Robot TCP 3D position and posture calibrator based on orthogonal vision

Abstract

The present invention discloses a technical solution of a robot TCP three-dimensional position and posture calibrator based on orthogonal vision. The image of the robot tool end is directly obtained through two built-in orthogonal visual sensors, and the three-dimensional coordinates and posture information of the robot end tool TCP can be obtained. By reading the robot's DH model parameters and the rotation angle values ​​of each axis, the automatic calibration of the TCP three-dimensional position and posture can be realized. The calibration technical solution has the outstanding advantages of high precision, intelligence, miniaturization, integration and generalization.

Description

Technical Field

[0001] The invention relates to a robot TCP calibrator, in particular to a robot TCP three-dimensional position and posture calibrator based on orthogonal vision. Background Art

[0002] The robot is an open-loop kinematic structure. The angle of joint rotation is obtained through an angle measurement device (usually an incremental encoder), and the spatial position of the current robot end effector is obtained through the robot kinematic model. Due to the inevitable errors in mechanical manufacturing and assembly, encoders, motion control and other links in the robot production process, gravity deformation, thermal deformation, clearance and wear and other random errors during the use of the robot, the actual spatial position of the robot end is usually significantly different from the ideal value, thus affecting the robot's motion accuracy.

[0003] By calibrating the robot and obtaining the accurate structure and posture parameters of the robot, the posture error of the robot can be greatly reduced, and then the absolute accuracy of the robot can be improved to the level of repeatability. In recent years, scholars from various countries have proposed many robot calibration methods, mainly kinematic loop method and axis measurement. The kinematic loop method is a method of obtaining the posture of the robot end through a measuring device and obtaining the robot joint parameters by solving the robot's kinematic equations. For example, Tianjin University, Zhejiang University, Ocean University of China, etc. use the method of using a camera at the end of the robot to collect image information of spatial feature points when the robot is in a certain posture, and obtain the three-dimensional coordinate data of the point through data processing, that is, to establish an accurate mapping relationship between the joint variables and the posture of the end flange coordinate system when the robot is in any posture. Nanjing University of Science and Technology fixes the laser at the end of the machine and places a position sensor position sensitive detector PSD in the robot workspace, so that the laser is positioned at the center of the position sensor in multiple postures to achieve point constraints. The axis measurement method is a method of abstracting the robot's joint axis into a straight line in space and using the geometric relationship between the joint axes to obtain the kinematic parameters in the model. Compared with the kinematic loop method, the axis measurement method is simple in calibration process and highly operable, so it is widely used in the process of robot calibration. For example, Shenyang Institute of Automation, Tianjin University and Huazhong University of Science and Technology use laser trackers and least square solutions of linear equations to calibrate robots.

[0004] However, various existing robot calibration methods have many problems: (1) High cost: Laser trackers are particularly expensive, with an investment of millions of dollars, which is several times the price of the robot itself, and has a very low cost-performance ratio.

[0005] (2) Low efficiency: Whether it is the binocular stereo vision calibration method or the laser tracker method, it is necessary to install the corresponding measurement components (target mirror or tool) at the end of the robot and cooperate with certain regular motion control and methods. Therefore, it is difficult to meet the needs of efficient production.

[0006] (3) Large size and weight: The existing calibration methods are basically based on laser and vision, and all of them are assembled from separate components (camera, lens, reflector, laser, etc.) to form a calibration system, which has the problems of large size and weight; (4) Installing electronic devices at the end of the robot will increase the number of electronic device cables installed on the robot arm. This will inevitably cause cable drag and rotation during the movement of the robot, greatly reducing the reliability and stability of the system.

[0007] (5) When installing tools on the robot's end tool, the tools vary greatly and each requires a separate tool to match it. This results in a variety of tools, greatly increasing the difficulty of customization, and limiting and inconvenient use.

[0008] (6) Most existing TCP calibrators can only calibrate the three-dimensional position of TCP, but not the three-dimensional posture of TCP, which affects the calibration effect and accuracy.

[0009] Therefore, there is an urgent need for a low-cost, high-efficiency, small-volume, highly integrated general robot TCP three-dimensional position and posture automatic calibration system. It is also easy to use, has no cable dragging, is not restricted by tool type, has no effect on the performance of the robot itself, and does not require robot programming. It is suitable for performance testing and verification of new robots, and can meet the general TCP calibration needs of in-service robots. Summary of the invention

[0010] The present invention aims at the defects and shortcomings of the current robot TCP calibration methods, such as low precision, large size, need for robot cooperation and provision of coordinate data, or need to install specific tools, and proposes a robot TCP three-dimensional position and posture calibrator technical solution based on orthogonal vision. The image of the robot tool end is directly obtained through two built-in orthogonal visual sensors, and then the three-dimensional coordinates and posture information of the robot end tool TCP can be obtained. By reading the robot's DH model parameters and the angle values ​​of each axis, the automatic calibration of the TCP three-dimensional position and posture can be realized. The calibration technical solution has the outstanding advantages of high precision, intelligence, miniaturization, integration, and generalization.

[0011] The present invention is achieved through the following technical solutions: The present invention proposes a robot TCP three-dimensional position and posture calibrator based on orthogonal vision, the calibrator consists of a host, a tripod, and a mobile terminal; the host consists of a visual sensor, a bracket, a bottom plate, a backlight source, a shell, a window sheet, an industrial computer, and a handle, wherein: The visual sensor is an industrial-grade visual sensor, there are two of them, which are installed inside the host and are respectively installed on the bottom plate through their own brackets, ensuring that the two visual sensors are at the same height and their optical axes are vertical, so that the three-dimensional coordinates and posture angle of the robot tool center point (TCP) can be synchronously and accurately measured; There are two brackets in total, which are completely consistent in shape and size, and correspond to the two visual sensors one by one; each visual sensor is positioned by one bracket and reliably fixed on the same base plate, so that the two visual sensors can be ensured to be at the same height and with their optical axes perpendicular; the brackets are made of metal materials to ensure accuracy and long-term stability; The bottom plate is an integrated plate structure, which can simultaneously position the two brackets to ensure that the two visual sensors are at the same height and the optical axes are vertical; the bottom plate is made of metal material to ensure accuracy and long-term stability, while having sufficient strength; The backlight source is an LED surface light source, there are two of them, which are respectively placed on the housing and are at the same height as the two visual sensors, and each backlight source corresponds to a visual sensor to provide backlight illumination for the visual sensor, thus forming a cross-cross orthogonal layout; The shell is a special-shaped structure, which is installed on the bottom plate. It can not only wrap the visual sensor to protect it from dust, water and touch, but also fix the backlight source, window sheet and industrial computer to facilitate operation and reading of information. The shell can be made of metal materials to ensure strength, or can be made of materials such as resin to reduce weight. The window piece is an optical window piece, which is made of optical materials to ensure transmittance; there are two window pieces, which are respectively installed on the housing and the front ends of the two visual sensors, and play the role of light transmission, dust prevention and waterproofing; The industrial computer is an integrated fanless industrial computer with a touch screen, which is directly mounted on the inclined surface of one side of the shell, so as to facilitate observation and operation by the operator; the industrial computer can receive image data from the two visual sensors, and perform data processing, and finally obtain the three-dimensional coordinates and posture angle of the robot tool center point (i.e. TCP), and display them on the screen; at the same time, the operator can set the parameters of the calibrator through the touch screen; There are two handles, which are installed on the bottom plate and on both sides of the housing to facilitate the transportation of the calibrator host. The tripod is a heavy-duty tripod used to support the main unit of the calibrator, and can conveniently adjust the position and posture of the main unit of the calibrator, and ensure that the position and posture of the main unit of the calibrator remain stable; The mobile terminal is an industrial-grade PDA with camera and wireless communication functions. It can obtain relevant parameters of the robot by shooting images of the robot teaching pendant interface; the mobile terminal can wirelessly communicate with the industrial computer and wirelessly send the obtained robot-related information to the industrial computer.

[0012] The working process of the robot TCP three-dimensional position and posture calibrator based on orthogonal vision provided by the present invention is as follows: first, the calibrator host is installed on a tripod, reliably fixed, and moved to a certain place in the robot workspace for stable support. Then, the operator controls the teach pendant to control the robot tool (such as a welding gun) to move to the vicinity of the center point within the working range of the calibrator host and stop. The two visual sensors in the calibrator host capture the image of the end of the tool in real time, and the TCP three-dimensional coordinates and posture angles are obtained through data processing. At the same time, the interface image of the robot teach pendant is photographed by a mobile terminal, and wirelessly transmitted to the industrial computer for image processing, and the DH model parameters of the robot and the angle values ​​of each axis are automatically identified and obtained. Finally, the industrial computer of the calibrator host obtains the three-dimensional coordinate value and posture angle of TCP in the robot base coordinate system through inverse calculation, thereby realizing the three-dimensional position and posture calibration of the robot TCP.

[0013] The robot TCP three-dimensional position and posture calibrator based on orthogonal vision provided by the present invention is also characterized in that: the visual sensor can be a combination of an industrial camera and an industrial lens to achieve high-precision image acquisition, or a visual module can be used to greatly reduce space occupancy.

[0014] The robot TCP three-dimensional position and posture calibrator based on orthogonal vision provided by the present invention is also characterized in that: the visual sensor can adopt either an industrial FA lens or a telecentric lens.

[0015] The present invention provides a robot TCP calibration method based on the robot TCP calibrator, which is specifically as follows: (1) First, open the robot teaching pendant to display the DH model parameters, use a mobile terminal to shoot the teaching pendant interface, and wirelessly transmit the teaching pendant image to the industrial computer; (2) The industrial computer processes the teaching pendant interface image transmitted by the mobile terminal, identifies and obtains the robot's DH model parameters; (3) Then, the robot drives the tool to move within the working range of the calibrator host, close to the center point, and then stops; (4) The two visual sensors in the calibrator host continuously and synchronously acquire the image sequence of the tool end, obtain the three-dimensional coordinates and attitude angle of the tool end TCP through data processing, and identify the three-dimensional coordinates and attitude angle values ​​after the tool stops through an intelligent judgment algorithm; (5) At this time, the robot teaching pendant displays the rotation angle parameters of each axis, and the teaching pendant interface is photographed by a mobile terminal, and the image is wirelessly transmitted to the industrial computer; (6) The industrial computer processes the teaching pendant interface image transmitted by the mobile terminal, identifies and obtains the rotation angle value of each axis of the robot; (7) The robot drives the tool to repeat the above process from multiple different angles and postures, obtaining multiple sets of tool three-dimensional coordinate values ​​and the angle values ​​of each axis of the robot each time; (8) The industrial computer obtains the three-dimensional coordinate value and posture angle of the robot TCP in the robot base coordinate system through inverse calculation, thereby realizing the three-dimensional position and posture calibration of the robot TCP. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the composition of the calibration instrument system of the present invention; Figure 2 It is a schematic diagram of the mainframe composition of the calibrator of the present invention; Figure 3 It is a schematic diagram of the layout of the visual sensor of the calibrator of the present invention; In the figure: 1 is the calibrator host, 2 is the tripod, 3 is the mobile terminal, 4 is the robot, 5 is the tool, 6 is the teaching pendant, 7 is the visual sensor, 8 is the bracket, 9 is the bottom plate, 10 is the backlight source, 11 is the shell, 12 is the window piece, 13 is the industrial computer, and 14 is the handle. DETAILED DESCRIPTION

[0017] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several variations and improvements may be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0018] The present invention is achieved through the following technical solutions: The present invention proposes a robot TCP three-dimensional position and posture calibrator based on orthogonal vision, the calibrator consists of a host 1, a tripod 2, and a mobile terminal 3. Figure 1 As shown; the host 1 is composed of a visual sensor 7, a bracket 8, a base plate 9, a backlight source 10, a housing 11, a window sheet 12, an industrial computer 13, and a handle 14, such as Figure 2 As shown, where: The visual sensor 7 is an industrial-grade visual sensor, and there are two of them, namely 7a and 7b, which are installed inside the host 1 and are respectively installed on the bottom plate 9 through their respective brackets 8, so that the two visual sensors 7a and 7b are at the same height and their optical axes are vertical, thereby being able to synchronously and accurately measure the three-dimensional coordinates and posture angle of the center point (i.e., TCP) of the robot tool 5; There are two brackets 8, namely 8a and 8b, which are completely consistent in shape and size, and correspond to the two visual sensors 7a and 7b respectively; each visual sensor 7a or 7b is positioned by a bracket 8a or 8b, and is reliably fixed on the same bottom plate 9, so that the two visual sensors 7a and 7b can be ensured to be at the same height and have vertical optical axes; the brackets 8a and 8b are made of metal materials, such as aluminum alloy, to ensure accuracy and long-term stability; The bottom plate 9 is an integrated plate structure, which can simultaneously position the two brackets 8a and 8b, thereby ensuring that the two visual sensors 7a and 7b are at the same height and the optical axes are vertical; the bottom plate 9 is made of metal material, such as aluminum alloy, to ensure accuracy and long-term stability, while having sufficient strength; The backlight source 10 is an LED surface light source, and there are two of them, namely 10a and 10b, which are respectively arranged on the housing 11 and are at the same height as the two visual sensors 7a and 7b. Moreover, each backlight source 10a or 10b corresponds to a visual sensor 7a or 7b, providing backlight illumination for the visual sensor 7a or 7b, thereby forming a cross-cross orthogonal layout, as shown in FIG. Figure 3 As shown; The housing 11 is a special-shaped structure, which is installed on the bottom plate 9. It can not only wrap the visual sensor 7 to protect it from dust, water and touch, but also fix the backlight source 10, the window sheet 12 and the industrial computer 13 to facilitate operation and reading of information. The housing 11 can be made of metal materials to ensure strength, or can be made of materials such as resin to reduce weight. The window sheet 12 is an optical window sheet, which is made of optical materials to ensure transmittance; there are two window sheets 12, namely 12a and 12b, which are respectively installed on the housing 11 and the front ends of the two visual sensors 7a and 7b, and play the role of light transmission, dust prevention and waterproofing; The industrial computer 13 is an integrated fanless industrial computer with a touch screen, which is directly mounted on the inclined surface of one side of the housing 11, so as to facilitate observation and operation by the operator; the industrial computer 13 can receive image data from the two visual sensors 7a and 7b, and perform data processing, and finally obtain the three-dimensional coordinates and posture angle of the center point (i.e., TCP) of the robot tool 5, and display them on the screen; at the same time, the operator can set the parameters of the calibrator through the touch screen; There are two handles 14, namely 14a and 14b, which are installed on the bottom plate 9 and on both sides of the housing 11 to facilitate the transportation of the calibrator host 1; The tripod 2 is a heavy-duty tripod used to support the main unit 1 of the calibrator, and can conveniently adjust the position and posture of the main unit 1 of the calibrator, and ensure that the position and posture of the main unit 1 of the calibrator remain stable; The mobile terminal 3 is an industrial-grade PDA with camera and wireless communication functions. It can obtain relevant parameters of the robot 4 by shooting the image of the interface of the teaching pendant 6 of the robot 4; the mobile terminal 3 can wirelessly communicate with the industrial computer 13 and wirelessly send the obtained relevant information of the robot 1 to the industrial computer 13.

[0019] The working process of the robot TCP three-dimensional position and posture calibrator based on orthogonal vision provided by the present invention is as follows: first, the calibrator host 1 is installed on the tripod 2, reliably fixed, and moved to a certain place in the working space of the robot 1 for stable support. Then, the operator controls the teach pendant 6 to control the tool 5 (such as a welding gun) of the robot 1 to move to the vicinity of the center point within the working range of the calibrator host 1 and stop. The two visual sensors 7a and 7b in the calibrator host 1 capture the image of the end of the tool 5 (i.e., TCP) in real time, and obtain the three-dimensional coordinates and posture angles of TCP through data processing. At the same time, the interface image of the robot 4 teach pendant 5 is photographed by the mobile terminal 3, and wirelessly transmitted to the industrial computer 13 for image processing, and the DH model parameters and angle values ​​of each axis of the robot 4 are automatically identified and obtained. Finally, the industrial computer 13 of the calibrator host 1 obtains the three-dimensional coordinate values ​​and posture angles of TCP in the base coordinate system of the robot 1 through inverse calculation, thereby realizing the three-dimensional position and posture calibration of the robot TCP.

[0020] The robot TCP three-dimensional position and posture calibrator based on orthogonal vision provided by the present invention is also characterized in that: the visual sensor 7 can be a combination of an industrial camera and an industrial lens to achieve high-precision image acquisition, or a visual module can be used to greatly reduce space occupancy.

[0021] The robot TCP three-dimensional position and posture calibrator based on orthogonal vision provided by the present invention is also characterized in that: the visual sensor 7 can adopt either an industrial FA lens or a telecentric lens.

[0022] The present invention provides a robot TCP calibration method based on the robot TCP calibrator, which is specifically as follows: (1) First, the teaching pendant 6 of the robot 4 is turned on to display the DH model parameters, the interface of the teaching pendant 6 is photographed by the mobile terminal 3, and the image of the teaching pendant 6 is wirelessly transmitted to the industrial computer 13; (2) The industrial computer 13 processes the interface image of the teaching pendant 6 transmitted by the mobile terminal 3, identifies and obtains the DH model parameters of the robot 4; (3) Then, the robot 4 drives the tool 5 to move into the working range of the calibrator main unit 1, close to the center point, and then stops; (4) The two visual sensors 7a and 7b in the calibrator host 1 continuously and synchronously acquire the image sequence of the end of the tool 5, obtain the three-dimensional coordinates and attitude angle of the TCP of the end of the tool 5 through data processing, and identify the three-dimensional coordinates and attitude angle values ​​of the tool 5 after it stops through an intelligent judgment algorithm; (5) At this time, the teaching pendant 6 of the robot 4 displays the rotation angle parameters of each axis, and the teaching pendant 6 interface is photographed by the mobile terminal 3, and the image is wirelessly transmitted to the industrial computer 13; (6) The industrial computer 13 processes the interface image of the teaching pendant 6 transmitted by the mobile terminal 3, identifies and obtains the rotation angle value of each axis of the robot 4; (7) The robot 4 drives the tool 5 to repeat the above process from multiple different angles and postures, and obtains multiple sets of three-dimensional coordinate values ​​of the tool 5 and the angle values ​​of each axis of the robot 4 each time; (8) The industrial computer 13 obtains the three-dimensional coordinate value and posture angle of the robot TCP in the robot base coordinate system through inverse calculation, thereby realizing the three-dimensional position and posture calibration of the robot TCP.

[0023] Compared with the prior art, the present invention has the following beneficial effects: (1) The TCP calibrator of the present invention uses two visual sensors to directly detect robot tools. It not only has high non-contact measurement accuracy, but also can calibrate the three-dimensional position and posture of TCP at the same time, which is very cost-effective. (2) The TCP calibrator of the present invention does not require any additional communication cables between the robot to be calibrated, and does not require any devices to be installed on the robot or tool. It has no associated impact on the robot and is not restricted by changes in tool types. It has strong versatility and does not require robot programming operations to complete the entire TCP calibration. The method of the present invention can be applied to both the design of new robots and the installation of accessories for robots in service, and has the best versatility; (3) The TCP calibrator described in the present invention uses a mobile terminal to capture the interface image of the robot teaching pendant, and directly identifies and obtains the required robot DH model parameters and the angle values ​​of each axis through the output distance. There is no need for manual recording and input, and the operation is simple and intelligent.

[0024] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art may make various modifications or variations within the scope of the claims, which do not affect the essence of the present invention.

Claims

1. A robot TCP three-dimensional position and posture calibrator based on orthogonal vision, the calibrator consists of a host, a tripod, and a mobile terminal; the host consists of a visual sensor, a bracket, a bottom plate, a backlight source, a shell, a window sheet, an industrial computer, and a handle, wherein: The visual sensor is an industrial-grade visual sensor, there are two of them, which are installed inside the host and are respectively installed on the bottom plate through their own brackets to ensure that the two visual sensors are at the same height and their optical axes are vertical; There are two brackets in total, which are completely consistent in shape and size, and correspond to the two visual sensors one by one; each visual sensor is positioned by one bracket and reliably fixed on the same base plate, so that the two visual sensors can be ensured to be at the same height and with their optical axes perpendicular; the brackets are made of metal materials to ensure accuracy and long-term stability; The bottom plate is an integrated plate structure, which can simultaneously position the two brackets to ensure that the two visual sensors are at the same height and the optical axes are vertical; the bottom plate is made of metal material to ensure accuracy and long-term stability, while having sufficient strength; The backlight source is an LED surface light source, there are two of them, which are respectively placed on the housing and are at the same height as the two visual sensors, and each backlight source corresponds to a visual sensor to provide backlight illumination for the visual sensor, thus forming a cross-cross orthogonal layout; The shell is a special-shaped structure, which is installed on the bottom plate. It can not only wrap the visual sensor to protect it from dust, water and touch, but also fix the backlight source, window sheet and industrial computer to facilitate operation and reading of information. The shell can be made of metal materials to ensure strength, or can be made of materials such as resin to reduce weight. The window piece is an optical window piece, which is made of optical materials to ensure transmittance; there are two window pieces, which are respectively installed on the housing and the front ends of the two visual sensors, and play the role of light transmission, dust prevention and waterproofing; The industrial computer is an integrated fanless industrial computer with a touch screen, which is directly mounted on the inclined surface of one side of the shell, so as to facilitate observation and operation by the operator; the industrial computer can receive image data from the two visual sensors, and perform data processing, and finally obtain the three-dimensional coordinates and posture angle of the robot tool center point (i.e. TCP), and display them on the screen; at the same time, the operator can set the parameters of the calibrator through the touch screen; There are two handles, which are installed on the bottom plate and on both sides of the housing to facilitate the transportation of the calibrator host. The tripod is a heavy-duty tripod used to support the main unit of the calibrator, and can conveniently adjust the position and posture of the main unit of the calibrator, and ensure that the position and posture of the main unit of the calibrator remain stable; The mobile terminal is an industrial-grade PDA with camera and wireless communication functions. It can obtain relevant parameters of the robot by shooting images of the robot teaching pendant interface; the mobile terminal can wirelessly communicate with the industrial computer and wirelessly send the obtained robot-related information to the industrial computer.

2. The orthogonal vision-based robot TCP three-dimensional position and posture calibrator according to claim 1, further characterized by: The visual sensor can be a combination of an industrial camera and an industrial lens to achieve high-precision image acquisition, or a visual module to significantly reduce space occupancy.

3. The robot TCP three-dimensional position and posture calibrator based on orthogonal vision according to claim 1 is also characterized in that: The visual sensor 7 may adopt an industrial FA lens or a telecentric lens.

4. A robot TCP three-dimensional position and posture calibration method based on any of the above items, specifically as follows: (1) First, open the robot teaching pendant to display the DH model parameters, use a mobile terminal to shoot the teaching pendant interface, and wirelessly transmit the teaching pendant image to the industrial computer; (2) The industrial computer processes the teaching pendant interface image transmitted by the mobile terminal, identifies and obtains the DH model parameters of the robot; (3) Then, the robot drives the tool to move within the working range of the calibrator host, close to the center point, and then stops; (4) The two visual sensors in the calibrator host continuously and synchronously acquire the image sequence of the tool end, obtain the three-dimensional coordinates and attitude angle of the tool end TCP through data processing, and identify the three-dimensional coordinates and attitude angle values ​​after the tool stops through an intelligent judgment algorithm; (5) At this time, the robot teaching pendant displays the rotation angle parameters of each axis, and the teaching pendant interface is photographed by a mobile terminal, and the image is wirelessly transmitted to the industrial computer; (6) The industrial computer processes the teaching pendant interface image transmitted by the mobile terminal, identifies and obtains the rotation angle value of each axis of the robot; (7) The robot drives the tool to repeat the above process from multiple different angles and postures, obtaining multiple sets of tool three-dimensional coordinate values ​​and the angle values ​​of each axis of the robot each time; (8) The industrial computer obtains the three-dimensional coordinate value and posture angle of the robot TCP in the robot base coordinate system through inverse calculation, thereby realizing the three-dimensional position and posture calibration of the robot TCP.

Images