Vision and force sense fused collaborative robot tool self-adaptive passivation method

Abstract

The invention relates to the technical field of machining, and discloses a visual sense and force sense fused collaborative robot tool self-adaptive passivation method which comprises the steps that when it is detected that a collaborative robot grabs a tool, three-dimensional feature information of the tool is extracted; obtaining the current contour of the wheel brush, and calculating the abrasion loss; determining a target attitude and a target position; operating the joint torque observer, and estimating the current contact force between the tail end of the robot and the wheel brush; moving to a target position in a target posture, and when the current contact force reaches the initial contact force, performing main movement along the edge line direction; if a force error exists between the current contact force and the target passivation force, generating a dynamic compensation motion instruction by using a PID algorithm, and planning the motion of the collaborative robot by combining the main motion and the compensation motion; according to the high-precision constant-force passivation method, machine vision and collaborative robot force perception are combined, high-precision constant-force passivation is achieved, the efficiency is improved, and the labor cost is reduced.

Description

An Adaptive Passivation Method for Collaborative Robot Tools Based on Vision and Force Sensing Technical Field

[0001] This invention relates to the field of machining technology, specifically to an adaptive blunting method for collaborative robot cutting tools that integrates vision and force perception. Background Technology

[0002] In modern manufacturing, cemented carbide cutting tools are core tools in machining. However, traditional tool passivation relies on the "feel" of skilled workers, resulting in significant deficiencies in automation of the cutting edge passivation process. It is inefficient, costly, and cannot meet the needs of large-scale production. Summary of the Invention

[0003] This invention provides a collaborative robot tool adaptive passivation method that integrates vision and force perception to solve the problems of low efficiency and high cost caused by the reliance on manual passivation processes in existing technologies.

[0004] In a first aspect, the present invention provides a vision-force fusion collaborative robot tool adaptive passivation method, applied to a vision-force fusion collaborative robot tool adaptive passivation system. The system includes a collaborative robot, a gripper, a tool, and a brush wheel. The joints of the collaborative robot have torque sensing capabilities, and the gripper is mounted on the end effector of the collaborative robot. The method includes: when the collaborative robot detects that it has grasped the tool, acquiring an image and scanning result of the tool, and extracting the tool's three-dimensional feature information; acquiring the current contour data of the brush wheel, comparing the current contour data with the initial contour data of the brush wheel, and calculating the wear amount; and fusing the three-dimensional features of the tool. Based on information and wear level, determine the target posture and position; run the joint torque observer and use it to estimate the current contact force between the collaborative robot's end effector and the brush wheel; control the collaborative robot to move towards the target position in the target posture, and when the current contact force reaches the initial contact force, control the collaborative robot to perform the main motion along the blade line direction in a constant force control mode; if there is a force error between the current contact force and the target passivation force, use the PID algorithm to generate dynamic compensation motion commands, and combine the main motion and compensation motion to plan the collaborative robot's motion; when passivation of the completed blade line trajectory is detected, control the collaborative robot to retreat to a safe position.

[0005] This invention extracts the three-dimensional features of the cutting tool to eliminate initial errors, calculates wear based on the contour, and dynamically compensates for tool wear, avoiding quality degradation caused by brush wear. By integrating the three-dimensional features and wear of the cutting tool, it provides precise pose for machining, avoiding the defects of one-cut machining. Without relying on external sensors, it achieves high-precision contact force sensing, controls the collaborative robot to move towards the target position in the target posture, and combines the main motion and the compensation motion generated by the PID algorithm with machine vision and the force perception of the collaborative robot to achieve high-precision constant force passivation. Compared with traditional methods, it effectively improves efficiency and reduces labor costs.

[0006] In one optional implementation, the system further includes a one-dimensional sensor and a vision camera. The three-dimensional feature information includes the spatial vector of the cutting edge to be blunted and the effective length of the tool extending from the end of the gripper. The system acquires images and scanning results of the tool and extracts the three-dimensional feature information of the tool, including: when the tool enters the detection range of the one-dimensional sensor, the one-dimensional sensor scans the tool and measures the effective length of the tool extending from the end of the gripper; the vision camera captures images of the tool, and the image processing algorithm is used to identify the spatial vector of the cutting edge to be blunted. The spatial vector is used to characterize the path and position of the cutting edge.

[0007] This invention uses a one-dimensional sensor to accurately measure the effective length of the cutting tool extending from the end of the gripper, eliminating the need for manual intervention. It is applicable to non-standard clamping scenarios, improving versatility. By capturing images of the cutting tool with a vision camera, the spatial vector of the cutting edge to be blunted is identified, avoiding the limitations of subjective human judgment and achieving the quantification and standardization of cutting edge positioning.

[0008] In one optional implementation, determining the target posture includes: calculating the end-effector posture based on the spatial vector of the blade to be blunted, the end-effector posture being used to characterize the posture of the collaborative robot when the blade to be blunted is parallel to the rotation axis of the brush; determining an additional posture based on preset process parameters, the additional posture being used to characterize the tilt angle when the tool axis is tilted to the horizontal plane at the same angle as the process parameters; and superimposing the end-effector posture and the additional posture to obtain the target posture.

[0009] This invention determines the end-effector posture based on the spatial vector of the blade line to be passivated, ensuring that the blade line and the rotation axis of the brush are parallel and achieving uniform contact. It also determines additional postures based on preset process parameters to match different passivation requirements. Combining the end-effector posture and additional postures, the target posture of the collaborative robot is determined, taking into account both uniformity and customization, and achieving automatic posture planning.

[0010] In one optional implementation, the current contact force between the end effector of the collaborative robot and the wheel brush is estimated using a joint torque observer, including: establishing the dynamic equations of the collaborative robot; obtaining the real-time position and real-time velocity of the collaborative robot joints, calculating the real-time acceleration based on the real-time velocity, substituting the real-time position, real-time velocity, and real-time acceleration into the dynamic equations to calculate the theoretical torque; determining the difference between the theoretical torque and the actual output torque as the residual torque; estimating the contact force corresponding to the residual torque using the pseudo-inverse of the Jacobian matrix transpose, and determining the estimated contact force as the current contact force between the end effector of the collaborative robot and the wheel brush.

[0011] This invention uses dynamic equations to calculate theoretical torque and calculates the difference between theoretical torque and actual torque as residual torque to estimate the contact force between the end effector of the collaborative robot and the wheel brush, providing a data basis for constant force passivation control.

[0012] In one alternative implementation, the theoretical torque is calculated according to the following formula:

[0013] in, For theoretical torque, The inertia matrix, Let be the acceleration vectors of each joint of the collaborative robot. For the Coriolis force and centrifugal force terms, This represents the velocity vectors of each joint of the collaborative robot. For the gravitational torque term, The torque represents the joint friction force; the contact force is calculated using the following formula:

[0014] in, For contact force, For residual torque, It is the pseudo-inverse of the transpose of the Jacobian matrix.

[0015] In one optional implementation, a dynamic compensation motion command is generated using a PID algorithm, including: determining the product of proportional gain and force error as a proportional term; determining the product of integral gain and integral force error as an integral term; determining the product of differential gain and rate of change of force error as a differential term; and determining the sum of the proportional term, integral term, and differential term as the compensation motion command.

[0016] This invention utilizes proportional, integral, and differential terms to dynamically adjust force errors in multiple dimensions, generating precise compensating motion commands to correct contact force errors between the tool and the brush wheel.

[0017] In one optional implementation, the collaborative robot motion is planned by combining the main motion and the compensation motion, including: applying the velocity value corresponding to the compensation motion command to the normal direction perpendicular to the brush surface and pointing to the center of the brush; superimposing the velocity vector corresponding to the main motion and the velocity vector corresponding to the compensation motion command after the direction is applied to obtain the final command, and driving the collaborative robot motion planning according to the final command.

[0018] This invention avoids the limitations of single trajectory control by superimposing the main motion and the compensation motion, and achieves complete blunting of the tool cutting edge by coordinating the main motion and the compensation motion.

[0019] Secondly, this invention provides a vision-force fusion collaborative robot tool adaptive passivation device, applied to a vision-force fusion collaborative robot tool adaptive passivation system. The system includes a collaborative robot, a gripper, a tool, and a brush wheel. The joints of the collaborative robot have torque sensing capabilities, and the gripper is installed at the end of the collaborative robot. The device includes: an extraction module, used to acquire an image and scan result of the tool when the collaborative robot is detected to have grasped the tool, and extract the tool's three-dimensional feature information; a comparison module, used to acquire the current contour data of the brush wheel, compare the current contour data with the initial contour data of the brush wheel, and calculate the wear amount; and a determination module, used to fuse the tool's three-dimensional feature information and wear. The system comprises the following modules: a measurement module to determine the target posture and position; an estimation module to run the joint torque observer and estimate the current contact force between the collaborative robot's end effector and the brush wheel; a main motion module to control the collaborative robot to move towards the target position in the target posture, and to control the collaborative robot to perform main motion along the blade direction in a constant force control mode when the current contact force reaches the initial contact force; a motion planning module to generate dynamic compensation motion commands using a PID algorithm if there is a force error between the current contact force and the target passivation force, and to plan the collaborative robot's motion by combining the main motion and the compensation motion; and a retraction module to control the collaborative robot to retract to a safe position when passivation of the completed blade trajectory is detected.

[0020] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the vision and force fusion collaborative robot tool adaptive passivation method described in the first aspect or any corresponding embodiment.

[0021] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the vision and force fusion adaptive passivation method for collaborative robot tools described in the first aspect or any corresponding embodiment. Attached Figure Description

[0022] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 is a structural schematic diagram of a collaborative robot tool adaptive passivation system based on vision and force fusion according to an embodiment of the present invention; Figure 2 is a flowchart of a collaborative robot tool adaptive passivation method based on vision and force fusion according to an embodiment of the present invention; Figure 3 is a schematic diagram of an adaptive passivation process based on an embodiment of the present invention; Figure 4 is a flowchart of constant force passivation control based on an embodiment of the present invention; Figure 5 is a structural block diagram of a collaborative robot tool adaptive passivation device based on vision and force fusion according to an embodiment of the present invention; Figure 6 is a hardware structure schematic diagram of an electronic device according to an embodiment of the present invention.

[0024] The attached figures are labeled 00, mounting plate; 10, collaborative robot; 20, gripper; 100, cutting tool; 30, brush passivation mechanism; 31, brush; 40, multimodal sensing mechanism; 41, vision camera; 42, one-dimensional sensor; 43, laser rangefinder; and 50, collaborative robot controller. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0027] In related technologies, automation techniques using dedicated rigid fixtures or basic robot teaching and reproduction are employed. However, existing automation technologies are essentially still "rigid" executions, with the following drawbacks: First, they cannot achieve self-adaptation. Existing solutions lack the ability to perceive the dynamic wear of the brush wheel during use and the minute positional errors of the tool clamping, resulting in extremely unstable actual passivation contact force and affecting processing quality. Second, they have poor production flexibility. When changing to different types of tools, it is necessary to replace and adjust the dedicated fixture or perform tedious robot teaching again, which cannot adapt to small-batch, multi-variety production modes.

[0028] This invention provides a collaborative robot tool adaptive passivation method that integrates vision and force perception. By integrating machine vision, ranging sensing, and the force perception capabilities of the robot body, it achieves intelligent closed-loop control of the entire passivation process.

[0029] According to an embodiment of the present invention, a method for adaptive blunting of collaborative robot cutting tools by fusion of vision and force is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] This embodiment provides a vision and force fusion collaborative robot tool adaptive passivation method, which is applied to a vision and force fusion collaborative robot tool adaptive passivation system, as shown in Figure 1. The system includes a mounting plate 00, a collaborative robot 10, a gripper 20, a tool 100, a brush passivation mechanism 30, a brush 31, a multimodal sensing mechanism 40, a vision camera 41, a one-dimensional sensor 42, a laser rangefinder 43, and a collaborative robot controller 50.

[0031] The collaborative robot 10, as the core of the entire process, has torque sensing capabilities in its joints. The gripper 20 is a general-purpose cutting tool 100, mounted at the end of the collaborative robot 10 to hold the tool 100 to be processed. The brush passivation mechanism 30 includes a brush 31, which is driven by a motor to rotate at high speed. The multimodal sensing mechanism 40 comprehensively acquires the status of the workpiece and tool, including a vision camera 41, a one-dimensional sensor 42 for measuring the length of the cutting tool 100, and a laser rangefinder 43 for measuring the wear of the brush 31. The collaborative robot controller 50 integrates and runs the core control algorithm. The mounting plate 00 is used to mount the worktable, the collaborative robot 10, the brush passivation mechanism 30, and the multimodal sensing mechanism 40. The space inside the cabinet below the mounting plate 00 can be used to place the robot controller and store related cables.

[0032] Figure 2 is a flowchart of a collaborative robot tool adaptive passivation method based on vision and force fusion according to an embodiment of the present invention. As shown in Figure 2, the process includes the following steps: Step S201, when the collaborative robot is detected to have grasped the tool, the image and scanning results of the tool are acquired, and the three-dimensional feature information of the tool is extracted.

[0033] In this embodiment of the invention, after the collaborative robot is detected to have grasped the tool, the collaborative robot is moved to the field of view of the multimodal sensing mechanism to collect images and scanning results of the tool, and the three-dimensional feature information of the tool is extracted from the images and scanning results.

[0034] Step S202: Obtain the current contour data of the brush wheel, compare the current contour data with the initial contour data of the brush wheel, and calculate the wear amount.

[0035] In this embodiment of the invention, to compensate for the inevitable wear of the brush wheel during use, wear detection is performed periodically. As shown in Figure 1, a laser rangefinder sensor is installed on the side of the brush wheel, with its installation direction consistent with the radial direction of the brush wheel, to scan or measure multiple points on the brush wheel surface. The laser rangefinder sensor collects the current outer diameter or surface contour data of the brush wheel, which is the current contour data. , will the current contour data and the initial contour data of the brush stored in the database. Compare and calculate the amount of wear. This will serve as a key basis for compensation in subsequent location planning. .

[0036] Step S203: Fuse the three-dimensional feature information and wear amount of the tool to determine the target attitude and target position.

[0037] In this embodiment of the invention, the three-dimensional feature information and wear amount of the tool obtained in steps S201 and S202 are fused to make intelligent decisions and dynamically generate the contact pose for this machining operation. This pose includes the target pose. and target location .

[0038] Step S204: Run the joint torque observer and use the joint torque observer to estimate the current contact force between the collaborative robot end effector and the wheel brush.

[0039] In this embodiment of the invention, the collaborative robot controller operates a joint torque observer based on the robot's dynamics model and real-time acquired joint torque, position, and velocity information. This observer can accurately estimate the contact force between the collaborative robot's end effector and the external environment (i.e., the wheel brush) in real time, without external force sensors. The current contact force between the collaborative robot's end effector and the wheel brush is estimated using this joint torque observer. .

[0040] Step S205: Control the collaborative robot to move towards the target position in the target posture. When the current contact force reaches the initial contact force, control the collaborative robot to perform the main motion along the blade line direction in a constant force control mode.

[0041] In this embodiment of the invention, the collaborative robot is controlled to achieve the target posture obtained in step S203. Towards the target location During the movement, the joint torque observer continues to operate, recording the current contact forces. With the preset initial contact force Compare them. Under the current contact force... Achieving a small initial contact force At this point, the collaborative robot confirms precise contact and begins passivation based on this. The collaborative robot then enters constant force control mode and begins its main motion along the blade line direction. .

[0042] Step S206: If there is a force error between the current contact force and the target passivation force, then use the PID algorithm to generate dynamic compensation motion commands, and combine the main motion and compensation motion to plan the collaborative robot motion.

[0043] In an embodiment of the invention, as shown in Figure 3, during the main motion of the collaborative robot, the current contact force is monitored in real time. and the preset target passivation force Compare, if the current contact force and the preset target passivation force The difference lies in the force error between the current contact force and the preset target passivation force. And force error If the value exceeds a preset threshold, a dynamic compensation motion command is generated in the normal direction of the brush. Then, combining the main motion from step S205, plan the collaborative robot's motion. If there is a force error... If the value is less than the preset threshold, it is determined whether the cutting edge trajectory has been completed. If the cutting edge trajectory has been completed, the process jumps to step S207. If the cutting edge trajectory has not been completed, the process returns to step S204.

[0044] Step S207: When the blunting of the completed blade trajectory is detected, control the collaborative robot to retreat to a safe position.

[0045] In this embodiment of the invention, after the blade trajectory is blunted, the collaborative robot smoothly retreats to a safe position along the brush normal direction, ending the current processing cycle.

[0046] The vision and force sensing fusion collaborative robot tool adaptive passivation method provided in this embodiment extracts the tool's three-dimensional features to eliminate initial errors, calculates wear based on the contour to dynamically compensate for tool wear, and avoids quality degradation caused by brush wear. By fusing the tool's three-dimensional features and wear amount, it provides precise pose for machining, avoiding the defects of one-cut machining. It achieves high-precision contact force sensing without relying on external sensors, controls the collaborative robot to move towards the target position in the target posture, and combines the main motion and the compensation motion generated by the PID algorithm with machine vision and collaborative robot force sensing to achieve high-precision constant force passivation. Compared with traditional methods, it effectively improves efficiency and reduces labor costs.

[0047] This embodiment provides a collaborative robot tool adaptive blunting method that integrates vision and force perception. The process includes the following steps: Step S301, when the collaborative robot is detected to have grasped the tool, the image and scanning results of the tool are acquired, and the three-dimensional feature information of the tool is extracted.

[0048] Specifically, step S301 includes: step S3011, when the tool reaches the detection range of the one-dimensional sensor, the one-dimensional sensor is used to scan the tool and measure the effective length of the tool extending from the end of the gripper.

[0049] Step S3012: Use a vision camera to capture an image of the cutting tool and use an image processing algorithm to identify the spatial vector of the cutting edge to be blunted.

[0050] In this embodiment of the invention, the three-dimensional feature information includes the spatial vector of the cutting edge to be blunted and the effective length of the tool extending from the end of the gripper.

[0051] As shown in Figure 3, the one-dimensional sensor uses a fork-shaped fiber optic sensor. When the tool enters the sensor's detection range, the one-dimensional sensor sends a signal to the collaborative robot, indicating that the depth position has been reached. The tool is scanned, and the effective length of the tool extending from the end-effector is accurately measured using the collaborative robot's end-tool coordinate system. .

[0052] After the cutting tool triggers the one-dimensional sensor, a vision camera, acting as another sensor, captures an image of the cutting edge. Using adaptive threshold segmentation combined with an edge detection algorithm, the edge of the cutting line is precisely separated from the background from the cutting edge image, and the spatial vector of the cutting line to be blunted is identified. This vector is used to characterize the path and position of the cutting edge.

[0053] One-dimensional sensors accurately measure the effective length of the tool extending from the end of the gripper, eliminating the need for manual intervention. This method is suitable for non-standard clamping scenarios, improving versatility. A vision camera captures tool images and identifies the spatial vector of the cutting edge to be blunted, avoiding the limitations of subjective human judgment and achieving the quantification and standardization of cutting edge positioning.

[0054] Step S302: Obtain the current contour data of the brush wheel, compare the current contour data with the initial contour data of the brush wheel, and calculate the wear amount.

[0055] For details, please refer to step S202 of the embodiment shown in Figure 2, which will not be repeated here.

[0056] Step S303: Combine the three-dimensional feature information and wear amount of the tool to determine the target attitude and target position.

[0057] Specifically, generating the target pose in step S303 includes: step S3031, calculating the end pose based on the spatial vector of the blade to be blunted.

[0058] Step S3032: Determine the additional posture according to the preset process parameters.

[0059] Step S3033: Superimpose the end effector attitude and the additional attitude to obtain the target attitude.

[0060] In this embodiment of the invention, firstly, based on the spatial vector of the cutting edge to be blunted... The robot end effector posture that allows the cutting edge to remain precisely parallel to the axis of rotation of the brush can be calculated. .

[0061] Based on preset process parameters, calculate the additional posture that makes the tool axis form a specific upward angle with the horizontal plane. For example, if the preset process parameter is an inclination angle of 20°, the corresponding additional posture is... It's 20°, which reflects the final effect of the tool tip tilting upwards by 20°.

[0062] By superimposing the end-effector attitude and the additional attitude, the final target attitude is obtained. : .

[0063] The end effector posture is determined by the spatial vector of the blade line to be blunted to ensure that the blade line and the rotation axis of the brush are parallel and achieve uniform contact. The additional posture is determined by the preset process parameters to match different blunting requirements. The target posture of the collaborative robot is determined by combining the end effector posture and the additional posture, taking into account both uniformity and customization, and realizing automatic posture planning.

[0064] Specifically, the target position is determined as follows: based on the effective length of the tool. and the current contour of the brush wheel (Wear compensation already included) ), calculate the target attitude The target position in three-dimensional space where the tool cutting edge comes into contact with the brush surface under the given posture. The calculation process is automatically performed by the collaborative robot's controller, which calculates the required rotation angles for each axis of the collaborative robot. Input parameters include the effective tool length, wear compensation, and the final position that the tool coordinate system at the tool tip needs to reach. The movement patterns of each axis of the robot are calculated automatically by the collaborative robot, ultimately determining the point at which the tool tip reaches the specified target.

[0065] Step S304: Run the joint torque observer and use the joint torque observer to estimate the current contact force between the collaborative robot end effector and the wheel brush.

[0066] Specifically, the step S304 above, which uses a joint torque observer to estimate the current contact force between the end effector of the collaborative robot and the wheel brush, includes: step S3041, establishing the dynamic equations of the collaborative robot.

[0067] Step S3042: Obtain the real-time position and real-time velocity of the collaborative robot joints, calculate the real-time acceleration based on the real-time velocity, and substitute the real-time position, real-time velocity, and real-time acceleration into the dynamic equation to calculate the theoretical torque.

[0068] Step S3043: The difference between the theoretical torque and the actual output torque is determined as the residual torque.

[0069] Step S3044: The pseudo-inverse of the Jacobian matrix transpose is used to estimate the contact force corresponding to the residual torque, and the estimated contact force is determined as the current contact force between the end effector of the collaborative robot and the wheel brush.

[0070] In this embodiment of the invention, the dynamic equations of the collaborative robot are first established: For a collaborative robot with N joints, its dynamic equations are expressed as follows:

[0071] in, The total torque vector required to drive the movement of each joint of the collaborative robot. , , Here, represents the position, velocity, and acceleration vectors of each joint of the collaborative robot, and is the inertia matrix. Used to describe the mass and inertia distribution of each link in a collaborative robot, which changes as the robot's posture changes. For the Coriolis force and centrifugal force terms, This is the gravitational torque term, used to describe the torque exerted by gravity on the links of the collaborative robot. This refers to the joint friction torque, used to describe the frictional force generated during joint rotation. For the contact force at the end of the collaborative robot, The transpose of the Jacobian matrix for collaborative robots is used to convert the contact force at the end effector. This is mapped to the corresponding torque generated at each joint.

[0072] Then, the theoretical torque is calculated. In each control cycle (typically 1 millisecond) of the collaborative robot's actual operation, the controller performs the following calculations: First, it obtains the real-time positions of all joints from the collaborative robot's dual encoders. and real-time speed By measuring real-time speed Differentiate to obtain real-time acceleration The second step is to obtain the location. ,speed acceleration Substituting the part of the dynamic equation that does not include external contact forces, the theoretical torque that should be generated if the collaborative robot is simply moving in free space according to its current state is calculated as follows:

[0073] in, This is the theoretical torque.

[0074] The real-time output current is read from the servo driver of the motor driving each joint. The actual output torque of the joint is calculated using the motor torque constant. For collaborative arms with built-in joint torque sensors, the torque measured by the sensors can be read directly.

[0075] Subtract the actual output torque from the theoretically calculated torque to obtain the residual torque. :

[0076] The residual torque is almost entirely generated by the contact between the collaborative robot's end effector and the external environment (wheel brush).

[0077] according to Estimate the contact force using the following formula:

[0078] in, It is the pseudo-inverse of the transpose of the Jacobian matrix. The pseudo-inverse is used because... It may not be a square matrix, or it may be irreversible at some singular points, while the pseudo-inverse provides a mathematically stable and meaningful solution.

[0079] By using dynamic equations to calculate the theoretical torque and calculating the difference between the theoretical torque and the actual torque as the residual torque, the contact force between the end effector of the collaborative robot and the wheel brush is estimated, providing a data basis for constant force passivation control.

[0080] Step S305: Control the collaborative robot to move towards the target position in the target posture. When the current contact force reaches the initial contact force, control the collaborative robot to perform the main motion along the blade line direction in a constant force control mode.

[0081] For details, please refer to step S205 of the embodiment shown in Figure 2, which will not be repeated here.

[0082] Step S306: If there is a force error between the current contact force and the target passivation force, then use the PID algorithm to generate dynamic compensation motion commands, and combine the main motion and compensation motion to plan the collaborative robot motion.

[0083] Specifically, the above step S306, which uses the PID algorithm to generate dynamic compensation motion commands, includes: step S3061, determining the product of the proportional gain and the force error as the proportional term.

[0084] Step S3062: The product of the integral gain and the integral result of the force error is determined as the integral term.

[0085] Step S3063: The product of the differential gain and the rate of change of the force error is determined as the differential term.

[0086] Step S3064: The sum of the proportional term, integral term, and differential term is determined as the compensation motion command.

[0087] In this embodiment of the invention, the proportional term provides a compensation speed proportional to the magnitude of the current force error. If the force error is large, the proportional term is also large, driving the robot to quickly move closer to the brush. If the force error is small, the proportional term is also small, allowing the robot to make fine adjustments.

[0088] The formula for calculating the proportional term is:

[0089] in, The proportional gain determines the "aggressiveness" of the controller's response. A larger proportional gain results in a faster response, but an excessively large proportional gain may cause oscillations and reduce stability. This is the force error.

[0090] When the blunting of the completed blade trajectory is detected, the robot is controlled to retreat to a safe position.

[0091] The proportional term serves as the primary driving force, used to quickly reduce most errors.

[0092] If only proportional terms are used for control, when the force error is very small, the calculated proportional terms may not be sufficient to overcome static friction, causing the collaborative robot to stop fine-tuning and eventually stabilize at a force slightly lower than the target passivation force, resulting in steady-state error. Therefore, integral terms are introduced to eliminate steady-state error.

[0093] As long as steady-state error exists, the cumulative... As time goes on, the integral term will continue to increase until the accumulated "push" is enough to overcome friction and completely eliminate the steady-state error.

[0094] The formula for calculating the integral term is:

[0095] in, The integral gain determines the speed at which steady-state errors are eliminated. An excessively large integral gain can lead to integral saturation and overshoot.

[0096] The integral term ensures the accuracy of the control command, guaranteeing that the final contact force can be precisely stabilized at the target value.

[0097] The differential term focuses on the rate of change of the error, serving both predictive and damping functions. When the collaborative robot rapidly approaches the target passivation force, the force error decreases quickly. The differential term detects a "rapidly changing" trend and generates a "brake" command in the opposite direction of movement to prevent exceeding the target passivation force. Conversely, if the force rapidly deviates from the target, the differential term generates a rapid corrective force.

[0098] The formula for calculating the differential term is:

[0099] in, The differential gain determines the degree of damping. The differential gain of the scheme can suppress oscillations, making the response smoother and more stable. However, an excessively large differential gain can lead to excessive sensitivity to noise.

[0100] The differential term ensures the stability of the control while preventing oscillations around the target value.

[0101] It should be noted that, as shown in Figure 4, the constant force passivation control process adopts a feedback control loop, and the input is the preset target passivation force. The target force and the estimated contact force from the feedback loop A comparison is made at a summation point, resulting in a force error. Force error The error is fed into a constant force controller, such as a PID controller. The controller calculates and outputs compensation motion commands based on the magnitude and trend of the error. This compensation command represents the fine-tuning speed or displacement required in the direction of the brush normal.

[0102] In addition, the collaborative robot also has a basic motion command along the cutting edge of the tool. These two commands are synthesized into the final joint drive commands in the collaborative robot's kinematics / dynamics controller and executed by the collaborative robot.

[0103] The motion of the collaborative robot acts on the contact process between the cutter and the brush wheel, generating an actual contact force. However, this process is affected by various disturbances D, such as the flexible deformation of the collaborative robot itself, brush wear, and initial positioning errors. These disturbances are the root cause of the actual force deviating from the target force.

[0104] Not direct measurement Instead, it reads the joint torques inside the collaborative robot. and joint position The signal is input into a joint torque observer. This observer, based on the robot's dynamic model, can accurately estimate the contact force at the end effector and output the estimated contact force. .

[0105] Ultimately, this estimated force As a closed-loop feedback signal, it is sent back to the initial summation point and compared with the target force, forming a continuous, automatically correcting closed loop. Through this closed loop, it can actively resist various disturbances, ensuring that the actual contact force always closely follows the target passivation force, thereby achieving high-precision constant force passivation.

[0106] By superimposing the main motion and the compensation motion, the limitations of single trajectory control are avoided, and the main motion and the compensation motion are coordinated to achieve complete blunting of the tool cutting edge.

[0107] Step S307: When the blunting of the completed blade trajectory is detected, control the collaborative robot to retreat to a safe position.

[0108] For details, please refer to step S207 of the embodiment shown in Figure 2, which will not be repeated here.

[0109] The visual and force-sensing fusion collaborative robot tool adaptive passivation method provided in this embodiment has the following beneficial effects: (1) Quality and consistency: It realizes a qualitative change from "position repetition" to "effect reproduction". Through constant force control, the present invention ensures that the cutting edge of each tool is subjected to the same and ideal force, fundamentally eliminating the influence of all physical variables such as robot deformation, brush wear, and clamping error on the processing effect, ensuring the high consistency and predictability of the processing effect, and realizing the leap from simple repetitive actions to repetitive high-quality results; (2) Intelligence and autonomy: It realizes full adaptation to the dynamic environment and greatly reduces human intervention. By fusing information from multiple sensors, the system can autonomously perceive and compensate for the wear of the brush and the individual differences of the workpiece. This means that the production line can operate autonomously and stably for a longer period of time without the need for frequent manual trajectory calibration or re-teaching, which greatly improves the production efficiency (OEE) and automation level; (3) Economy and reliability: While achieving high performance, it significantly reduces costs and improves system robustness. The core "sensorless force control" technology completely eliminates the need for expensive and easily damaged external six-dimensional force / torque sensors. This not only directly reduces hardware costs and system integration complexity, but more importantly, it removes a critical point of failure, making the system more reliable and durable in industrial environments with lower maintenance costs; (4) Flexibility and versatility: It enables rapid adaptation to different processes and workpieces. Due to the adoption of universal collaborative robots and adaptive intelligent algorithms, the production line can quickly switch to process different types of cutting tools or adjust different passivation process parameters (such as passivation force and angle). Operators only need to modify the parameters in the software without replacing expensive special mechanical fixtures, which greatly improves the flexibility of the production line and perfectly adapts to the needs of modern manufacturing for small batches and multiple varieties.

[0110] This embodiment also provides a vision-force fusion collaborative robot tool adaptive blunting device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0111] This embodiment provides a collaborative robot tool adaptive blunting device that integrates vision and force perception, as shown in Figure 5. It includes an extraction module 501, which is used to acquire the image and scanning results of the tool when the collaborative robot is detected to have grasped the tool, and to extract the three-dimensional feature information of the tool.

[0112] The comparison module 502 is used to obtain the current contour data of the brush and compare the current contour data with the initial contour data of the brush to calculate the wear amount.

[0113] The determination module 503 is used to fuse the three-dimensional feature information and wear amount of the tool to determine the target attitude and target position.

[0114] The estimation module 504 is used to run the joint torque observer and use the joint torque observer to estimate the current contact force between the collaborative robot end effector and the wheel brush.

[0115] The main motion module 505 is used to control the collaborative robot to move towards the target position in the target posture. When the current contact force reaches the initial contact force, it controls the collaborative robot to perform main motion along the blade line in a constant force control mode.

[0116] The motion planning module 506 is used to generate dynamic compensation motion commands using a PID algorithm if there is a force error between the current contact force and the target passivation force, and to plan the collaborative robot motion by combining the main motion and the compensation motion.

[0117] The retraction module 507 is used to control the collaborative robot to retract to a safe position when the blunting of the completed blade trajectory is detected.

[0118] In some optional implementations, the extraction module 501 includes: a first extraction unit, used to scan the tool with a one-dimensional sensor when the tool reaches the detection range of the one-dimensional sensor, and measure the effective length of the tool extending from the end of the gripper; and a second extraction unit, used to capture the tool image with a vision camera and use an image processing algorithm to identify the spatial vector of the cutting edge to be blunted, wherein the spatial vector is used to characterize the path and position of the cutting edge.

[0119] In some optional implementations, the determining module 503 includes: a first determining unit, used to calculate the end-effector posture based on the spatial vector of the blade to be blunted, the end-effector posture being used to characterize the posture of the collaborative robot when the blade to be blunted is parallel to the rotation axis of the brush; a second determining unit, used to determine an additional posture based on preset process parameters, the additional posture being used to characterize the tilt angle when the tool axis is tilted to the horizontal plane at the same angle as the process parameters; and a third determining unit, used to superimpose the end-effector posture and the additional posture to obtain the target posture.

[0120] In some optional implementations, the estimation module 504 includes: an equation establishment unit for establishing the dynamic equations of the collaborative robot; a first calculation unit for acquiring the real-time position and real-time velocity of the collaborative robot joints, calculating the real-time acceleration based on the real-time velocity, and substituting the real-time position, real-time velocity, and real-time acceleration into the dynamic equations to calculate the theoretical torque; a second calculation unit for determining the difference between the theoretical torque and the actual output torque as the residual torque; and a third calculation unit for estimating the contact force corresponding to the residual torque using the pseudo-inverse of the Jacobian matrix transpose, and determining the estimated contact force as the current contact force between the end effector of the collaborative robot and the wheel brush.

[0121] In some alternative implementations, motion planning module 506 includes a proportional term determination unit for determining the product of proportional gain and force error as a proportional term.

[0122] The integral term determination unit is used to determine the product of the integral gain and the integral result of the force error as the integral term.

[0123] The differential term determination unit is used to determine the product of the differential gain and the rate of change of the force error as the differential term.

[0124] The accumulation unit is used to determine the accumulated result of the proportional, integral, and differential terms as the compensation motion command.

[0125] In some optional implementations, the motion planning module 506 further includes a direction application unit, used to apply the velocity value corresponding to the compensation motion command to a normal direction perpendicular to the brush surface and pointing towards the center of the brush.

[0126] The motion planning unit is used to superimpose the velocity vector corresponding to the main motion and the velocity vector corresponding to the compensated motion command after the direction is applied to obtain the final command, and drive the collaborative robot's motion planning according to the final command.

[0127] The vision-force fusion collaborative robot tool adaptive passivation device provided in this embodiment of the invention can execute the vision-force fusion collaborative robot tool adaptive passivation method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0128] Figure 6 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0129] Referring specifically to FIG6, a schematic diagram of a suitable electronic device for implementing embodiments of the present invention is shown below. The electronic device may include a processor (e.g., a central processing unit, a graphics processor, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from memory 608 into random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the electronic device. The processor 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0130] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows the electronic device to communicate wirelessly or wiredly with other devices to exchange data. Although FIG6 shows an electronic device with various devices, it should be understood that it is not required to implement or have all the devices shown, and more or fewer devices may be implemented or have instead.

[0131] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a memory 608, or installed from a ROM 602. When the computer program is executed by the processor 601, it performs the functions defined in the vision and force fusion collaborative robot tool adaptive passivation method of the embodiments of the present invention.

[0132] The electronic device shown in Figure 6 is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0133] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the vision and force fusion collaborative robot tool adaptive passivation method shown in the above embodiments is implemented.

[0134] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0135] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and all such modifications and variations fall within the scope defined by the appended invention.

Claims

1. A method for adaptive blunting of cutting tools in collaborative robots that integrates vision and force perception, characterized in that, An adaptive passivation system for collaborative robots using vision and force fusion is disclosed. The system includes a collaborative robot, a gripper, a cutting tool, and a brush wheel. The joints of the collaborative robot have torque sensing capabilities. The gripper is mounted on the end effector of the collaborative robot. The method includes: when the collaborative robot detects that it has grasped the cutting tool, acquiring an image and scan result of the cutting tool and extracting its three-dimensional feature information; acquiring the current contour data of the brush wheel and comparing it with the initial contour data of the brush wheel to calculate the wear amount; fusing the three-dimensional feature information of the cutting tool and the wear amount to determine the target posture and target position; running a joint torque observer and using it to estimate the current contact force between the end effector of the collaborative robot and the brush wheel; controlling the collaborative robot to move towards the target position in the target posture; when the current contact force reaches the initial contact force, controlling the collaborative robot to perform a main motion along the cutting edge direction in a constant force control mode; if there is a force error between the current contact force and the target passivation force, generating a dynamic compensation motion command using a PID algorithm, and combining the main motion and the compensation motion to plan the collaborative robot's motion; when passivation of the completed cutting edge trajectory is detected, controlling the collaborative robot to retreat to a safe position.

2. The method according to claim 1, characterized in that, The system also includes a one-dimensional sensor and a vision camera. The three-dimensional feature information includes the spatial vector of the cutting edge to be blunted and the effective length of the tool extending from the end of the gripper. Acquiring the image and scanning results of the tool and extracting the three-dimensional feature information of the tool includes: when the tool enters the detection range of the one-dimensional sensor, scanning the tool with the one-dimensional sensor and measuring the effective length of the tool extending from the end of the gripper; capturing the image of the tool with the vision camera and using an image processing algorithm to identify the spatial vector of the cutting edge to be blunted. The spatial vector is used to characterize the path and position of the cutting edge.

3. The method according to claim 2, characterized in that, The determination of the target posture includes: calculating the end-effector posture based on the spatial vector of the blade line to be blunted, wherein the end-effector posture is used to characterize the posture of the collaborative robot when the blade line to be blunted is parallel to the rotation axis of the brush; determining an additional posture based on preset process parameters, wherein the additional posture is used to characterize the tilt angle that makes the tool axis tilt to the horizontal plane at the same angle as the process parameters; and superimposing the end-effector posture and the additional posture to obtain the target posture.

4. The method according to claim 1, characterized in that, The step of estimating the current contact force between the end effector of the collaborative robot and the wheel brush using the joint torque observer includes: establishing the dynamic equations of the collaborative robot; obtaining the real-time position and real-time velocity of the collaborative robot joints, calculating the real-time acceleration based on the real-time velocity, substituting the real-time position, real-time velocity, and real-time acceleration into the dynamic equations to calculate the theoretical torque; determining the difference between the theoretical torque and the actual output torque as the residual torque; estimating the contact force corresponding to the residual torque using the pseudo-inverse of the Jacobian matrix transpose, and determining the estimated contact force as the current contact force between the end effector of the collaborative robot and the wheel brush.

5. The method according to claim 4, characterized in that, Calculate the theoretical torque using the following formula: in, For theoretical torque, The inertia matrix, Let be the acceleration vectors of each joint of the collaborative robot. For the Coriolis force and centrifugal force terms, This represents the velocity vectors of each joint of the collaborative robot. For the gravitational torque term, The torque represents the joint friction force; the contact force is calculated using the following formula: in, For contact force, For residual torque, It is the pseudo-inverse of the transpose of the Jacobian matrix.

6. The method according to claim 1, characterized in that, The method of generating dynamic compensation motion commands using the PID algorithm includes: determining the product of proportional gain and force error as the proportional term; determining the product of integral gain and integral force error as the integral term; determining the product of differential gain and rate of change of force error as the differential term; and determining the sum of the proportional term, integral term, and differential term as the compensation motion command.

7. The method according to claim 1, characterized in that, The method of combining main motion and compensation motion planning for collaborative robot motion includes: applying the velocity value corresponding to the compensation motion command to the normal direction perpendicular to the brush surface and pointing towards the center of the brush; superimposing the velocity vector corresponding to the main motion and the velocity vector corresponding to the compensation motion command after applying the direction to obtain the final command, and driving the collaborative robot motion planning according to the final command.

8. A collaborative robot tool adaptive blunting device integrating vision and force perception, characterized in that, An adaptive passivation system for collaborative robots using vision and force fusion is disclosed. The system includes a collaborative robot, a gripper, a cutting tool, and a brush wheel. The joints of the collaborative robot have torque sensing capabilities. The gripper is mounted at the end effector of the collaborative robot. The device includes: an extraction module for acquiring an image and scan result of the cutting tool when the collaborative robot detects that it has grasped the tool, and extracting the tool's three-dimensional feature information; a comparison module for acquiring the current contour data of the brush wheel, comparing the current contour data with the initial contour data of the brush wheel, and calculating the wear amount; and a determination module for fusing the three-dimensional feature information of the cutting tool and the wear amount to determine the target posture and target position; and an estimation module. The system comprises the following modules: a calculation module for running a joint torque observer and estimating the current contact force between the collaborative robot's end effector and the wheel brush; a main motion module for controlling the collaborative robot to move towards the target position in the target posture, and controlling the collaborative robot to perform main motion along the blade direction in a constant force control mode when the current contact force reaches the initial contact force; a motion planning module for generating dynamic compensation motion commands using a PID algorithm if there is a force error between the current contact force and the target passivation force, and combining the main motion and compensation motion to plan the collaborative robot's motion; and a retreat module for controlling the collaborative robot to retreat to a safe position when passivation of the completed blade trajectory is detected.

9. An electronic device, characterized in that, include: A memory and a processor are interconnected, the memory storing computer instructions, and the processor executing the computer instructions to perform the adaptive passivation method for collaborative robot tools based on vision and force fusion as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the vision and force fusion collaborative robot tool adaptive passivation method according to any one of claims 1 to 7.

Images