A high-precision automatic teaching device and teaching method for the full working space of a robot

Abstract

The present invention discloses a high-precision automatic teaching device for the entire workspace of a robot in the field of machine vision technology, comprising: a robot for realizing full-workspace movements; a quick-change disk installed at the end of the industrial robot for installing a binocular camera, a laser displacement component, and a customized device, wherein the binocular camera is fixedly installed on one side of the quick-change disk, and the laser displacement component and the customized device are replaceably installed on the bottom of the quick-change disk; a customized mold installed below the quick-change disk for reference teaching, wherein a plurality of features are provided on the customized mold, and the feature center is used as a reference point or teaching point. The present invention completes robot 3D vision grasping, assembly and other tasks with an overall accuracy requirement of more than 0.2 mm through three stages: system calibration, automatic teaching and online alignment.

Description

Technical Field

[0001] The present invention relates to the technical field of machine vision, and in particular to a robot automatic teaching device and a teaching method. Background Art

[0002] The working ability of industrial robots is mainly determined by their positioning accuracy and the quality of the products they produce. Accuracy refers to absolute positioning accuracy and repeat positioning accuracy. The current repeat positioning accuracy can reach up to 0.01mm, but because the robot is affected by processing conditions, the absolute positioning accuracy of the robot is relatively low. Generally speaking, robot errors are divided into geometric errors and non-geometric errors. Here, geometric errors include rod parameter errors, errors between the theoretical reference coordinate system and the actual reference coordinate system, non-parallelism of the joint axis, zero position deviation, etc. Non-geometric factors include elastic deformation of joints and links, gear clearance, gear transmission error, thermal deformation, etc. The superposition of various error factors makes it difficult and costly to improve the absolute accuracy of the robot from the model, which has an adverse effect on the application of robots in fields with high precision requirements such as aerospace [1].

[0003] Therefore, compensating for robot errors and improving the absolute accuracy of the robot are urgent issues that need to be solved in the current field of robotics technology. The present invention proposes a high-precision automatic teaching device and method for the entire workspace of the robot, which improves the overall accuracy of the robot's 3D visual grasping and assembly system to above 0.2mm.

[0004] [1] Wang Wenbo. Research on the application of robot absolute positioning error compensation in aircraft canopy hole making[D]. Shenyang University of Technology, 2023. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides a high-precision automatic teaching device and teaching method for the entire workspace of a robot. Through three stages of system calibration, automatic teaching, and online alignment, it ultimately completes robot 3D visual grasping, assembly and other tasks with an overall accuracy requirement of more than 0.2mm.

[0006] The object of the present invention is achieved as follows: a high-precision automatic teaching device for the entire workspace of a robot, characterized by comprising:

[0007] Robots, used to achieve full workspace motion;

[0008] A quick-change plate is installed at the end of the industrial robot to install a binocular camera, a laser displacement component, and a customized device. The binocular camera is fixedly installed on one side of the quick-change plate, and the laser displacement component and the customized device are replaceably installed at the bottom of the quick-change plate.

[0009] Custom molds are installed below the quick-change plate for reference and teaching. Several features are set on them, with the center of the feature as the reference point or teaching point.

[0010] As an optimal technical solution for the high-precision automatic teaching device for the entire workspace of a robot described in the present invention, the laser displacement component includes a laser displacement sensor and a monocular camera; the customized equipment includes a customized gripper and a customized suction cup.

[0011] A method for high-precision automatic teaching of a robot's entire workspace, using the above-mentioned high-precision automatic teaching device for the robot's entire workspace, includes three stages: system calibration, automatic teaching, and online alignment:

[0012] System calibration phase:

[0013] Step S1) calibrating the parameters of the monocular camera and the binocular camera, calibrating the hand-eye relationship, and calibrating the height difference between the laser displacement sensor and the monocular camera;

[0014] Step S2) teaching the photographing pose and alignment pose of the reference point;

[0015] Automatic teaching stage:

[0016] Step S3) Locate the three reference points, calculate the relative pose of the robot flange and the teaching point when the binocular camera photographs the teaching point, then calculate the robot pose of the teaching point on the mold photographed by the binocular camera, and calculate the relative pose of the flange and the teaching point when the laser spot is aligned with the teaching point;

[0017] Step S4) The robot drives the binocular camera to precisely locate the teaching point and calculates the robot posture when the laser spot is aligned with the teaching point;

[0018] Step S5) guiding the robot to move according to the calculated alignment posture so that the laser displacement sensor spot is aligned with the teaching point, the monocular camera takes a picture of the spot, and calculates the position deviation between the spot and the teaching point;

[0019] Step S6) guiding the robot to deflect so that the light spot is accurately aligned with the teaching point, and recording the posture of the teaching point and the relative posture of the teaching point and the flange during alignment;

[0020] Online alignment stage:

[0021] Step S7) During operation, the customized device is first replaced by the quick-change disk, and then the binocular camera accurately recognizes the target's posture, selects the nearest teaching relative posture, and calculates the precise posture of the customized device aligned with the target;

[0022] Step S8) Guide the robot to move according to the calculated precise posture to complete the precise task.

[0023] As a preferred technical solution of the high-precision automatic teaching method for the entire workspace of a robot described in the present invention, step S1 is specifically as follows:

[0024] First, calibrate the intrinsic parameters of the left and right cameras in the binocular camera, calibrate the relative pose of the left and right cameras, and the relative pose of the left camera and the robot flange; then calibrate the intrinsic parameters of the monocular camera and the hand-eye relationship; operate the robot so that the camera optical axis is perpendicular to the calibration plate plane, take a picture of the calibration plate, and estimate the distance between the camera imaging plane and the calibration plate plane based on the image , while collecting the laser displacement sensor value , find the difference between the two, that is, the distance between the center of the camera imaging plane and the center of the laser in the direction of the camera optical axis .

[0025] As a preferred technical solution of the high-precision automatic teaching method for the entire workspace of a robot described in the present invention, step S2 is specifically as follows:

[0026] Operate the robot to drive the binocular camera to take pictures of three reference points on the mold ,in, Represents the rotation of the reference point coordinate system in the robot coordinate system, Represents the translation of the reference point coordinate system in the robot coordinate system; when taking pictures, ensure that the ring at the reference point is clearly imaged in the left and right cameras, and the camera optical axis is perpendicular to the plane of the ring. Record the robot's posture at this time, and you will get the photographic posture of the reference point. ; Operate the robot so that the center of the laser spot is aligned with the first reference point , ensure that the optical axis of the monocular camera is perpendicular to the plane of the ring where the reference point is located, and the ring is clearly imaged in the monocular camera, and record the laser reading at this time , and record the robot posture at this time, that is, the alignment posture of the reference point is obtained .

[0027] As a preferred technical solution of the high-precision automatic teaching method for the entire workspace of a robot described in the present invention, step S3 is specifically as follows:

[0028] The robot takes three pictures according to the teaching , guide the binocular camera to take pictures of the three reference points and locate them, and get the precise positions of the three reference points , combined with the three-dimensional model information of the mold, calculate the projection of the reference point and the teaching point on the model in the robot coordinate system ; The total number of reference points plus teaching points; the exact position of the three reference points Alternative , get the precise pose of the three reference points ; Calculate the relative position of the flange and the teaching point when the binocular camera takes a picture of the teaching point , and then calculate the projection of the teaching point taken by the binocular camera Robot pose , calculate the relative position of the robot flange and the teaching point when the laser spot is aligned with the teaching point .

[0029] As a preferred technical solution of the high-precision automatic teaching method for the entire workspace of a robot described in the present invention, step S4 is specifically as follows:

[0030] The robot drives the binocular camera to precisely locate the teaching points and obtain the precise position of all teaching points , which replaces the translation part of the teaching point projection , get the precise pose of the teaching point ; Calculate the robot pose when the laser spot is aligned with the teaching point .

[0031] As a preferred technical solution of the high-precision automatic teaching method for the entire workspace of a robot described in the present invention, step S5 is specifically as follows:

[0032] According to the calculated alignment pose Guide the robot to move so that the laser displacement sensor spot is aligned with the teaching point Due to the influence of factors such as the absolute accuracy of the robot and the measurement error of the camera, there will be a position error when the light spot is aligned with the teaching point. At this time, the monocular camera takes a picture of the light spot and reads the reading of the laser displacement sensor. , depth information of the monocular camera , locate the image center of the light spot, and calculate the position deviation between the light spot and the teaching point in the robot coordinate system based on the camera internal parameters .

[0033] As a preferred technical solution of the high-precision automatic teaching method for the entire workspace of a robot described in the present invention, step S6 is specifically as follows:

[0034] In the coarse alignment pose Based on the offset, the precise alignment pose is obtained , guide the robot to deviate so that the light spot is accurately aligned with the teaching point and record the precise position of the teaching point The relative position of the flange and the teaching point during precise alignment .

[0035] As a preferred technical solution of the high-precision automatic teaching method for the entire workspace of a robot described in the present invention, step S7 is specifically as follows:

[0036] According to different application scenarios, the robot can replace the end tool through the quick change plate, including the suction cup and the tool with rigid connection of the clamp. The center of the tool is on the laser beam and the distance from the laser center is ; Then the binocular camera accurately recognizes the target's position and obtains the target's precise position , calculate the target point With all teaching points distance, select the nearest teaching point , find the corresponding relative position of the flange and the teaching point when accurately aligning , that is, get the nearest teaching relative pose and calculate the precise pose of the tool to the target .

[0037] Compared with the existing technology, the beneficial effect of the present invention is that the present invention realizes non-contact high-precision automatic teaching of the entire workspace of the robot by flexibly combining robot technology, 3D vision technology, monocular vision technology and laser ranging technology, and greatly improves the final alignment accuracy of the entire workspace of the robot 3D vision system. Before this, the final alignment accuracy of the robot's 3D vision was generally lower than 0.5mm, and it could only guarantee the alignment accuracy of a small range of workspace. In addition, the teaching mainly relied on human eye judgment, and the teaching accuracy was inconsistent and the efficiency was low. The device and method can be applied to different industrial scenarios. By replacing the end tool, tasks such as alignment, grasping, and assembly with an accuracy of 0.2mm in the entire workspace of the robot can be achieved, and the intensity of the teaching can be changed according to actual needs, shortening the teaching cycle while ensuring high precision. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0039] Figure 1 Schematic diagram of the robot's full workspace automatic teaching device in the present invention.

[0040] Figure 2 This is a flow chart of the robot's full workspace automatic teaching method in the present invention.

[0041] Among them, there are 100 robots, 200 conversion disks, 300 binocular cameras, 400 laser displacement sensors, 500 monocular cameras, 600 customized equipment, 700 customized molds, 701 reference points, and 702 teaching points. DETAILED DESCRIPTION

[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0043] Example 1

[0044] like Figure 1 The device shows a high-precision automatic teaching device for the entire workspace of a robot, including an industrial robot 100, a quick-change disk 200, a laser displacement sensor 400, a binocular camera 300, a monocular camera 500, a light source, a customized mold 700, etc. The industrial robot 100 is a 6-axis robot 100, and the laser beam is parallel to the optical axis of the monocular camera 500; the binocular camera 300 (or other 3D vision system) is rigidly mounted at the end of the robot 100 to locate the center of the ring on the mold; the manipulator side of the quick-change disk 200 is mounted on the flange at the end of the robot 100, and customized equipment 600 is mounted on multiple fixture sides of the quick-change disk 200, including customized grippers, suction cups or other tools (after switching through the quick-change disk 200, the end of the tool is on the laser beam).

[0045] The custom mold 700 has several rings (or other features) with a diameter of 2 mm, with the centers of the rings serving as reference points 701 or teaching points 702 .

[0046] Example 2

[0047] like Figure 2 A high-precision automatic teaching method for the entire workspace of a robot 100 is shown, using the teaching device of Example 1, and includes three stages: system calibration, automatic teaching, and online alignment:

[0048] System calibration phase:

[0049] Step S1: calibrate the parameters of the monocular and binocular cameras 300, calibrate the hand-eye relationship, and calibrate the height difference between the laser and the monocular camera 500;

[0050] Step S2: teaching the photographing pose and alignment pose of the reference point 701;

[0051] Automatic teaching stage:

[0052] Step S3: Locate the three reference points 701, calculate the relative pose of the flange and the teaching point 702 when the binocular camera 300 photographs the teaching point 702, then calculate the pose of the robot 100 when the binocular camera 300 photographs the teaching point 702 on the mold, and calculate the relative pose of the flange and the teaching point 702 when the laser spot is aligned with the teaching point 702;

[0053] Step S4: The robot 100 drives the binocular camera 300 to precisely locate the teaching point 702 and calculate the posture of the robot 100 when the laser spot is aligned with the teaching point 702;

[0054] Step S5: guiding the robot 100 to move according to the calculated alignment posture so that the light spot of the laser displacement sensor 400 is aligned with the teaching point 702, the monocular camera 500 takes a picture of the light spot, and calculates the position deviation between the light spot and the teaching point 702;

[0055] Step S6: guiding the robot 100 to deflect so that the light spot is accurately aligned with the teaching point 702, and recording the posture of the teaching point 702 and the relative posture of the teaching point 702 and the flange during alignment;

[0056] Online alignment stage:

[0057] Step S7: When working, first use the quick-change tray 200 to replace the tool (suction cup, gripper, etc.), then the binocular camera 300 accurately recognizes the target's posture, selects the nearest teaching relative posture to calculate the precise posture of the tool alignment target;

[0058] Step S8: guiding the robot 100 to move according to the calculated precise posture to complete the precise alignment (grasping, assembly, etc.) task.

[0059] Step S1 is specifically as follows:

[0060] First, calibrate the intrinsic parameters of the left and right cameras in the binocular camera 300, calibrate the relative pose of the left and right cameras, and the relative pose of the left camera and the robot 100 flange, and obtain the conversion relationship between the image coordinate system in the binocular camera 300 and the robot 100 coordinate system; then calibrate the intrinsic parameters of the monocular camera 500 and the hand-eye relationship; operate the robot 100 so that the camera optical axis is perpendicular to the calibration plate plane, take a picture of the calibration plate, and estimate the distance between the camera imaging plane and the calibration plate plane based on the image , while collecting 400 values of laser displacement sensor , find the difference between the two, that is, the distance between the center of the camera imaging plane and the center of the laser in the direction of the camera optical axis .

[0061] It should be noted that during alignment, the distance between the center of the light spot and the camera imaging plane can be calculated by reading the laser reading, and finally the image coordinates measured by the monocular camera 500 are converted to the coordinate system of the robot 100.

[0062] Step S2 is specifically as follows:

[0063] The robot 100 drives the binocular camera 300 to take pictures of three reference points 701 on the mold. ,in represents the rotation of the reference point 701 coordinate system in the robot 100 coordinate system, Represents the translation of the reference point 701 coordinate system in the robot 100 coordinate system; when taking pictures, ensure that the ring at the reference point 701 is clearly imaged in the left and right cameras, and the camera optical axis is perpendicular to the plane of the ring. Record the posture of the robot 100 at this time, and you will get the photographic posture of the reference point 701 ; Operate the robot 100 so that the center of the laser spot is aligned with the first reference point 701 , ensure that the optical axis of the monocular camera 500 is perpendicular to the plane of the ring where the reference point 701 is located, and the ring is clearly imaged in the monocular camera 500, and record the laser reading at this time , and record the robot 100 posture at this time, that is, the alignment posture of the reference point 701 is obtained .

[0064] It should be noted that this step provides a basis for calculating the relative posture of the flange and the teaching point 702 when the laser spot is aligned with the teaching point 702 in step S3.

[0065] Step S3 is specifically as follows:

[0066] The robot 100 takes three pictures according to the teaching , guide the binocular camera 300 to take pictures of the three reference points 701 and locate them, and obtain the precise positions of the three reference points 701 , combined with the three-dimensional model information of the mold, calculate the projection of the reference point 701 and the teaching point 702 on the model in the robot 100 coordinate system , The total number of reference points 701 plus teaching points 702; the exact position of the three reference points 701 Alternative , get the precise pose of the three reference points 701 ; Calculate the relative position of the flange and the teaching point 702 when the binocular camera 300 takes a picture of the teaching point 702 , then calculate the projection of the teaching point 702 taken by the binocular camera 300 100 robot poses , calculate the relative position of the flange and the teaching point 702 when the laser spot is aligned with the teaching point 702 .

[0067] It should be noted that After that, the robot 100 posture when taking photos of all the teaching points 702 can be obtained after positioning the three reference points 701 on the positioning model, which is used to accurately position the teaching point 702. After accurately positioning the teaching point 702 in step S4, the robot 100 posture when taking photos of all the teaching points 702 can be obtained. Calculate the pose of the robot 100 when the laser spot is aligned with the teaching point 702.

[0068] Step S4 is specifically as follows:

[0069] The robot 100 drives the binocular camera 300 to precisely locate the teaching point 702, and obtains the precise position of all the teaching points 702. , which replaces the translation part of the projection of the teaching point 702 , get the precise pose of teaching point 702 ; Calculate the robot 100's pose when the laser spot is aligned with the teaching point 702 .

[0070] It should be noted that the posture of the robot 100 during alignment is used to project the light spot around the teaching point 702. At this time, due to factors such as the low absolute accuracy of the robot 100, the center of the light spot cannot be accurately aligned with the center of the teaching point 702, and the error is subsequently measured.

[0071] Step S5 is specifically as follows:

[0072] According to the calculated alignment pose Guide the robot 100 to move so that the laser displacement sensor 400 spot is aligned with the teaching point 702 Due to factors such as the absolute accuracy of the robot 100 and the camera measurement error, a position error will occur when the light spot is aligned with the teaching point 702. At this time, the monocular camera 500 takes a picture of the light spot and reads the reading of the laser displacement sensor 400. , depth information of monocular camera 500 , locate the image center of the light spot, and calculate the position deviation between the light spot and the teaching point 702 in the robot 100 coordinate system during alignment by combining the camera internal parameters , that is, the overall error of the system aligning with the target point during the teaching phase is obtained.

[0073] Step S6 is specifically as follows:

[0074] In the coarse alignment pose Based on the offset, the precise alignment pose is obtained , guide the robot 100 to deviate so that the light spot is accurately aligned with the teaching point 702, and record the precise position and posture of the teaching point 702 The relative position of the flange and the teaching point 702 during precise alignment .

[0075] It should be noted that a database with one-to-one correspondence between spatial points and precise alignment matrices is established. After the spatial points are precisely positioned in the online stage, the nearest taught spatial points in the database are found, and the final alignment posture of the robot 100 is calculated according to the corresponding precise alignment matrix to achieve high-precision alignment of the target.

[0076] Step S7 is specifically as follows:

[0077] According to different application scenarios, the robot 100 can replace the end tool through the quick change plate 200, including rigidly connected tools such as suction cups and grippers. The center of the tool is on the laser beam and the distance from the laser center is ; Then the binocular camera 300 accurately recognizes the target position and obtains the accurate position of the target point , calculate the target point With all teaching points 702 distance, select the nearest teaching point 702 , find the corresponding relative position of the flange and the teaching point 702 when accurately aligning , that is, get the nearest teaching relative pose and calculate the precise pose of the tool to the target After establishing a database of correspondence between spatial points and precise alignment matrices in the automatic teaching phase, high-precision alignment of the target can be achieved by searching the database data in the online phase.

[0078] It should be noted that the general industrial robot 100 3D vision system is limited by factors such as the absolute accuracy of the robot 100 and can only achieve a final accuracy of up to 0.5mm. The present invention can stably achieve a maximum error of less than 0.2mm between the tool coordinate system of the robot 100 end during alignment and the target point in space, and can be widely used in the field of high-precision machining.

[0079] The above embodiments are only intended to help understand the method and core concept of the present invention. It should be noted that, without departing from the principles of the present invention, a number of improvements and modifications may be made to the present invention by those skilled in the art, and such improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A method for high-precision automatic teaching of a robot's entire workspace, using a high-precision automatic teaching device for the robot's entire workspace, comprising: Robots, used to achieve full workspace motion; A quick-change tray is installed at the end of the industrial robot and is used to install a binocular camera, a laser displacement assembly, and customized equipment. The binocular camera is fixedly installed on one side of the quick-change tray, and the laser displacement assembly and customized equipment are replaceably installed at the bottom of the quick-change tray. The laser displacement assembly includes a laser displacement sensor and a monocular camera; the customized equipment includes a customized gripper and a customized suction cup. The custom mold is installed below the quick-change plate for reference teaching. It has several features on it, with the feature center as the reference point or teaching point. It is characterized by three stages: system calibration, automatic teaching, and online alignment: System calibration phase: Step S1) calibrating the parameters of the monocular camera and the binocular camera, calibrating the hand-eye relationship, and calibrating the height difference between the laser displacement sensor and the monocular camera; Step S2) teaching the photographing pose and alignment pose of the reference point; Automatic teaching stage: Step S3) Locate three reference points, calculate the relative pose of the robot flange and the teaching point when the binocular camera photographs the teaching point, then calculate the robot pose of the teaching point on the mold photographed by the binocular camera, and calculate the relative pose of the flange and the teaching point when the laser displacement sensor spot is aligned with the teaching point; Step S4) The robot drives the binocular camera to precisely locate the teaching point and calculates the robot posture when the laser displacement sensor spot is aligned with the teaching point; Step S5) guiding the robot to move according to the calculated alignment posture so that the laser displacement sensor spot is aligned with the teaching point, the monocular camera takes a picture of the spot, and calculates the position deviation between the spot and the teaching point; Step S6) guiding the robot to deflect so that the light spot is accurately aligned with the teaching point, and recording the posture of the teaching point and the relative posture of the teaching point and the flange during alignment; Online alignment stage: Step S7) During operation, the customized device is first replaced by the quick-change disk, and then the binocular camera accurately recognizes the target's posture, selects the nearest teaching relative posture, and calculates the precise posture of the customized device aligned with the target; Step S8) Guide the robot to move according to the calculated precise posture to complete the precise task.

2. A high-precision automatic teaching method for the entire workspace of a robot according to claim 1, characterized in that: Step S1 is specifically as follows: First, calibrate the intrinsic parameters of the left and right cameras in the binocular camera, calibrate the relative pose of the left and right cameras, and the relative pose of the left camera and the robot flange; then calibrate the intrinsic parameters of the monocular camera and the hand-eye relationship; operate the robot so that the optical axis of the monocular camera is perpendicular to the plane of the calibration plate, take a picture of the calibration plate, and estimate the distance between the camera imaging plane and the calibration plate plane based on the image , while collecting the laser displacement sensor value , calculate the difference between the two, that is, the distance between the center of the camera imaging plane and the center of the laser displacement sensor in the direction of the camera optical axis .

3. A high-precision automatic teaching method for the entire workspace of a robot according to claim 2, characterized in that: Step S2 is specifically as follows: Operate the robot to drive the binocular camera to take pictures of three reference points on the mold ,in, Represents the rotation of the reference point coordinate system in the robot coordinate system, Represents the translation of the reference point coordinate system in the robot coordinate system; when taking pictures, ensure that the ring at the reference point is clearly imaged in the left and right cameras, and the camera optical axis is perpendicular to the plane of the ring. Record the robot's posture at this time, and you will get the photographic posture of the reference point. ; Operate the robot so that the center of the laser displacement sensor spot is aligned with the first reference point , ensure that the optical axis of the monocular camera is perpendicular to the plane of the ring where the reference point is located, and the ring is clearly imaged in the monocular camera, and record the reading of the laser displacement sensor at this time , and record the robot posture at this time, that is, the alignment posture of the reference point is obtained .

4. A high-precision automatic teaching method for the entire workspace of a robot according to claim 3, characterized in that: Step S3 is specifically as follows: The robot takes three photos according to the teaching , guide the binocular camera to take pictures of the three reference points and locate them, and get the precise positions of the three reference points , combined with the three-dimensional model information of the mold, calculate the projection of the reference point and the teaching point on the model in the robot coordinate system ; The total number of reference points plus teaching points; the exact position of the three reference points Alternative , get the precise pose of the three reference points ; Calculate the relative position of the flange and the teaching point when the binocular camera takes a picture of the teaching point , and then calculate the projection of the teaching point taken by the binocular camera Robot pose , calculate the relative position of the robot flange and the teaching point when the laser displacement sensor spot is aligned with the teaching point .

5. A high-precision automatic teaching method for the entire workspace of a robot according to claim 4, characterized in that: Step S4 is specifically as follows: The robot drives the binocular camera to precisely locate the teaching points and obtain the precise position of all teaching points , which replaces the translation part of the teaching point projection , get the precise pose of the teaching point ; Calculate the robot's position when the laser displacement sensor spot is aligned with the teaching point .

6. A high-precision automatic teaching method for the entire workspace of a robot according to claim 5, characterized in that: Step S5 is specifically as follows: According to the calculated alignment pose Guide the robot to move so that the laser displacement sensor spot is aligned with the teaching point When the light spot is aligned with the teaching point, there will be a position error. At this time, the monocular camera takes a picture of the light spot and reads the reading of the laser displacement sensor. , depth information of the monocular camera , locate the image center of the light spot, and calculate the position deviation between the light spot and the teaching point in the robot coordinate system based on the camera internal parameters .

7. A high-precision automatic teaching method for the entire workspace of a robot according to claim 6, characterized in that: Step S6 is specifically as follows: In the coarse alignment pose Based on the offset, the precise alignment pose is obtained , guide the robot to deviate so that the light spot is accurately aligned with the teaching point and record the precise position of the teaching point The relative position of the flange and the teaching point during precise alignment .

8. A high-precision automatic teaching method for the entire workspace of a robot according to claim 7, characterized in that: Step S7 is specifically as follows: According to different application scenarios, the robot can replace the end tool through the quick change plate, including the suction cup and the tool with rigid connection of the gripper. The center of the tool is on the laser displacement sensor line and the distance from the center of the laser displacement sensor is ; Then the binocular camera accurately recognizes the target's position and obtains the target's precise position , calculate the target point With all teaching points distance, select the nearest teaching point , find the corresponding relative position of the flange and the teaching point when accurately aligning , that is, get the nearest teaching relative pose and calculate the precise pose of the tool to the target .

Images