Three-degree-of-freedom high-speed parallel robot drag teaching method, equipment and storage medium

Abstract

This invention discloses a method, device, and storage medium for teaching a three-degree-of-freedom high-speed parallel robot to move it. The method enables the servo motors driving the robot to operate in torque control mode, allowing the robot to operate in zero-force control mode. When the robot is in zero-force control mode, it is moved by dragging it. The robot's movement trajectory is sampled and recorded, along with the actual output torque and command torque of the servo motors at corresponding sampling points. A genetic algorithm and / or neural network are used to process each point to obtain a reproduced trajectory and corresponding compensation torques at each point in the reproduced trajectory. The robot is then put into reproduction mode. Based on the compensation torques at each point in the reproduced trajectory, a command torque is output, causing the robot to move along the reproduced trajectory. This invention does not require additional sensors or external force observers; it only requires the robot's own sensors and the robot's dynamic equations to achieve zero-force dragging, making it applicable to existing robots.

Description

Technical Field

[0001] This invention relates to the field of robotics, and in particular to a method, device, and storage medium for teaching a three-degree-of-freedom high-speed parallel robot by dragging. Background Technology

[0002] Currently, in recent years, parallel robots have been increasingly used in the automotive manufacturing, food, and new energy 3C industries. Traditional parallel robots mainly pursue high speed and high precision, and usually only work on pre-programmed trajectories. However, their teaching programming is relatively complex, requiring higher professional skills from workers. The programming time is long and reprogramming is inconvenient. Therefore, there is a greater demand for reducing the difficulty of teaching functions for parallel robots.

[0003] To meet this requirement, we consider introducing drag-and-drop teaching technology into parallel robots to simplify the teaching process and reduce the difficulty of teaching, while ensuring that the parallel robot can be easily dragged and that the trajectory can be reproduced accurately and completely.

[0004] Currently, industry solutions for drag-and-drop teaching mainly fall into two categories: drag-and-drop teaching methods with external sensors and drag-and-drop teaching methods without external sensors. Dragging-and-drop teaching methods with external sensors primarily rely on force sensors, allowing the robot to detect the torque applied to each joint. However, force sensors are typically expensive and require adjustments to the configuration of parallel robots. For existing parallel robots, modifying the mechanical structure to install force sensors is a complex and costly process. Compared to sensor-based drag-and-drop teaching methods, sensorless drag-and-drop teaching methods have more advantages. These methods utilize the robot's own sensors and its dynamic model to achieve drag-and-drop teaching. The most common method in the industry is drag-and-drop teaching based on external force observers and impedance control. However, this method requires high modeling accuracy and often necessitates the use of more complex external force observers, which reduces the robot's response speed during drag-and-drop teaching.

[0005] Furthermore, the design of motion patterns for trajectory reproduction is mostly based on the characteristics of serial robots, without much consideration for the mechanical characteristics of parallel robots. In trajectory planning, optimization should be carried out in combination with the characteristics of parallel robots. Summary of the Invention

[0006] This invention provides a three-degree-of-freedom high-speed parallel robot drag teaching method, device, and storage medium to solve the technical problems existing in the prior art.

[0007] The technical solution adopted by this invention to solve the technical problems existing in the prior art is as follows:

[0008] A three-degree-of-freedom high-speed parallel robot drag teaching method is proposed. The servo motors driving the robot operate in torque control mode, while the robot operates in zero-force control mode. When the robot is in zero-force control mode, it is dragged and moved. The robot's movement trajectory is sampled and recorded, along with the actual output torque and command torque of the servo motors at corresponding sampling points. A genetic algorithm and / or neural network are used to process each point to obtain a reproduced trajectory and corresponding compensation torques at each point in the reproduced trajectory. The robot is then put into reproduction mode. Based on the compensation torques at each point in the reproduced trajectory, a command torque is output, causing the robot to move along the reproduced trajectory.

[0009] Furthermore, the method for enabling the robot to operate in zero-force control mode includes the following steps:

[0010] In torque control mode, the actual output torque of the servo motor driving the robot joint is equal to the sum of the command torque output by the control system and the compensation torque of the servo motor, as expressed by the formula: τ op =τ imp +τ s ;

[0011] In the formula:

[0012] τ imp The command torque output by the control system;

[0013] τ op This refers to the actual output torque of the servo motor.

[0014] τ s This is the compensation torque for the servo motor;

[0015] The compensation torque τ of the servo motor s It was calculated using the robot's simplified dynamic equations;

[0016] The command torque τ output by the control system imp The following impedance control model was used for calculation:

[0017]

[0018] In the formula:

[0019] e represents the deviation between the actual joint position and its set value;

[0020] This represents the deviation between the actual joint speed and its set value.

[0021] This represents the deviation between the actual value of the joint acceleration and its set value.

[0022] M is the inertia coefficient of the servo motor under the impedance control model;

[0023] D is the damping coefficient of the servo motor under the impedance control model;

[0024] K is the stiffness coefficient of the servo motor under the impedance control model;

[0025] When the stiffness coefficient K of the servo motor is set to 0 under the impedance control model, the impedance control cannot reduce e, meaning the robot cannot return to the set initial position. At the same time, the set position of the joint is set to be the same as the real-time position. At this time, the output torque of the servo motor is equal to its compensation torque to maintain the robot's current pose and motion state. When there is no external force, the robot is in a stationary or uniform rotation state, which realizes zero-force control of the robot and makes the robot work in zero-force control mode.

[0026] Furthermore, the method for calculating the compensation torque using the robot's simplified dynamic equations includes the following steps:

[0027] The compensation torque of the servo motor is calculated using the following dynamic equation;

[0028] τ s =τ a +τ v +τ g ;

[0029]

[0030] τ v =m A r A g(cosθ1 cosθ2 cosθ3) T ;

[0031]

[0032]

[0033] In the formula:

[0034] τ s This is the compensation torque for the servo motor;

[0035] τ a This refers to the inertial term in the compensating torque;

[0036] τ v The velocity term in the compensating torque;

[0037] τ g This refers to the gravity term in the compensating torque.

[0038] m is the equivalent mass of the moving platform;

[0039] To accelerate the moving platform;

[0040] J is the Jacobian matrix;

[0041] J -T It is the transpose of the inverse of the Jacobian matrix;

[0042] I A The moment of inertia of the active arm relative to its rotation axis;

[0043] g is the acceleration due to gravity;

[0044] m A r A The product of the mass and radius of the active arm with respect to its axis of rotation;

[0045] m A The mass of the active arm relative to its pivot axis;

[0046] r A The radius of rotation of the active arm about its axis;

[0047] θ1, θ2, and θ3 correspond to the angular displacements of the three active joints;

[0048] It is a unit column vector in the vertical direction;

[0049] For a parallel robot servo motor, its torque output in torque mode is: τ op =τ imp +τ s When the parallel robot operates in zero-force control mode, the command torque τ output by the control system is... imp If the torque is 0, the actual output torque of the servo motor is equal to the compensation torque of the servo motor; therefore:

[0050] τ op =τ s .

[0051] Furthermore, when sampling and recording the robot's movement trajectory, the sampling period is 0.5 to 2 seconds.

[0052] Furthermore, when sampling and recording the robot's movement trajectory, the sampling period is changed, and sampling is repeated along the set trajectory.

[0053] Furthermore, a neural network is set up, the sampled data is compiled into training samples, the neural network is trained, and the trained neural network is used to predict and reproduce the compensation torque at each point in the trajectory.

[0054] Furthermore, after obtaining the reproduced trajectory, the reproduced trajectory is divided into several segments to obtain multiple segmentation points on the reproduced trajectory; the robot moves through the segmentation points and performs point-to-point displacement function.

[0055] The present invention also provides an apparatus for a three-degree-of-freedom high-speed parallel robot drag teaching method, comprising a memory and a processor, wherein the memory is used to store a computer program; and the processor is used to execute the computer program and, when executing the computer program, implement the steps of the three-degree-of-freedom high-speed parallel robot drag teaching method as described above.

[0056] The present invention also provides a storage medium storing a computer program, which, when executed by a processor, implements the steps of the three-degree-of-freedom high-speed parallel robot drag teaching method described above.

[0057] The advantages and positive effects of this invention are:

[0058] (1) The present invention proposes a three-degree-of-freedom high-speed parallel robot drag teaching method based on a zero-force control scheme with torque compensation. It does not require the use of additional sensors and external force observers. It only requires the use of robot body sensors and combined with robot dynamic equations to achieve zero-force drag of the robot, which is applicable to existing robots.

[0059] (2) This invention combines the advantages of impedance control and adjusts its parameters to improve the response speed of the parallel robot's compensation torque output during dragging.

[0060] (3) This invention conducts an engineering analysis of the robot sorting and grasping process, and uses a genetic algorithm to optimize the constructed reproducible motion trajectory. Under the condition of ensuring the shortest motion time, it can be better applied to the actual application of parallel robots. Attached Figure Description

[0061] Figure 1 This is a flowchart of a three-degree-of-freedom high-speed parallel robot drag teaching method according to the present invention.

[0062] Figure 2 This is a flowchart of trajectory recording and reproduction in the zero-force control mode of a parallel robot.

[0063] Figure 3 This is a schematic diagram of the robot's portal frame path.

[0064] In the diagram: P1 to P6 are the dividing points on the path of the door frame.

[0065] CST: Periodic synchronous torque mode for servo motors.

[0066] ON: Enable zero-force control.

[0067] OFF: Zero force control is off / not enabled.

[0068] FIFO: First-In-First-Out control function. Detailed Implementation

[0069] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0070] Please see Figures 1 to 3 A three-degree-of-freedom high-speed parallel robot drag teaching method is proposed. This method enables the servo motors driving the robot to operate in torque control mode, and the robot to operate in zero-force control mode. When the robot is operating in zero-force control mode, it is dragged to move. The robot's movement trajectory is sampled and recorded, along with the actual output torque and command torque of the robot's servo motors at the corresponding sampling points. A genetic algorithm and / or neural network are used to process each point to obtain a reproduced trajectory and the corresponding compensation torque at each point in the reproduced trajectory. The robot is then put into reproduction mode. Based on the compensation torque at each point in the reproduced trajectory, a command torque is output to make the robot move along the reproduced trajectory.

[0071] A position-based impedance control algorithm and a torque-compensated zero-force control method enable the robot to operate in a zero-force control mode. The robot is dragged and its trajectory is sampled and recorded. A genetic algorithm is used to process each point to obtain a reconstructed trajectory. The robot then enters a reconstructed mode via a corresponding command. Figure 3 For example, after obtaining the reproduced trajectory, running the point-to-point displacement function can reproduce the original trajectory relatively completely and accurately.

[0072] Appendix Figure 1 and appendix Figure 2 middle:

[0073] CST: Periodic Synchronous Torque Mode of Servo Motor, also known as Torque Control Mode. In this mode, the servo motor controller sends the target torque value and compensation torque value to the servo motor once in each fixed cycle.

[0074] ON and OFF: These are switch signals. When triggered, it is ON; when canceled, it is OFF. When the ON signal is issued, the zero-force control mode is turned on; when the OFF signal is issued, the zero-force control mode is turned off, and the system returns to the default working mode.

[0075] FIFO: First-In-First-Out control function, generally used for position-based motion control, and in this invention used to implement motion control when reproducing a trajectory.

[0076] Preferably, the method for enabling the robot to operate in a zero-force control mode may include the following steps:

[0077] In torque control mode, the actual output torque of the servo motor driving the robot joint is equal to the sum of the command torque output by the control system and the compensation torque of the servo motor, as expressed by the formula: τ op =τ imp +τ s ;

[0078] In the formula:

[0079] τ imp The command torque output by the control system;

[0080] τ op This refers to the actual output torque of the servo motor.

[0081] τ s This is the compensation torque for the servo motor;

[0082] The compensation torque τ of the servo motor s It was calculated using the robot's simplified dynamic equations;

[0083] The command torque τ output by the control system imp The following impedance control model was used for calculation:

[0084]

[0085] In the formula:

[0086] e represents the deviation between the actual joint position and its set value;

[0087] This represents the deviation between the actual joint speed and its set value.

[0088] This represents the deviation between the actual value of the joint acceleration and its set value.

[0089] M is the inertia coefficient of the servo motor under the impedance control model;

[0090] D is the damping coefficient of the servo motor under the impedance control model;

[0091] K is the stiffness coefficient of the servo motor under the impedance control model;

[0092] The stiffness coefficient K of the servo motor can be set to 0 under the impedance control model. Impedance control cannot reduce e, meaning the robot cannot return to the set initial position. At the same time, the set position of the joint is set to be the same as the real-time position. At this time, the output torque of the servo motor is equal to its compensation torque to maintain the robot's current posture and motion state. When there is no external force, the robot is in a stationary or uniform rotation state, which realizes zero-force control of the robot and makes the robot work in zero-force control mode.

[0093] The servo motor is controlled by a servo motor torque control module, which allows the robot's servo motor to operate in torque control mode. Alternatively, a servo motor driver with torque control mode can be used to drive the servo motor.

[0094] Preferably, the method for calculating the compensation torque using the robot's simplified dynamic equations may include the following steps:

[0095] The compensation torque of the servo motor is calculated using the following dynamic equation;

[0096] τ s =τ a +τ v +τ g ;

[0097]

[0098] τ v =m A r A g(cosθ1 cosθ2 cosθ3) T ;

[0099]

[0100]

[0101] In the formula:

[0102] τ s This is the compensation torque for the servo motor;

[0103] τ a This refers to the inertial term in the compensating torque;

[0104] τ v The velocity term in the compensating torque;

[0105] τ g This refers to the gravity term in the compensating torque.

[0106] m is the equivalent mass of the moving platform;

[0107] To accelerate the moving platform;

[0108] J is the Jacobian matrix;

[0109] J -T It is the transpose of the inverse of the Jacobian matrix;

[0110] I A The moment of inertia of the active arm relative to its rotation axis;

[0111] g is the acceleration due to gravity;

[0112] m A r A The product of the mass and radius of the active arm with respect to its axis of rotation;

[0113] m A The mass of the active arm relative to its pivot axis;

[0114] r A The radius of rotation of the active arm about its axis;

[0115] θ1, θ2, and θ3 correspond to the angular displacements of the three active joints;

[0116] It is a unit column vector in the vertical direction;

[0117] For a parallel robot servo motor, its torque output in torque mode is: τ op =τ imp +τ s When the parallel robot operates in zero-force control mode, the command torque τ output by the control system is... imp If the torque is 0, the actual output torque of the servo motor is equal to the compensation torque of the servo motor; therefore:

[0118] τ op =τ s .

[0119] Preferably, when sampling and recording the robot's movement trajectory, the sampling period can be 0.5 to 2 seconds. A sampling period of 1 second is preferred.

[0120] Preferably, when sampling and recording the robot's movement trajectory, the sampling period can be changed, and sampling can be repeated along the set trajectory.

[0121] Preferably, a neural network can be set up, the sampled data can be compiled into training samples, the neural network can be trained, and the trained neural network can be used to predict and reproduce the compensation torque at each point in the trajectory.

[0122] Preferably, after obtaining the reproduced trajectory, the reproduced trajectory can be divided into several segments to obtain multiple segmentation points on the reproduced trajectory; the robot can then move through the segmentation points and perform point-to-point displacement function.

[0123] The present invention also provides an apparatus for a three-degree-of-freedom high-speed parallel robot drag teaching method, comprising a memory and a processor, wherein the memory is used to store a computer program; and the processor is used to execute the computer program and, when executing the computer program, implement the steps of the three-degree-of-freedom high-speed parallel robot drag teaching method as described above.

[0124] The present invention also provides a storage medium storing a computer program, which, when executed by a processor, implements the steps of the three-degree-of-freedom high-speed parallel robot drag teaching method described above.

[0125] The workflow and working principle of the present invention will be further described below with reference to a preferred embodiment:

[0126] A three-degree-of-freedom high-speed parallel robot drag teaching method is characterized by: operating the servo motor driving the robot in torque control mode, operating the robot in zero-force control mode, dragging the robot while it is in zero-force control mode, sampling and recording the robot's movement trajectory, and recording the actual output torque and command torque of the robot's servo motor at the corresponding sampling points; processing each point using a genetic algorithm and / or neural network to obtain the reproduced trajectory and the compensation torque at each point in the reproduced trajectory; operating the robot in reproduction mode; and outputting the command torque based on the compensation torque at each point in the reproduced trajectory to make the robot move along the reproduced trajectory.

[0127] The dynamic equations of the parallel robot model are set in order to obtain the motor driving torque, which prepares for torque compensation in subsequent zero-force control.

[0128] Assume that all joint rotations of the parallel robot are driven by servo motors; combining the robot's structural parameters, the inertia and center of gravity of each component, the robot's dynamic equations are constructed using the principle of virtual work, and the equations are as follows:

[0129]

[0130] Where θ is the joint angular displacement. The joint angular velocity, Let θ be the joint angular acceleration. M(θ) is the inertia matrix. Let G(θ) be the Coriolis force and centrifugal force matrix, and G(θ) be the gravity term. When the parallel robot operates in zero-force control mode, the actual output torque of the servo motor is equal to the compensation torque of the servo motor; τ s This is the compensation torque of the servo motor, which is also the joint driving torque.

[0131] Based on the dynamic equations of the parallel robot and its position θ and velocity when it is being dragged. acceleration The information is used to obtain the joint driving torque. To facilitate the subsequent calculation of the servo motor compensation torque, the dynamic equations of the parallel robot are simplified. The simplified dynamic equations are as follows:

[0132] τ s =τ a +τ v +τg ;

[0133]

[0134] τ v =m A r A g(cosθ1 cosθ2 cosθ3) T ;

[0135]

[0136]

[0137] In the formula:

[0138] τ s This is the compensation torque for the servo motor;

[0139] τ a This refers to the inertial term in the compensating torque;

[0140] τ v The velocity term in the compensating torque;

[0141] τ g This refers to the gravity term in the compensating torque.

[0142] m is the equivalent mass of the moving platform;

[0143] To accelerate the moving platform;

[0144] J is the Jacobian matrix;

[0145] J -T It is the transpose of the inverse of the Jacobian matrix;

[0146] I A The moment of inertia of the active arm relative to its rotation axis;

[0147] g is the acceleration due to gravity;

[0148] m A r A The product of the mass and radius of the active arm with respect to its axis of rotation;

[0149] m A The mass of the active arm relative to its pivot axis;

[0150] r A The radius of rotation of the active arm about its axis;

[0151] θ1, θ2, and θ3 correspond to the angular displacements of the three active joints;

[0152] It is a unit column vector in the vertical direction;

[0153] These include the gravity term at the end effector and the gravity term of the active arm; the remaining parameters are constants and are related to the robot's characteristics.

[0154] Based on the principle of impedance control, the relationship between the force on the robot joint and its position, expressed by the position-based impedance control algorithm, is as follows:

[0155]

[0156] In the formula:

[0157] e represents the deviation between the actual joint position and its set value;

[0158] This represents the deviation between the actual joint speed and its set value.

[0159] This represents the deviation between the actual value of the joint acceleration and its set value.

[0160] M is the inertia coefficient of the servo motor under the impedance control model;

[0161] D is the damping coefficient of the servo motor under the impedance control model;

[0162] K is the stiffness coefficient of the servo motor under the impedance control model;

[0163] For a parallel robot servo motor, its torque output in torque mode is:

[0164] τ op =τ imp +τ s ;

[0165] In the formula:

[0166] τ imp The command torque output by the control system;

[0167] τ op This refers to the actual output torque of the servo motor.

[0168] τ s This is the compensation torque for the servo motor;

[0169] If the parallel robot needs to operate in zero-force control mode, the command torque τ output by the control system... imp It must be 0, that is:

[0170] τ s =τ op ;

[0171] The compensation torque τ of the servo motor sThe calculations obtained from the above kinetic equations are relatively accurate.

[0172] An improvement is made to the position-based impedance control algorithm by setting the stiffness coefficient K of the servo motor to 0 under the impedance control model. At this time, the servo motor only outputs the driving torque calculated by the dynamic equation as the compensation torque. The real-time position is set to the set position. At this time, the deviation e between the actual value of the joint position and its set value is always 0. At this time, the servo motor only outputs the compensation torque, which completes the zero-force control based on torque compensation.

[0173] Upon entering zero-force control mode, the system begins recording the trajectory. The sampling period is adjusted appropriately based on the trajectory length, recording the coordinates of a portion of the trajectory during dragging. After dragging is complete, trajectory recording is confirmed, and the coordinates of each point are saved to obtain the initial reproducible trajectory. A genetic algorithm is used to process adjacent points, searching for points-to-point paths based on the shortest movement time, thus obtaining an optimized complex trajectory. After obtaining the reproducible trajectory, linear interpolation and circular interpolation methods are used to divide the trajectory into several segments, obtaining multiple segmentation points on the trajectory. Point-to-point displacement is used to ensure the robot moves through these segmentation points. After completing the above process, the robot enters reproducible mode via corresponding commands. Figure 3 For example, after obtaining the reproduced trajectory, running the point-to-point displacement function can reproduce the required original trajectory relatively completely and accurately.

[0174] The aforementioned algorithms and functional modules, such as genetic algorithms, neural networks, linear interpolation, circular interpolation, point-to-point displacement, trajectory recording, position-based impedance control algorithms, and servo motor torque control modules, can all adopt algorithms and functional modules applicable in the existing technology; or they can adopt algorithms and functional modules in the existing technology and be constructed using conventional technical means.

[0175] The embodiments described above are only used to illustrate the technical ideas and features of the present invention. Their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The patent scope of the present invention should not be limited by these embodiments. That is, any equivalent changes or modifications made in accordance with the spirit disclosed in the present invention still fall within the patent scope of the present invention.

Claims

1. A method for teaching a three-degree-of-freedom high-speed parallel robot by dragging, characterized in that, The servo motors driving the robot operate in torque control mode, while the robot operates in zero-force control mode. When the robot is operating in zero-force control mode, it is dragged to move. The robot's movement trajectory is sampled and recorded, and the actual output torque and command torque of the robot's servo motors at the corresponding sampling points are recorded. Genetic algorithms and / or neural networks are used to process each point to obtain the reproduced trajectory and the corresponding compensation torque at each point in the reproduced trajectory. The robot is then put into reproduction mode. Based on the compensation torque at each point in the reproduced trajectory, the command torque is output to make the robot move along the reproduced trajectory. The method for enabling a robot to operate in zero-force control mode includes the following steps: In torque control mode, the actual output torque of the servo motor driving the robot joint is equal to the sum of the command torque output by the control system and the compensation torque of the servo motor. The relationship is as follows: ; In the formula: The command torque output by the control system; This refers to the actual output torque of the servo motor. This is the compensation torque for the servo motor; Compensation torque of servo motor It was calculated using the robot's simplified dynamic equations; Command torque output by the control system The following impedance control model was used for calculation: ； In the formula: This represents the deviation between the actual joint position and its set value. This represents the deviation between the actual joint speed and its set value. This represents the deviation between the actual value of the joint acceleration and its set value. The inertia coefficient of the servo motor under the impedance control model; Here is the damping coefficient of the servo motor under the impedance control model; This represents the stiffness coefficient of the servo motor under the impedance control model. The stiffness coefficient of the servo motor under the impedance control model Setting it to 0, impedance control cannot be enabled. Reduce, meaning the robot cannot return to the set initial position, and at the same time set the joint's set position is the same as the real-time position. At this time, the servo motor output torque is equal to its compensation torque to maintain the robot's current posture and motion state. When there is no external force, it is in a stationary or uniform rotation state, that is, to achieve zero force control of the robot and make the robot work in zero force control mode. The method for calculating the compensation torque using the simplified dynamic equations of the robot includes the following steps: The compensation torque of the servo motor is calculated using the following dynamic equation; ； ； ； ； = In the formula: This is the compensation torque for the servo motor; This refers to the inertial term in the compensating torque; The velocity term in the compensating torque; This refers to the gravity term in the compensating torque. For the equivalent quality of the moving platform; To accelerate the moving platform; It is a Jacobian matrix; It is the transpose of the inverse of the Jacobian matrix; The moment of inertia of the active arm relative to its rotation axis; It is the acceleration due to gravity; The product of the mass and radius of the active arm with respect to its axis of rotation; The mass of the active arm relative to its pivot axis; The radius of rotation of the active arm about its axis; , , This corresponds to the angular displacement of the three active joints; It is a unit column vector in the vertical direction; For a parallel robot servo motor, its torque output in torque mode is: When the parallel robot operates in zero-force control mode, the command torque output by the control system is... If the torque is 0, the actual output torque of the servo motor is equal to the compensation torque of the servo motor; therefore: 。 2. The three-degree-of-freedom high-speed parallel robot drag teaching method according to claim 1, characterized in that, When sampling and recording the robot's movement trajectory, the sampling period is 0.5 to 2 seconds.

3. The three-degree-of-freedom high-speed parallel robot drag teaching method according to claim 1, characterized in that, When sampling and recording the robot's movement trajectory, the sampling period is changed, and sampling is repeated along the set trajectory.

4. The three-degree-of-freedom high-speed parallel robot drag teaching method according to claim 1, characterized in that, Set up a neural network, compile the sampled data into training samples, train the neural network, and use the trained neural network to predict and reproduce the compensation torque at each point in the trajectory.

5. The three-degree-of-freedom high-speed parallel robot drag teaching method according to claim 1, characterized in that, After obtaining the reproduced trajectory, the reproduced trajectory is divided into several segments, resulting in multiple segmentation points on the reproduced trajectory; the robot moves through the segmentation points and performs point-to-point displacement function.

6. A device for a three-degree-of-freedom high-speed parallel robot drag teaching method, comprising a memory and a processor, characterized in that, The memory is used to store computer programs; the processor is used to execute the computer programs and, when executing the computer programs, implement the steps of the three-degree-of-freedom high-speed parallel robot drag teaching method as described in any one of claims 1 to 5.

7. A storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the three-degree-of-freedom high-speed parallel robot drag teaching method as described in any one of claims 1 to 5.

Images