult to meet the needs of high-end manufacturing. Through the innovative series-parallel hybrid robot structure design and generalized predictive decoupling algorithm of the present invention, precise force-position control in complex curved surface machining is achieved, significantly improving the flexibility and adaptability of the robot system, enabling it to be competent for more complex machining tasks.

[0107] In summary, the present invention not only solves multiple key problems in the existing technology, but also achieves significant technological progress in aspects such as force-position control accuracy, error compensation ability, and complex curved surface machining ability, providing strong support for the machining of high-precision curved surface parts in high-end manufacturing. Description of the Drawings

[0108] Figure 1 is a schematic diagram of a three-degree-of-freedom grinding and polishing force control device provided by an embodiment of the present invention;

[0109] Figure 2 is a high-precision force-position control flow chart of "robot-end effector" two-stage collaboration provided by an embodiment of the present invention;

[0110] Figure 3 is a schematic diagram of the conversion of the position error of the robot end provided by an embodiment of the present invention;

[0111] Figure 4 is a schematic diagram of the generalized predictive decoupling control algorithm provided by an embodiment of the present invention;

[0112] Figure 5 is a schematic diagram of the grinding and polishing normal force control provided by an embodiment of the present invention;

[0113] Figure 6 is a schematic diagram of the experimental comparison of the trajectory tracking error provided by an embodiment of the present invention;

[0114] In the figure: 1. Flange; 2. X-direction servo motor; 3. X and Y-direction ball screw platform; 4. Y-direction spring; 5. Y-direction servo motor; 6. X-direction spring; 7. Z-direction servo motor; 8. Z-direction ball screw platform; 9. Three-dimensional force sensor; 10. Grinding and polishing electric spindle; 11. Grinding head. Detailed Embodiment

[0115] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0116] In industrial applications, the high-precision force-position control method of the present invention can play a significant role in the following two fields:

[0117] 1. High-precision grinding and polishing in complex surface machining: In high-end manufacturing fields such as aerospace and automotive manufacturing, the surface accuracy requirements of parts are extremely high, especially for components with complex curved surfaces, such as aero-engine blades, automotive body molds, etc. Traditional machining methods are difficult to meet the accuracy requirements of these curved surface parts. Through dual-stage collaborative control, the present invention can achieve precise control of force and displacement during the grinding and polishing process, significantly improve the machining quality, and ensure surface finish and shape accuracy. At the same time, the intelligent error compensation mechanism of the system can effectively reduce the influence of human factors on machining accuracy, further improving production efficiency and product consistency.

[0118] 2. Precision control in optical mirror machining: In the manufacturing of precision instruments and the production of optical devices, such as large-aperture optical lenses, reflectors, etc., the surface is required to have extremely high smoothness and a flawless curved surface shape. The control method provided by the present invention can precisely control the force and displacement between the end effector and the workpiece, ensuring that every step in the grinding and polishing process meets the design requirements and avoiding surface damage or defects caused by uneven force or position error. This high-precision force-position control makes the manufacturing of complex curved surface optical devices possible, greatly improving the machining accuracy and surface quality of the products.

[0119] To achieve the above object, according to one aspect of the present invention, there is provided a high-precision force-position control method, system and terminal with "robot-end effector" dual-stage collaboration, characterized in that:

[0120] For a specific grinding and polishing scenario, a new type of three-degree-of-freedom force-position control end effector is designed.

[0121] For a specific grinding and polishing scenario, a method for identifying robot geometric errors and converting end position errors is designed.

[0122] For a specific grinding and polishing scenario, a new type of multi-input multi-output force-position decoupling control method is designed.

[0123] Preferably, a new type of three-degree-of-freedom force-position control end effector is as follows:

[0124] The described new three - degree - of - freedom force - position control end - effector consists of a flange, an X - axis servo motor, a Y - axis servo motor, X - axis and Y - axis ball - screw platforms, an X - axis spring, a Y - axis spring, a Z - axis servo motor, a Z - axis ball - screw platform, a three - dimensional force sensor, a grinding and polishing electric spindle, a grinding head, and some connecting parts. The flange is connected to the flange at the end of the robot; the X - axis spring and Y - axis spring improve the compliance of the end - effector in the X - direction and Y - direction; the force sensor is connected to the grinding and polishing electric spindle and the Z - axis ball - screw platform to detect the contact force during the grinding and polishing process; the grinding head is connected to the grinding and polishing electric spindle.

[0125] The three - degree - of - freedom force - position control end - effector can achieve motion in three translational degrees of freedom in space. The servo motor is equipped with an encoder, and it can obtain the X - axis motor displacement Xx, Y - axis motor displacement Xy, and Z - axis motor displacement Xz in real - time.

[0126] The three - degree - of - freedom force - position control end - effector can compensate for the positioning error of the robot end in real - time while controlling the normal contact force, thereby improving the tracking accuracy of the grinding and polishing trajectory.

[0127] Preferably, a method for identifying the geometric error of a robot and converting the end - position error is as follows: First, use the kinematic model method to identify the actual geometric parameters of the robot body offline; then, substitute the actual position information of each joint of the robot during the grinding and polishing process and the identified actual geometric parameters of the robot body into the forward kinematic model to predict the relative displacement error between the tool and the workpiece in the workpiece coordinate system; finally, through the transformation matrix between the workpiece coordinate system and the local coordinate system of the grinding and polishing contact point, convert the workpiece displacement error into the tangential displacement error and the binormal displacement error as the reference displacement of the three - degree - of - freedom force - position control end - effector, thereby realizing the online compensation of the robot end - position error.

[0128] The present invention proposes a high - precision force - position control method, system, and terminal with "robot - end - effector" two - stage collaboration. First, a brief description of the end - grinding and polishing force - control system is given, as Figure 1 shown, including a flange 1, an x - direction servo motor 2, an x - and y - direction ball - screw platform 3, a y - direction spring 4, a y - direction servo motor 5, an x - direction spring 6, a z - direction servo motor 7, a z - direction ball - screw platform 8, a three - dimensional force sensor 9, a grinding and polishing electric spindle 10, and a grinding head 11. The entire grinding and polishing force - control device is connected to the flange at the end of the robot through the flange 1. The three - dimensional force sensor 9 is installed on the z - direction servo motor 7. The grinding and polishing electric spindle 10 is connected to the three - dimensional force sensor 9, and the grinding head 11 is installed on the grinding and polishing electric spindle 10. The y - direction spring 4 and the x - direction spring 6 improve the compliance.

[0129] The working principle of the present invention is first based on the mechanical structure of the end - grinding and polishing force - control system, as Figure 1As shown in the figure. The entire system connects the force control device to the end flange of the robot through the flange 1, which can ensure the integrated operation of the force control system and the robot. The core component of the force control system is the three-dimensional force sensor 9, which is installed on the z-direction servo motor 7 and is used to detect the force acting on the grinding head during the grinding and polishing process in real time, ensuring precise control of the force during the machining process.

[0130] Secondly, multiple servo motors are set in the system to precisely control the displacement of the grinding and polishing head in the x, y, and z directions. Among them, the x-direction servo motor 2 drives the grinding head to move in the x direction through the ball screw platform 3, and increases the compliance through the x-direction spring 6, enabling the system to adapt to the complex shape of the curved surface parts. The y-direction servo motor 5 also controls the movement of the grinding head in the y direction through the ball screw platform 3, and at the same time, the y-direction spring 4 also improves the compliance in this direction to reduce the sudden change of force that may occur during the machining process.

[0131] The control in the z direction is completed by the z-direction servo motor 7, which drives the z-direction ball screw platform 8 to enable the grinding head to move precisely up and down in the z direction. In addition, the grinding and polishing electric spindle 10 is directly connected to the three-dimensional force sensor 9 to drive the grinding head 11 to perform the grinding and polishing operation. The three-dimensional force sensor real-time feedbacks the force change during the grinding and polishing process. Combining with the precise control of the servo motor, the system can adjust the contact force in real time during the grinding and polishing process, thus ensuring high-precision force-position control.

[0132] Finally, the entire system realizes high-precision force-position control during the grinding and polishing process of complex curved surface parts by comprehensively applying multi-axis servo motors, ball screw platforms, and spring systems, combined with the real-time feedback of the three-dimensional force sensor. Through the compliance design in the x, y, and z directions and real-time force feedback, the system can flexibly respond to the processing requirements of different curved surfaces, ensure the stability and precision during the grinding and polishing process, and greatly improve the quality and efficiency of curved surface machining.

[0133] Then, a high-precision force-position control method, system, and terminal for "robot-end effector" dual-stage coordination are described, and the process is as Figure 2 shown. The steps are as follows:

[0134] S1, establish a robot kinematic model and a robot end position error model.

[0135] The forward kinematic expression of the robot based on POE is:

[0136]

[0137] where g is the transformation from the base coordinate system to the tool coordinate system, is the robot joint screw, and q i (i = 1, 2,..., 6) is the joint angle, is the initial transformation twist related to the selection of the base coordinate system and the tool coordinate system. The robot position error with respect to and has the following model:

[0138]

[0139] According to the error model, the following linear iterative equation is obtained:

[0140] y = Jx

[0141] where

[0142] y = [δgg -1 ∨

[0143] J = [J 1 , J 2 , J 3 , J 4 , J 5 , J 6 , J st

[0144] x = [δξ 1 , δξ 2 , δξ 3 , δξ 4 , δξ 5 , δξ 6 , δξ st T

[0145] where y represents the difference between the actually measured pose of the tool center point of the end effector and the pose calculated using the theoretical robot geometric parameters, J is the identification Jacobian matrix, and x is the parameter to be identified.

[0146] S2, identify the geometric parameters of the robot body.

[0147] The position P of the fixed point on the end effector in the tool coordinate system 0 and the position P in the robot base coordinate system e are related as

[0148]

[0149] If P 0 can be measured, then the above equation can be written as

[0150]

[0151] where δP e = P e a - P​​​e n , P e a and P e n are respectively the actual position and the theoretical position of a fixed point on the end effector measured relative to the base coordinate system.

[0152] Therefore, it can be obtained that

[0153]

[0154] wherein, is the skew-symmetric matrix of P e .

[0155] A new linear iterative equation is obtained as follows

[0156] z = Kx

[0157] wherein, z = δP e , By measuring the positions of three non-collinear center points at the end of the robot in m different poses respectively, the following equation can be obtained:

[0158]

[0159] Therefore, the parameter x to be identified is calculated by the least squares method as

[0160]

[0161] S3. Through the actual joint information of each axis of the robot fed back during the grinding and polishing process and the geometric parameters of the robot body identified, the relative displacement error between the "tool - workpiece" in the workpiece coordinate system is calculated.

[0162] After obtaining the true geometric parameters of the robot, substituting the actual position information of each joint of the robot during the grinding and polishing process and the actual geometric parameters of the robot body identified into the forward kinematic model, the relative displacement error between the tool and the workpiece in the workpiece coordinate system can be calculated as

[0163] ΔP = g(ξ act , θ act ) - P a

[0164] wherein, P a represents the position of the tool center point in the workpiece coordinate system calculated by the theoretical geometric parameters of the robot.

[0165] S4. Construct a decoupling model of the tracking error from the end displacement error to the tangential and binormal directions of the grinding and polishing trajectory, and convert the workpiece error into the tangential displacement error and the binormal displacement error.

[0166] Through the transformation matrix The displacement error ΔP in the workpiece coordinate system is transformed into the tangential position error Δp and the secondary normal position error Δp at the grinding and polishing contact point in the local coordinate system of the workpiece. t and the secondary normal position error Δp s .

[0167] To obtain the transformation matrix A tool coordinate system (X t Y t Z t ) is established at the center of the grinding head, and a local coordinate system of the workpiece (X p Y p Z p ) is established at the grinding and polishing point on the workpiece, where Z p is the normal vector of the surface point, and X p , Y p are the maximum and minimum principal curvature directions of the grinding and polishing point respectively, which is the grinding and polishing posture for path planning at this point. A tool contact point coordinate system (X pt Y pt Z pt ) is established at the contact point between the grinding and polishing tool and the workpiece, where X pt has the same direction as X t , Y pt has the opposite direction to Z t , and Z pt has the opposite direction to Y t . The schematic diagram of the conversion of the end position error is as shown in Figure 3 .

[0168] During the grinding and polishing process, the direction of the grinding head tool axis is parallel to the minimum principal curvature direction. Therefore, the directions of Y pt , Y p are the same. Thus, the homogeneous coordinate transformation matrix between the tool contact point coordinate system (X

[0169] Y pt Y pt Z pt ) and the local coordinate system of the workpiece (X p Y p Z p ) is

[0170]

[0171] where α is the contact angle, which can be obtained by offline calculation of the workpiece model and the planned grinding and polishing path points.

[0172] The tool coordinate system (X t Y t Z t ) and the tool contact point coordinate system (X pt Y pt Z ptThe homogeneous coordinate transformation matrix between

[0173]

[0174] where R is the radius of the grinding head, and K e is the stiffness of the grinding head, and F n is the normal contact force of grinding and polishing.

[0175] Therefore, the homogeneous coordinate transformation matrix between the tool coordinate system (X t Y t Z t ) and the workpiece local coordinate system (X p Y p Z p ) is

[0176]

[0177] The transformation matrix between the workpiece coordinate system (X b Y b Z b ) and the workpiece local coordinate system (X p Y p Z p ) is

[0178]

[0179] Since the end effector controller obtains the pose of the tool coordinate system (X t Y t Z t ) relative to the workpiece coordinate system (X b Y b Z b ) through real-time communication with the ABB robot controller via the EGM and ADS protocols, thus it can be calculated.

[0180] Therefore, the tangential position error Δp t and the binormal position error Δp s in the workpiece local coordinate system are

[0181]

[0182] In S5, the three-degree-of-freedom end effector controls the normal contact force of grinding and polishing through the generalized predictive decoupling algorithm to compensate for the tangential displacement error and the binormal displacement error.

[0183] The end effector uses the Y-axis servo motor to control the normal contact force F n , the X-axis servo motor compensates for the tangential displacement error Δp t , and the Z-axis servo motor compensates for the binormal displacement error Δp s, so this system has three inputs and three outputs, and the normal force and tangential position control are coupled with each other. The system control block diagram is as shown in Figure 4 shown.

[0184] The control model of the system is as follows:

[0185]

[0186] where, A 11 (z -1 ), A 22 (z -1 ), A 33 (z -1 ), B 11 (z -1 ), B 12 (z -1 ), B 21 (z -1 ), B 22 (z -1 ) and B 33 (z -1 ) are polynomial expressions, y 1 (k), y 2 (k) and y 3 (k) are the system outputs, representing the normal contact force, tangential displacement error, and secondary normal displacement error measured by the force sensor and the servo motor encoder respectively. u 1 (k), u 2 (k) and u 3 (k) are the system inputs, and ζ 1 (k), ζ 2 (k) and ζ 3 (k) are white noises.

[0187] Decomposing the above equation into 3 subsystems, we can get

[0188] A 11 (z -1 )y 1 (k) = B 11 (z -1 )u 1 (k - 1) + B 12 (z -1 )u 2 (k - 1) + ζ 1 (k) / Δ

[0189] A 22 (z -1 )y 2 (k) = B 21 (z -1 )u 1 (k - 1) + B 22 (z-1 )u 2 (k - 1)+ζ 2 (k) / Δ

[0190] A 33 (z -1 )y 3 (k)=B 33 (z -1 )u 3 (k - 1)+ζ 3 (k) / Δ

[0191] Combining the Diophantine equations, the optimal output prediction value can be obtained as

[0192] Y 1 =G 11 ΔU 1 +G 12 ΔU 2 +H 11 Δu 1 (k - j)+H 12 Δu 2 (k - j)+F 1 y 1 (k)

[0193] Y 2 =G 21 ΔU 1 +G 22 ΔU 2 +H 21 Δu 1 (k - j)+H 22 Δu 2 (k - j)+F 2 y 2 (k)

[0194] Y 3 =G 3 ΔU 3 +H 3 Δu 3 (k - j)+F 3 y 3 (k)

[0195] where G 11 ,G 12 ,H 11 ,H 12 ,F 1 ,G 21 ,G 22 ,H 21 ,H 22 ,F 2 ,G 3 ,H 3 ,F3 It is a matrix polynomial calculated according to Diophantine equations and three subsystem equations.

[0196] For the three-input three-output model control system represented above, the following objective function is adopted

[0197]

[0198] where n is the prediction length, m is the control length, and λ is the control weight, In predictive control, in order to make the output y(k + j) smoothly transition to the set value y at a certain response speed r , the reference trajectory is generated by the following formula

[0199]

[0200] Then, the performance index of the three-input three-output system is decomposed into the performance indices of three subsystems, and we can obtain

[0201] J = J 1 + J 2 + J 3

[0202] where,

[0203]

[0204] Replacing y i (k + j) with Y i (i = 1, 2, 3), we can obtain

[0205] J 1 = (Y 1 - W 1 ) T (Y 1 - W 1 ) + λΔU 1 T ΔU 1

[0206] J 2 = (Y 2 - W 2 ) T (Y 2 - W 2 ) + λΔU 2 T ΔU 2

[0207] J 3 = (Y 3 - W 3 ) T (Y3 -W 3 ) + λΔU 3 T ΔU 3

[0208] To minimize J i , make It can be obtained that

[0209] ΔU 1 =(G 11 T G 11 + λI) -1 G 11 T (W 1 - G 12 ΔU 2 - f 1 )

[0210] ΔU 2 =(G 22 T G 22 + λI) -1 G 22 T (W 2 - G 21 ΔU 1 - f 2 )

[0211] ΔU 3 =(G 3 T G 3 + λI) -1 G 3 T (W 3 - f 3 )

[0212] In the above equation, ΔU 1 and ΔU 2 on the right - hand side of the equation are regarded as perturbations and eliminated based on feed - forward decoupling control. Since ΔU 1 and ΔU 2 are unknown at the k - th step, ΔU 1 and ΔU 2 at the (k - 1) - th step are used instead. After the above equation is solved by rolling optimization, the first element Δu i of the matrix ΔU i (k)(i = 1, 2, 3) is applied to the system and repeated at each sampling period. Then, the control signal is calculated as

[0213] u i (k)=u i(k - 1)+Δu i (k)

[0214] The present invention designs a grinding and polishing actuator that can move in three degrees of freedom, and utilizes the characteristics of high control precision and response speed of the end effector to perform high-speed dynamic compensation for the position error at the end of the robot, breaking through the restriction of low trajectory precision in the large working space of the robot on the grinding and polishing contour precision.

[0215] Compared with the current robot offline position compensation method, using the high-precision force-position control method of "robot-end effector" two-stage collaboration proposed by the present invention, while controlling the normal contact force accuracy of grinding and polishing to be ±0.4 N, the average tracking error of the robot trajectory is 0.15 mm. Compared with the robot offline position compensation method, the trajectory tracking accuracy of the method of the present invention is improved by 72%. The experimental results are as Figure 5 、 6 shown. Therefore, the method proposed by the present invention can effectively improve the grinding and polishing processing accuracy.

[0216] An application embodiment of the present invention provides a computer device, which includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the high-precision force-position control method of "robot-end effector" two-stage collaboration.

[0217] An application embodiment of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the high-precision force-position control method of "robot-end effector" two-stage collaboration.

[0218] An application embodiment of the present invention provides an information data processing terminal, which includes a high-precision force-position control system of "robot-end effector" two-stage collaboration.

[0219] I. The specific application fields or related products of the present invention.

[0220] A high-precision force-position control method, system and terminal of "robot-end effector" two-stage collaboration proposed by the present invention. The end effector can move in three degrees of freedom. Through the force-position decoupling control algorithm, it respectively controls the normal contact force during the grinding and polishing process and compensates the tangential and binormal displacements in the local coordinate system of the workpiece. While ensuring the force control accuracy of grinding and polishing, it also improves the robot trajectory tracking accuracy. The present invention can be well applied to the grinding and polishing processing of complex curved surface parts such as blades and integral blisks, ensuring the force-position control accuracy during the grinding and polishing process, solving problems such as poor grinding and polishing quality of complex curved surface parts such as blades and integral blisks, and realizing high-precision and high-quality grinding and polishing processing.

[0221] II. Evidence related to the technical effects obtained in the embodiments of the present invention.

[0222] Compared with the current robot offline position compensation method, by using the high-precision force-position control method of "robot-end effector" dual-stage coordination proposed in the present invention, while the accuracy of the controlled grinding and polishing normal contact force is ±0.4 N, the average tracking error of the robot trajectory is 0.15 mm. Compared with the robot offline position compensation method, the trajectory tracking accuracy of the method of the present invention is improved by 72%. The experimental results are as Figure 5 、 6 shown. Therefore, the method proposed in the present invention can effectively improve the grinding and polishing processing accuracy.

[0223] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated designed hardware. Those of ordinary skill in the art can understand that the above devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code is provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits of programmable hardware devices such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable logic devices such as field programmable gate arrays, or can be implemented by software executed by various types of processors, or can be implemented by a combination of the above hardware circuits and software, such as firmware.

[0224] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention by those skilled in the art within the technical scope disclosed by the present invention shall be covered by the protection scope of the present invention.

Claims

1. A high-precision force-position control method for "robot-end effector" two-stage collaboration, characterized in that: include: S1, establish robot kinematics model and robot end position error model; S2, identifying the geometric parameters of the robot body; S3, calculate the relative displacement error of "tool-workpiece" in the workpiece coordinate system through the actual joint information of each axis of the robot fed back during the grinding and polishing process and the robot body geometric parameters obtained by identification; S4, constructing a tracking error decoupling model from the end displacement error to the tangential and sub-normal directions of the grinding and polishing trajectory, and converting the workpiece error into the tangential displacement error and sub-normal displacement error; S5, the three-degree-of-freedom end effector controls the grinding and polishing normal contact force through the generalized predictive decoupling algorithm to compensate for the tangential displacement error and the secondary normal displacement error; S5 specifically includes: the end effector uses a Y-axis servo motor to control the normal contact force F n , X-axis servo motor compensates for tangential displacement error Δp t , Z-axis servo motor compensation pair normal displacement error Δp s , so the system has three inputs and three outputs, and the normal force and tangential position control are coupled to each other; The control model of the system is as follows: In the formula, y1(k), y2(k) and y3(k) are the system outputs, which represent the normal contact force, tangential displacement error and secondary normal displacement error measured by the force sensor and the servo motor encoder respectively; u1(k), u2(k) and u3(k) are the system inputs, ζ1(k), ζ2(k) and ζ3(k) are white noises; Decomposing the above equation into three subsystems, we can get And 11 (from -1 )y1(k)=B 11 (from -1 )u1(k-1)+B 12 (from -1 )u2(k-1)+ζ1(k) / Δ A 22 (z -1 )y2(k)=B 21 (z -1 )u1(k-1)+B 22 (z -1 )u2(k-1)+ζ2(k) / Δ A 33 (z -1 )y3(k)=B 33 (z -1 )u3(k-1)+ζ3(k) / Δ Combined with the Diophantine equation, the optimal output prediction value can be obtained as Y1=G 11 ΔU1+G 12 ΔU2+H 11 Δu1(k-j)+H 12 Δu2(k-j)+F1y1(k) Y2=G 21 ΔU1+G 22 ΔU2+H 21 Δu1(k-j)+H 22 Δu2(k-j)+F2y2(k) Y3=G3ΔU3+H3Δu3(kj)+F3y3(k) Among them, G 11 ,G 12 ,H 11 ,H 12 ,F1,G 21 ,G 22 ,H 21 ,H 22 ,F2,G3,H3,F3 are matrix polynomials calculated based on the Diophantine equation and the three subsystem equations; For the three-input three-output model control system represented above, the following objective function is adopted Where n is the prediction length, m is the control length, and λ is the control weight. In predictive control, in order to make the output y(k+j) smoothly transition to the set value y at a certain response speed r , the reference trajectory is generated by Then the performance index of the three-input and three-output system is decomposed into the performance index of three subsystems, and we can get J=J1+J2+J3 in, y i (k+j) is replaced by Y i (i=1, 2, 3), we can get J1=(Y1-W1) T (Y1-W1)+λΔU1 T ΔU1 J2=(Y2-W2) T (Y2-W2)+λΔU2 T ΔU2 <h2 style=";text-align:left;direction:ltr">J3=(Y3-W3)<h2 style=";text-align:left;direction:ltr"> T <h2 style=";text-align:left;direction:ltr"> (Y3-W3)+λΔU3<h2 style=";text-align:left;direction:ltr"> T <h2 style=";text-align:left;direction:ltr"> ΔU3 In order to minimize J i ,make Available ΔU1=(G 11 T G 11 +λI) -1 G 11 T (W1-G 12 ΔU2-f1) ΔU2=(G 22 T G 22 +λI) -1 G 22 T (W2-G 21 ΔU1-f2) <h2 style=";text-align:left;direction:ltr">ΔU3=(G3<h2 style=";text-align:left;direction:ltr"> T <h2 style=";text-align:left;direction:ltr"> G3+λI)<h2 style=";text-align:left;direction:ltr"> -1 <h2 style=";text-align:left;direction:ltr"> G3<h2 style=";text-align:left;direction:ltr"> T <h2 style=";text-align:left;direction:ltr"> (W3-f3) In the above equation, ΔU1 and ΔU2 on the right side of the equation are regarded as disturbances and eliminated based on feedforward decoupling control; since ΔU1 and ΔU2 on the right side of the equation are unknown at the kth step, they are replaced by ΔU1 and ΔU2 at the k-1th step; after the rolling optimization solution, the matrix ΔU i The first element Δu i (k) (i=1, 2, 3) is applied to the system and repeated once in each sampling period; then, the control signal is calculated as at i (k)=u i (k-1)+Δu i (k)。 2. The high-precision force-position control method of "robot-end effector" two-stage coordination as claimed in claim 1 is characterized in that: S1 specifically includes: The forward kinematics expression of the robot based on POE is: Among them, g is the transformation from the base coordinate system to the tool coordinate system, is the robot joint rotation, q i (i=1,2,…,6) is the joint angle, is the initial transformation rotation related to the selection of the base coordinate system and the tool coordinate system; the robot position error is about and The model is as follows: According to the error model, the linear iterative equation is as follows: y=Jx in, y=[δgg -1 ] ∨ J=[J1,J2,J3,J4,J5,J6,J st ] x=[δξ1,δξ2,δξ3,δξ4,δξ5,δξ6,δξ st ] T Where y represents the difference between the actual measured pose of the end effector tool center point and the pose calculated using the theoretical robot geometric parameters, J is the identification Jacobian matrix, and x is the parameter to be identified.

3. The high-precision force-position control method of "robot-end effector" two-stage coordination as claimed in claim 1 is characterized in that: S2 specifically includes: the position P0 of the fixed point on the end effector in the tool coordinate system and the position P0 in the robot base coordinate system e The relationship is P0 can be measured, so the above formula can be written as Among them, δP e =P e a -P e n , P e a and P e n are the actual position and theoretical position of the fixed point on the end effector measured relative to the base coordinate system, respectively; Therefore, we can get in, YesP e The antisymmetric matrix of ; The new linear iterative equation is as follows z=Kx Where z = δP e , By measuring the positions of the three non-collinear center points of the robot end in m different positions, the following equation can be obtained: Therefore, the parameter to be identified x is calculated by the least squares method as 4. The high-precision force-position control method of "robot-end effector" two-stage coordination as claimed in claim 1 is characterized in that: S3 specifically includes: after obtaining the real geometric parameters of the robot, the actual position information of each joint of the robot during the grinding and polishing process and the actual geometric parameters of the robot body obtained by identification are substituted into the kinematics forward model, and the relative displacement error between the tool and the workpiece in the workpiece coordinate system can be calculated as: ΔP=g(ξ act ,i act )-P a Among them, P a Indicates the position of the tool center point in the workpiece coordinate system calculated by the robot theoretical geometric parameters.

5. The high-precision force-position control method of "robot-end effector" two-stage coordination as claimed in claim 4 is characterized in that: S4 specifically includes: The displacement error ΔP in the workpiece coordinate system is converted into the tangential position error Δp ​​of the grinding and polishing contact point in the local coordinate system of the workpiece t and the subnormal position error Δp s ; To get the transformation matrix Establish the tool coordinate system (X t Y t Z t ), establish the workpiece local coordinate system (X p Y p Z p ), where Z p is the normal vector of the surface point, X p , Y p are the maximum and minimum principal curvature directions of the grinding and polishing point, respectively. This is the grinding and polishing posture of the path planning at the surface point. The tool contact point coordinate system (X pt Y pt Z pt ), where X pt With X t Same direction, Y pt With Z t In the opposite direction, Z pt With Y t in the opposite direction; During the grinding and polishing process, the grinding head blade axis direction is parallel to the direction of the minimum principal curvature, so Y pt , Y p The direction is consistent, so the tool contact point coordinate system (X pt Y pt Z pt ) and the workpiece local coordinate system (X p Y p Z p ) is the homogeneous coordinate transformation matrix between Among them, α is the contact angle, which can be obtained by offline calculation of the workpiece model and the planned grinding and polishing path points; Tool coordinate system (X t Y t Z t ) and the tool contact point coordinate system (X pt Y pt Z pt ) is the homogeneous coordinate transformation matrix between Where R is the radius of the grinding head, K e is the grinding head stiffness, F n is the normal contact force in grinding and polishing; Therefore, the tool coordinate system (X t Y t Z t ) and the workpiece local coordinate system (X p Y p Z p ) is the homogeneous coordinate transformation matrix between Workpiece coordinate system (X b Y b Z b ) and the workpiece local coordinate system (X p Y p Z p ) is the transformation matrix between Since the end effector controller communicates with the ABB robot controller in real time through the EGM and ADS protocols to obtain the tool coordinate system (X t Y t Z t ) relative to the workpiece coordinate system (X b Y b Z b ) position, so It can be calculated; Therefore, the tangential position error Δp ​​of the grinding and polishing contact point in the local coordinate system of the workpiece is t and the subnormal position error Δp s for 6. A "robot-end effector" two-stage coordinated high-precision force-position control system that implements the "robot-end effector" two-stage coordinated high-precision force-position control method as described in any one of claims 1 to 5, comprising a flange, an x-direction servo motor, x- and y-direction ball screw platforms, a y-direction spring, a y-direction servo motor, an x-direction spring, a z-direction servo motor, a z-direction ball screw platform, a three-dimensional force sensor, a grinding and polishing electric spindle, and a grinding head; the entire grinding and polishing force control device is connected to the robot end flange through the flange, the three-dimensional force sensor is installed on the z-direction servo motor, the grinding and polishing electric spindle is connected to the three-dimensional force sensor, the grinding head is installed on the grinding and polishing electric spindle, and the y-direction spring and the x-direction spring improve flexibility.

7. A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the "robot-end effector" two-stage collaborative high-precision force position control method as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to execute the steps of the "robot-end effector" two-stage collaborative high-precision force position control method as described in any one of claims 1 to 5.

9. An information data processing terminal, comprising the "robot-end effector" two-stage coordinated high-precision force position control system as described in claim 6.

Images