Composite robot CNC feeding and discharging method and device

By combining 3D cameras and 2D cameras in the composite robot CNC loading and unloading system, the laser SLAM navigation system is used to achieve automatic navigation, which solves the problems of low deployment efficiency, high cost and accuracy, and achieves efficient and accurate loading and unloading operations.

CN120206521AActive Publication Date: 2025-06-27BEIJING XIAOER INTELLIGENT TECHNOLOGY CO LTD

Patent Information

Application Number
CN202510434223.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-04-08
Publication Date
2025-06-27
Estimated Expiration
2045-04-08

Smart Images

  • Figure CN120206521A_ABST
    Figure CN120206521A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of machine vision fusion application, and particularly relates to a composite robot CNC feeding and discharging method and device. Comprising a composite robot and a CNC machine tool; the composite robot comprises a movable chassis, a mechanical arm, a feeding and discharging clamping jaw, a 3D camera and a 2D camera. The 3D camera and the 2D camera are mounted on a mounting platform at the tail end of the mechanical arm; the 2D camera and the 3D camera are added in the first deployment stage and the nth deployment stage to assist deployment, teaching of each CNC machine tool is not needed, the 3D camera is dismounted in the feeding and discharging stage, only the 2D camera is used in the feeding and discharging stage, and the cost is lower; and meanwhile, a 2D positioning identification plate or a 3D positioning identification plate is adopted, the feeding base does not need to be photographed, and the influence of cutting liquid, metal filings, discharging workpieces and the like on the camera algorithm precision is avoided.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of machine vision fusion applications, and specifically relates to a method and device for loading and unloading a CNC machine tool by a composite robot. Background Art

[0002] In the process of the development of manufacturing automation, the application of composite robots for loading and unloading CNC machine tools is becoming increasingly widespread. A composite robot integrates a movable chassis and a robotic arm, and the end of the robotic arm can carry a loading and unloading gripper and a camera. Among them, the camera, as a key 3D positioning sensor, mainly includes two types: 2D cameras and 3D cameras. The working principles of the two are significantly different, and each has its own advantages and disadvantages in the CNC loading and unloading scenario.

[0003] A 2D camera only has two-dimensional information. When used for CNC loading and unloading, it needs to be paired with a special 2D feature pattern (i.e., a 2D positioning identification plate) to obtain 3D information for aligning the material with the loading seat. The size of this 2D feature pattern is fixed and its orientation can be recognized. When in use, it needs to be fixed near the loading seat and maintain a fixed spatial relationship with the loading seat during each loading and unloading. Through algorithms, the precise pose (including three-dimensional coordinates xyz and three-axis rotation angles rpy) in the 2D camera coordinate system can be calculated based on its fixed size. When realizing loading and unloading based on a 2D camera, during the deployment stage, the optimal TCP pose for loading and unloading and the pose phase difference of the 2D positioning identification plate need to be determined through teaching. During the subsequent loading and unloading stage, the 2D camera takes pictures of the 2D positioning identification plate for positioning, and precise loading and unloading are completed using the data saved during the deployment stage. However, teaching is required for the loading seat of each CNC machine tool. If there are multiple loading seats in one machine tool, the teaching time during the deployment stage will increase significantly, which is not conducive to the flexible arrangement of production.

[0004] A 3D camera has depth information and can directly photograph the loading seat, and uses a template matching algorithm to obtain the precise pose (including three-dimensional coordinates xyz and three-axis rotation angles rpy) of the center of the matching template of the loading seat in the 3D camera coordinate system. When realizing loading and unloading based on a 3D camera, as long as multiple CNC machine tools use loading seats of the same shape, only the first one needs to be taught to determine the optimal TCP pose for loading and unloading, and the loading seats of other machine tools do not need to be taught repeatedly. During the subsequent loading and unloading stage, the teaching results of the first machine tool can be directly used. However, this solution has obvious defects. The photographing and matching positioning of the loading seat are easily affected by cutting fluid, metal chips, workpieces, etc. To ensure the positioning accuracy, the loading seat needs to be specially designed, and the actual operation is difficult; at the same time, high-precision 3D cameras are costly and large in size, increasing the equipment investment and space occupation.

[0005] Currently, in the field of CNC loading and unloading of composite robots, issues such as deployment efficiency, camera cost and volume, and loading and unloading accuracy restrict each other, and innovative solutions are urgently needed. Based on these actual dilemmas, the present invention aims to make up for each other's weaknesses, enabling the composite robot to improve deployment efficiency, reduce costs while ensuring accuracy in the CNC loading and unloading scenario, and promoting the wide application of this technology. Summary of the Invention

[0006] To make up for the deficiencies of the prior art, the present invention provides a CNC loading and unloading device for a composite robot, including a composite robot and a CNC machine tool;

[0007] The composite robot includes a mobile chassis, a robotic arm, a loading and unloading gripper, and a 3D camera; one end of the robotic arm is installed on the mobile chassis, and the other end installs the loading and unloading gripper through the installation platform at the end of the robotic arm; the 3D camera is installed on the installation platform;

[0008] Each CNC machine tool has a loading seat and a positioning identification plate with the same shape and size; the relative positions of the loading seat and the positioning identification plate on each CNC machine tool are fixed;

[0009] The mobile chassis of the composite robot uses laser SLAM navigation to navigate to a specified position in front of multiple CNC machine tools; each CNC machine tool is provided with its own navigation station, and all navigation stations are marked on the laser SLAM navigation map. The composite robot is scheduled to the navigation station corresponding to any CNC machine tool with loading and unloading requirements through the laser SLAM navigation system and the navigation stations on the navigation map.

[0010] Preferably, another 2D camera is installed on the installation platform; the positioning identification plate is a 2D positioning identification plate; the 3D camera is used in the deployment stage of the CNC machine tool and the composite robot; the 2D camera combines with the 2D positioning identification plate to form pose data, and the data is recorded through the control software, and the best loading pose of the composite robot for each CNC machine tool is recorded for use in the loading and unloading stage.

[0011] Preferably, the positioning identification plate is a 3D positioning identification plate; the 3D camera is used in the deployment stage and the loading and unloading stage of the CNC machine tool and the composite robot; the 3D camera combines with the 3D positioning identification plate to form pose data, and the data is recorded through the control software, and the best loading pose of the composite robot for each CNC machine tool is recorded for use in the loading and unloading stage.

[0012] A method for CNC loading and unloading of a composite robot, the deployment stage of the first CNC machine tool includes the following steps:

[0013] Step 1: Obtain the best loading and unloading pose of the TCP through teaching;

[0014] Step 2: Adjust the pose of the end of the robotic arm to the optimal pose for the 3D camera to take pictures of the loading seat. Record the adjustment amount ΔT3D1 for Step 3 and record the coordinate ZTCP3D of the TCP for use during the 3D camera's pose positioning in the deployment process of other CNC machines. Calculate and record the pose T3D of the optimal loading and unloading pose in the 3D camera coordinate system;

[0015] Step 3: Adjust the pose of the end of the robotic arm again to the optimal pose for the 2D camera to take pictures of the 2D positioning identification plate. Record the TCP coordinate ZTCP2D for use during the 2D camera's pose positioning in the loading and unloading stage. Record the first optimal loading and unloading pose T2D1 in the 2D camera's three-dimensional coordinate system relative to the center of the 2D positioning identification plate on the first CNC;

[0016] The specific content of Step 2 is as follows:

[0017] Adjust the pose of the robotic arm so that the loading seat is within the field of view of the 3D camera and in the middle position. Record the relative adjustment amount ΔT3D1 of the TCP and the TCP coordinate ZTCP3D at this time. The high-precision 3D camera fixed at the end of the robotic arm takes pictures of the loading seat to obtain the matching template of the loading seat and the pose T3D1 of the center point of the loading seat in the 3D camera's three-dimensional coordinate system. Calculate the pose T3D of the optimal loading and unloading pose in the 3D camera coordinate system according to the adjustment amount of the pose of the end of the robotic arm: T3D = T3D1 - ΔT3D1;

[0018] The specific content of Step 3 is as follows:

[0019] Adjust the end of the robotic arm so that the 2D positioning identification plate is within the field of view of the 2D camera and in the middle position. Record the total adjustment amount ΔT2D1 of the robotic arm TCP relative to the optimal loading and unloading pose at this time and record the TCP coordinate at this time as ZTCP2D. The 2D camera takes pictures of the 2D positioning identification plate to obtain the pose Z2D1 of the center point of the 2D positioning identification plate in the 2D camera's three-dimensional coordinate system. Obtain the first optimal loading and unloading pose T2D1 = ΔT2D1 - Z2D1 in the 2D camera's three-dimensional coordinate system relative to the center of the 2D positioning identification plate on the first CNC and record this value for use in the loading and unloading stage.

[0020] Preferably, it further includes the deployment stage of the nth CNC machine tool, where n > 1; the composite robot remains stable at the entrance of the nth CNC, referring to the ZTCP3D recorded in the deployment of the first CNC, adjusts the photographing position of the 3D camera, takes a photo of the loading seat, uses the matching template of the loading seat obtained in the deployment of the first CNC, and obtains the pose T3Dn of the center point of the current loading seat in the 3D camera's three-dimensional coordinate system; at the same time, referring to the ZTCP2D recorded in the deployment of the first CNC, adjusts the photographing position of the 2D camera, records the adjustment amount ΔT2Dn relative to the TCP of the robotic arm this time, the 2D camera fixed at the end of the robotic arm takes a photo of the 2D positioning identification plate, and obtains the pose Z2Dn of the center point of the 2D positioning identification plate in the 2D camera's three-dimensional coordinate system; obtains the best loading and unloading pose T2Dn of the nth CNC in the 2D camera's three-dimensional coordinate system relative to the center of the 2D positioning identification plate on the nth CNC, T2Dn = ΔT2Dn + (T3Dn - T3D) - Z2Dn, and records this value for use in the loading and unloading stage.

[0021] Preferably, the loading and unloading stage includes the following:

[0022] Only use the 2D camera and do not use the 3D camera;

[0023] Dispatch the mobile chassis of the composite robot to remain stable at the entrance of the mth CNC, where m ≥ 1; referring to the ZTCP2D recorded in the deployment of the first CNC, adjust the end of the robotic arm to the photographing position of the 2D camera, the 2D camera fixed at the end of the robotic arm takes a photo of the 2D positioning identification plate fixed on the mth CNC, and obtains the pose Z2Dm of the center point of the 2D positioning identification plate in the 2D camera's three-dimensional coordinate system at this photographing moment; read the best loading and unloading pose T2Dm of the mth CNC relative to the center of the 2D positioning identification plate on the mth CNC recorded previously in the 2D camera's three-dimensional coordinate system; calculate the relative increment W2Dm that the TCP needs to adjust to reach the best loading and unloading pose of the mth CNC for this loading task, W2Dm = Z2Dm + T2Dm.

[0024] A method for a composite robot to load and unload a CNC, the deployment stage of the first CNC includes the following steps:

[0025] Step 1: Obtain the best loading and unloading pose of the TCP through teaching.

[0026] Step 2: Adjust the pose of the end of the robotic arm to the best photographing pose of the 3D camera for the loading seat, record the adjustment amount ΔT3D1 this time for use in Step 3, record the coordinate ZTCP3D of the TCP for use in the 3D camera photographing and positioning during the deployment of other CNC machine tools; calculate and record the pose T3D of the best loading and unloading pose in the 3D camera coordinate system.

[0027] Step 3: Adjust the pose of the end effector of the robotic arm again to the optimal pose for the 3D camera to take pictures of the 3D positioning identification board, and record the TCP coordinate ZTCP2D for use in the 3D camera pose positioning during the loading and unloading stage. Record, in the 3D camera's three-dimensional coordinate system, the optimal loading and unloading pose T2D1 of the first unit relative to the center of the 3D positioning identification board on the first CNC machine.

[0028] The specific steps of Step 2 are as follows:

[0029] Adjust the pose of the robotic arm so that the loading seat is within the field of view of the 3D camera and in the middle position, record the relative adjustment amount ΔT3D1 of the TCP and the TCP coordinate ZTCP3D at this time; the high-precision 3D camera fixed at the end effector of the robotic arm takes pictures of the loading seat to obtain the matching template of the loading seat and the pose T3D1 of the center point of the loading seat in the 3D camera's three-dimensional coordinate system; calculate the pose T3D of the optimal loading and unloading pose in the 3D camera coordinate system according to the adjustment amount of the pose of the end effector of the robotic arm, where T3D = T3D1 - ΔT3D1.

[0030] The specific steps of Step 3 are as follows:

[0031] Adjust the end effector of the robotic arm so that the 3D positioning identification board is within the field of view of the 3D camera and in the middle position, record the total adjustment amount ΔT2D1 of the robotic arm TCP relative to the optimal loading and unloading pose at this time, and record the TCP coordinate at this time as ZTCP2D. The 3D camera takes pictures of the 3D positioning identification board to obtain the pose Z2D1 of the center point of the 3D positioning identification board in the 3D camera's three-dimensional coordinate system; obtain the optimal loading and unloading pose T2D1 of the first unit relative to the center of the 3D positioning identification board on the first CNC machine in the 3D camera's three-dimensional coordinate system, where T2D1 = ΔT2D1 - Z2D1, and record this value for use in the loading and unloading stage.

[0032] Preferably, the loading and unloading stage includes the following:

[0033] It also includes the deployment stage of the nth CNC machine tool, where n > 1; the composite robot remains stable at the entrance of the nth CNC, referring to the ZTCP3D recorded in the deployment of the first CNC, adjusts the photographing position of the 3D camera, takes a photo of the loading seat, uses the matching template of the loading seat obtained from the deployment of the first CNC, and obtains the pose T3Dn of the center point of the current loading seat in the 3D camera's three-dimensional coordinate system; at the same time, referring to the ZTCP2D recorded in the deployment of the first CNC, adjusts the photographing position of the 3D camera, records the adjustment amount ΔT2Dn relative to the TCP of the robotic arm this time, and the 3D camera fixed at the end of the robotic arm takes a photo of the 3D positioning identification board, obtaining the pose Z2Dn of the center point of the 3D positioning identification board in the 3D camera's three-dimensional coordinate system; obtains, in the 3D camera's three-dimensional coordinate system, relative to the center of the 3D positioning identification board on the nth CNC, the optimal loading and unloading pose T2Dn of the nth CNC as T2Dn = ΔT2Dn + (T3Dn - T3D) - Z2Dn, and records this value for use in the loading and unloading stage.

[0034] Preferably, the loading and unloading stage includes the following:

[0035] Dispatch the mobile chassis of the composite robot to remain stable at the entrance of the mth CNC, where m ≥ 1; referring to the ZTCP2D recorded in the deployment of the first CNC, adjust the end of the robotic arm to the photographing position of the 3D camera, and the 3D camera fixed at the end of the robotic arm takes a photo of the 3D positioning identification board fixed on the mth CNC, obtaining the pose Z2Dm of the center point of the 3D positioning identification board in the 3D camera's three-dimensional coordinate system at the moment of this photo; read the optimal loading and unloading pose T2Dm of the mth CNC relative to the center of the 3D positioning identification board on the mth CNC recorded previously in the 3D camera's three-dimensional coordinate system; calculate the relative increment W2Dm that the TCP needs to adjust to reach the optimal loading and unloading pose of the mth CNC for this loading task as W2Dm = Z2Dm + T2Dm.

[0036] Preferably, in the deployment stage of the first CNC machine tool and the deployment stage of the nth CNC machine tool, after determining the optimal TCP loading and unloading pose, the position of the current composite robot's mobile chassis in the laser SLAM navigation system is calibrated and recorded.

[0037] The beneficial effects of the present invention are as follows:

[0038] 1. In the deployment stage of the present invention, teaching is carried out during the deployment of the first CNC machine tool, combined with the 3D camera taking a photo of the loading seat, and the data for the deployment stage is calculated for the deployment of other CNC machine tools later. It is not necessary to carry out teaching for each CNC machine tool, effectively improving the deployment efficiency.

[0039] 2. In the deployment stage, the present invention can obtain and record the relative pose difference between the 2D positioning identification plate and the loading seat by using a 2D camera in combination with a 3D camera. During the loading and unloading stage, the relative pose difference can be directly used without taking pictures of the loading seat, thus avoiding the influence of cutting fluid, metal chips, unloading workpieces, etc. on the accuracy of the camera algorithm. At the same time, in the loading and unloading stage, the form of cooperation between the 3D camera and the 3D positioning identification plate can also be adopted. Without using the 2D camera and the 2D positioning identification plate, the influence of cutting fluid, metal chips, unloading workpieces, etc. on the accuracy of the camera algorithm can also be effectively avoided.

[0040] 3. In the deployment stage, the present invention uses a 2D camera in combination with a 2D positioning identification plate to measure the relative pose adjustment amount, calculates the optimal TCP loading and unloading position, and realizes removing the 3D camera and only using the 2D camera during the loading and unloading stage. This not only has a lower cost but also avoids the large volume and weight of the 3D camera, which causes many inconveniences during actual use. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The present invention will be further described below with reference to the accompanying drawings.

[0042] Figure 1 is a flowchart of a method for loading and unloading a CNC by a composite robot provided by the present invention;

[0043] Figure 2 is a 3D camera and a loading seat used in an embodiment of the present invention;

[0044] Figure 3 is a 2D camera and a 2D positioning identification plate used in an embodiment of the present invention;

[0045] Figure 4 is a 3D camera and a 3D positioning identification plate used in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] In order to make the technical means, creative features, achieved purposes and effects of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0047] Embodiment 1: As Figures 1 - 3 , a composite robot CNC loading and unloading device provided by the present invention includes a composite robot and a CNC machine tool;

[0048] The composite robot includes a mobile chassis, a robotic arm, a loading and unloading gripper, a 3D camera, and a 2D camera; one end of the robotic arm is installed on the mobile chassis, and the other end installs the loading and unloading gripper through a mounting platform; the 3D camera and the 2D camera are installed on the mounting platform;

[0049] Each CNC machine tool has a loading seat and a 2D positioning identification plate with the same shape and size; the relative positions of the loading seat and the 2D positioning identification plate on each CNC machine tool are fixed;

[0050] As Figure 1 shown, a method for loading and unloading a composite robot CNC uses a composite robot CNC loading and unloading device to realize the deployment and loading and unloading of a composite robot and multiple CNC machine tools, specifically including the following:

[0051] (1) In the deployment stage of the first CNC, the mobile chassis of the composite robot remains stable at the entrance of the first CNC. First, the optimal loading pose of the TCP (Tool Center Point) at the end of the robotic arm is determined by teaching.

[0052] Figure 2 The 3D camera and the loading seat used in the embodiment of the present invention are shown. The loading seat is used to place the workpiece to be processed on the CNC and has sufficient 3D features. Then, the end of the robotic arm is adjusted to a position suitable for the 3D camera to take a picture. At this time, the position suitable for the camera to take a picture means that the loading seat is basically within the field of view of the 3D camera and in the middle position. Record the adjustment amount ΔT3D1 of the TCP (for example, ΔZ = 300mm, and the others are 0) and the TCP coordinate ZTCP3D at this time (for example, Z = 500mm, X = 50mm, and the others are 0); According to the adjustment amount of the end pose of the robotic arm, the optimal loading and unloading pose in the 3D camera coordinate system is T3D = T3D1 - ΔT3D1 (obtaining Z = 200mm, X = 50mm, and the others are 0);

[0053] Since both the 3D camera and the TCP are fixed at the end of the robotic arm, this pose coordinate is applicable to the subsequent nth CNC. Record this value for use in the deployment stage of other CNC machine tools using the same-shaped loading seat. Therefore, it is not necessary to teach each CNC, which greatly saves the deployment difficulty and deployment time.

[0054] Figure 3The 2D camera and 2D positioning identification plate used in the embodiments of the present invention. The black square frame composed of a square with an outer side length of 50 mm and a square with an inner side length of 40 mm is a 2D feature pattern, and the TM character inside is used for orientation recognition. Then, adjust the end of the robotic arm to a suitable position for the 2D camera to take a picture. At this time, the suitable picture-taking position means that the 2D positioning identification plate is basically within the field of view of the 2D camera and in the middle position. Record the total adjustment amount ΔT2D1 of the TCP of the robotic arm relative to the optimal loading and unloading pose at this time (for example, ΔZ = 500 mm, ΔX = 100, and the others are 0) and the TCP coordinate ZTCP2D at this time. The 2D camera fixed at the end of the robotic arm takes a picture of the 2D positioning identification plate fixed on the first CNC, and obtains the pose Z2D1 of the center point of the 2D positioning identification plate in the stereo coordinate system of the 2D camera (for example, Z = 200 mm, X = 5 mm, Y = 10 mm, and the others are 0); furthermore, calculate that in the stereo coordinate system of the 2D camera, relative to the center of the 2D positioning identification plate on the first CNC, the first optimal loading and unloading pose T2D1 = ΔT2D1 - Z2D1 (obtaining Z = 300 mm, X = 95 mm, Y = -10 mm, and the others are 0), and record this value for use in the loading stage.

[0055] (2) In the deployment stage of the nth (>1) CNC, the mobile chassis of the composite robot remains stable at the door of the nth CNC. Referring to the ZTCP3D recorded in the deployment of the first CNC machine tool, adjust the end of the robotic arm to a suitable position for the 3D camera to take a picture. The high-precision 3D camera fixed at the end of the robotic arm takes a picture of the loading seat, and uses the matching template of the loading seat obtained in the first deployment to obtain the pose T3Dn of the center point of the current loading seat in the stereo coordinate system of the 3D camera (for example, Z = 500 mm, X = 55 mm, Y = 5 mm, and the others are 0);

[0056] Referring to the ZTCP2D recorded in the first deployment, adjust the end of the robotic arm to a suitable position for the 2D camera to take a picture again, and record the adjustment amount ΔT2Dn relative to the TCP of the robotic arm this time (for example, ΔZ = 200 mm, ΔX = 100, and the others are 0). The 2D camera fixed at the end of the robotic arm takes a picture of the 2D positioning identification plate fixed on the nth CNC, and obtains the pose Z2Dn of the center point of the 2D positioning identification plate in the stereo coordinate system of the 2D camera (for example, Z = 202 mm, X = 5 mm, Y = 15 mm, and the others are 0); furthermore, calculate that in the stereo coordinate system of the 2D camera, relative to the center of the 2D positioning identification plate on the nth CNC, the nth optimal loading and unloading pose T2Dn = ΔT2Dn + (T3Dn - T3D) - Z2Dn, obtaining (Z = 200 + 300 - 202 = 298 mm, X = 100 + 5 - 5 = 100 mm, Y = 0 + 5 - 15 = -10 mm, and the others are 0), and record this value for use in the loading stage.

[0057] During the deployment of the first CNC machine tool, teach by combining a 3D camera to take pictures of the 3D positioning identification board or the loading seat, calculate the deployment stage data for the deployment of other CNC machine tools in the later stage, and there is no need for each CNC machine tool to be taught, effectively improving the deployment efficiency.

[0058] (3) Loading and unloading stage: According to the needs of CNC loading and unloading, dispatch the mobile chassis of the composite robot to stay stable at the entrance of the m(≥1)th CNC machine tool. Referring to the ZTCP2D recorded in the first deployment, adjust the end of the robotic arm to a suitable position for the 2D camera to take pictures. The 2D camera fixed at the end of the robotic arm takes pictures of the 2D positioning identification board fixed on the mth CNC machine tool, and obtains the pose Z2Dm of the center point of the 2D positioning identification board in the 2D camera stereo coordinate system at the moment of this picture taking (for example, Z = 201mm, X = 0mm, Y = 5mm, and others are 0); read the pose T2Dm of the optimal loading pose of the mth CNC machine tool relative to the center of the 2D positioning identification board on the mth CNC machine tool, which was previously recorded in the 2D camera stereo coordinate system (for example, Z = 300mm, X = 95mm, Y = -10mm, and others are 0); calculate the relative increment W2Dm that needs to be adjusted for the TCP to reach the optimal loading pose of the mth CNC machine tool for this loading task, W2Dm = Z2Dm + T2Dm (obtained, Z = 300 + 201 = 501mm, X = 0 + 95 = 95mm, Y = 5 - 10 = -5mm, and others are 0).

[0059] The mobile chassis of the composite robot stops at the entrance of each CNC machine tool through an automatic navigation method. When using laser SLAM navigation as the standard, the positioning accuracy is ±10mm, which is relatively small compared to the imaging field of view of the 3D camera and the 2D camera. Therefore, in a better practice, during the deployment stage and the loading and unloading stage of the n(>1)th CNC machine tool, the ZTCP3D or ZTCP2D recorded in the deployment stage of the first CNC machine tool can be used to adjust the end of the robotic arm to a suitable position for the 3D camera to take pictures or a suitable position for the 2D camera to take pictures; of course, methodologically, it is not necessary to use the ZTCP3D or ZTCP2D recorded in the deployment stage of the first CNC machine tool, as long as the end of the robotic arm can be adjusted to a suitable position for the 3D camera to take pictures or a suitable position for the 2D camera to take pictures before each picture taking;

[0060] By using the 2D camera and the 3D camera, the relative pose difference between the 2D positioning identification board and the loading seat can be obtained and recorded during the deployment stage, and the relative pose difference can be directly used during the loading and unloading stage without taking pictures of the loading seat, avoiding the influence of cutting fluid, metal chips, unloading workpieces, etc. on the accuracy of the camera algorithm;

[0061] Moreover, in the first deployment stage and the nth deployment stage, 2D cameras and 3D cameras are added to assist in deployment, eliminating the need for teaching each CNC machine tool, thus greatly improving the deployment efficiency. Meanwhile, the 3D camera is removed during the loading and unloading stage, and only 2D cameras are used during the loading and unloading stage, which not only reduces costs but also avoids the inconvenience caused by the large volume of the 3D camera in some scenarios.

[0062] Embodiment 2: The difference between this embodiment and Embodiment 1 is that only 3D cameras and 3D positioning identification plates are used during both the deployment stage and the loading and unloading stage, while 2D cameras and 2D identification plates are not used; the 3D positioning identification plate is located on the CNC machine tool and has a relative pose difference with the loading seat, and this relative pose difference on each CNC machine tool may be slightly different, but remains unchanged during the deployment and loading and unloading stages; during the deployment stage, the 3D camera takes pictures of the 3D positioning identification plate, and the template matching algorithm calculates the pose T2D1 of the optimal loading pose relative to the center of the 3D positioning identification plate on the first CNC in the 3D camera's stereo coordinate system, and the pose T2Dn of the optimal loading pose relative to the center of the 2D positioning identification plate on the nth CNC. The calculation formula is the same as that in Embodiment 1.

[0063] Figure 4 This is the 3D camera and the 3D positioning identification plate used in the embodiments of the present invention. The 3D positioning identification plate is an oxidized aluminum block with a size of 5 cm (length) x 5 cm (width) x 0.5 cm (thickness), and has a through hole with a diameter of 1 cm on the left for positioning the orientation. Specifically, it includes the following:

[0064] (1) During the deployment stage of the first CNC, the mobile chassis of the composite robot remains stable at the entrance of the first CNC. First, the optimal loading pose of the TCP (Tool Center Point) at the end of the robotic arm is determined through teaching.

[0065] Then, the end of the robotic arm is adjusted to a position suitable for the 3D camera to take pictures. At this time, the suitable position for the camera to take pictures means that the loading seat is basically within the field of view of the 3D camera and in the middle position. Record the relative adjustment amount ΔT3D1 of the TCP (for example, ΔZ = 300 mm, and the others are 0) and the TCP coordinate ZTCP3D at this time (for example, Z = 500 mm, X = 50 mm, and the others are 0); according to the adjustment amount of the end pose of the robotic arm, the pose T3D of the optimal loading and unloading pose in the 3D camera coordinate system is T3D = T3D1 - ΔT3D1 (resulting in Z = 200 mm, X = 50 mm, and the others are 0);

[0066] Since both the 3D camera and the TCP are fixed at the end of the robotic arm, this pose coordinate is applicable to the subsequent nth CNC. Record this value for use in the deployment stage of other CNC machine tools with the same-shaped loading seat, so there is no need to teach each CNC, greatly saving the deployment difficulty and deployment time.

[0067] Then, adjust the end of the robotic arm to a position suitable for the 3D camera to take pictures. At this time, the suitable picture-taking position means that the 3D positioning identification board is basically within the field of view of the 3D camera and in the middle position. Record the total adjustment amount ΔT2D1 of the robotic arm TCP relative to the optimal loading and unloading pose at this time (for example, ΔZ = 500mm, ΔX = 100, and the others are 0) and the TCP coordinate ZTCP2D at this time. The 3D camera fixed at the end of the robotic arm takes pictures of the 3D positioning identification board fixed on the first CNC, and obtains the pose Z2D1 of the center point of the 3D positioning identification board in the 3D camera's three-dimensional coordinate system (for example, Z = 200mm, X = 5mm, Y = 10mm, and the others are 0); furthermore, calculate that in the 3D camera's three-dimensional coordinate system, relative to the center of the 3D positioning identification board on the first CNC, the first optimal loading and unloading pose T2D1 = ΔT2D1 - Z2D1 (obtaining Z = 300mm, X = 95mm, Y = -10mm, and the others are 0), and record this value for use in the loading stage.

[0068] (2) During the deployment stage of the nth (>1) CNC, the mobile chassis of the composite robot remains stable at the entrance of the nth CNC. Referring to the ZTCP3D recorded during the deployment of the first CNC machine tool, adjust the end of the robotic arm to a position suitable for the 3D camera to take pictures. The high-precision 3D camera fixed at the end of the robotic arm takes pictures of the loading seat, and uses the matching template of the loading seat obtained during the deployment of the first one to obtain the pose T3Dn of the center point of the current loading seat in the 3D camera's three-dimensional coordinate system (for example, Z = 500mm, X = 55mm, Y = 5mm, and the others are 0);

[0069] Referring to the ZTCP2D recorded during the deployment of the first one, adjust the end of the robotic arm to a position suitable for the 3D camera to take pictures again, and record the adjustment amount ΔT2Dn relative to the robotic arm TCP this time (for example, ΔZ = 200mm, ΔX = 100, and the others are 0). The 3D camera fixed at the end of the robotic arm takes pictures of the 3D positioning identification board fixed on the nth CNC, and obtains the pose Z2Dn of the center point of the 3D positioning identification board in the 3D camera's three-dimensional coordinate system (for example, Z = 202mm, X = 5mm, Y = 15mm, and the others are 0); furthermore, calculate that in the 3D camera's three-dimensional coordinate system, relative to the center of the 3D positioning identification board on the nth CNC, the nth optimal loading and unloading pose T2Dn = ΔT2Dn + (T3Dn - T3D) - Z2Dn, obtaining (Z = 200 + 300 - 202 = 298mm, X = 100 + 5 - 5 = 100mm, Y = 0 + 5 - 15 = -10mm, and the others are 0), and record this value for use in the loading stage.

[0070] During the teaching process when deploying the first CNC machine tool, the 3D camera is used to take pictures of the 3D positioning identification board or the loading seat, and the data obtained during the deployment stage is calculated for the subsequent deployment of other CNC machine tools. There is no need for each CNC machine tool to be taught, effectively improving the deployment efficiency.

[0071] (3) Loading and unloading stage: According to the requirements of CNC loading and unloading, the mobile chassis of the composite robot is scheduled to remain stable at the entrance of the m-th (m≥1) CNC machine tool. Referring to the ZTCP2D recorded during the deployment of the first CNC machine tool, the end of the robotic arm is adjusted to a suitable position for the 3D camera to take pictures. The 3D camera fixed at the end of the robotic arm takes pictures of the 3D positioning identification board fixed on the m-th CNC machine tool, and obtains the pose Z2Dm of the center point of the 3D positioning identification board in the 3D camera's stereo coordinate system at the moment of this picture taking (for example, Z = 201mm, X = 0mm, Y = 5mm, and others are 0); Read the pose T2Dm of the optimal loading pose of the m-th CNC machine tool relative to the center of the 3D positioning identification board on the m-th CNC machine tool recorded in the 3D camera's stereo coordinate system before (for example, Z = 300mm, X = 95mm, Y = -10mm, and others are 0); Calculate the relative increment W2Dm that needs to be adjusted for the TCP to reach the optimal loading pose of the m-th CNC machine tool for this loading task, where W2Dm = Z2Dm + T2Dm (obtained, Z = 300 + 201 = 501mm, X = 0 + 95 = 95mm, Y = 5 - 10 = -5mm, and others are 0).

[0072] The mobile chassis of the composite robot stops at the entrance of each CNC machine tool through an automatic navigation method. When using the laser SLAM navigation standard configuration, the positioning accuracy is ±10mm, which is relatively small compared to the imaging field of view of the 3D camera. Therefore, in a better practice, during the deployment stage and the loading and unloading stage of the n-th (n>1) CNC machine tool, the ZTCP3D or ZTCP2D recorded during the deployment stage of the first CNC machine tool can be used to adjust the end of the robotic arm to a suitable position for the 3D camera to take pictures; Of course, in terms of method, it is not necessary to use the ZTCP3D or ZTCP2D recorded during the deployment stage of the first CNC machine tool, as long as the end of the robotic arm can be adjusted to a suitable position for the 3D camera to take pictures or a suitable position for the 2D camera to take pictures before each picture taking;

[0073] Through the cooperation of the 3D camera and the 3D positioning identification board during the deployment stage, the 3D camera takes pictures of the 3D positioning identification board during the loading and unloading stage. Combining the relative pose difference between the 3D positioning identification board and the loading seat recorded during the deployment stage, and then moving to the best TCP loading and unloading position through calculation; Compared with the 3D camera directly taking pictures of the loading seat, it can effectively avoid the influence of cutting fluid, metal chips, unloading workpieces, etc. on the accuracy of the camera algorithm.

[0074] The above two embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent substitutions can still be made to the specific implementation manners of the present invention, and any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention shall be covered within the protection scope of the claims of the present invention.

Claims

1. A composite robot CNC loading and unloading device, characterized in that: Including composite robots and CNC machine tools; The composite robot comprises a mobile chassis, a mechanical arm, a loading and unloading clamp, and a 3D camera; one end of the mechanical arm is mounted on the mobile chassis, and the other end is mounted with the loading and unloading clamp through a mounting platform at the end of the mechanical arm; the 3D camera is mounted on the mounting platform; Each of the CNC machine tools has a loading seat and a positioning identification plate of the same shape and size; the loading seat and the positioning identification plate on each of the CNC machine tools are fixed in relative position; The mobile chassis of the composite robot uses laser SLAM navigation to navigate to the designated position in front of multiple CNC machine tools; each of the CNC machine tools is provided with its own navigation site, and all navigation sites are marked on the laser SLAM navigation map. The composite robot is dispatched to the navigation site corresponding to any CNC machine tool with loading and unloading needs through the laser SLAM navigation system and the navigation sites on the navigation map.

2. A composite robot CNC loading and unloading device according to claim 1, characterized in that: A 2D camera is installed on the installation platform; the positioning identification plate is a 2D positioning identification plate; the 3D camera is used in the deployment stage of CNC machine tools and composite robots; the 2D camera combines with the 2D positioning identification plate to form posture data, and records the data through the control software, and records the loading posture of the composite robot for each CNC machine tool for use in the loading and unloading stages.

3. A composite robot CNC loading and unloading device according to claim 1, characterized in that: The positioning identification plate is a 3D positioning identification plate; the 3D camera is used in the deployment stage and the loading and unloading stage of the CNC machine tool and the composite robot; the 3D camera combines with the 3D positioning identification plate to form posture data, and records the data through the control software, and records the loading posture of the composite robot for each CNC machine tool for use in the loading and unloading stage.

4. A composite robot CNC loading and unloading method, applicable to the composite robot CNC loading and unloading device according to any one of claims 1 to 3, characterized in that: The first CNC deployment phase includes the following steps: Step 1: Get the loading and unloading posture of TCP through teaching; Step 2: Adjust the end position of the robot arm to the position for the 3D camera to take photos of the loading seat, record the adjustment amount ΔT3D1 for use in step 3, and record the coordinates ZTCP3D of the TCP for use when the 3D camera takes photos and positions other CNC machine tools during deployment; calculate and record the position T3D of the loading and unloading position in the 3D camera coordinate system; Step 3: Adjust the end position of the robot arm again to the 2D camera's photo pose for the 2D positioning mark plate, and record the TCP coordinate ZTCP2D for use when the 2D camera is taking photos and positioning during the loading and unloading stage; record it in the 2D camera's stereo coordinate system, relative to the center of the 2D positioning mark plate on the first CNC, the first optimal loading and unloading pose T2D1. The step 2 specifically includes the following: Adjust the robot arm posture so that the loading seat is within the field of view of the 3D camera and in the middle position, record the relative adjustment amount ΔT3D1 of TCP and the TCP coordinate ZTCP3D at this time; take a picture of the loading seat with a high-precision 3D camera fixed at the end of the robot arm, obtain the matching template of the loading seat and the posture T3D1 of the center point of the loading seat in the 3D camera stereo coordinate system; according to the adjustment amount of the posture of the end of the robot arm, calculate the best loading and unloading posture in the 3D camera coordinate system T3D = T3D1-ΔT3D1; The step three specifically includes the following: Adjust the end of the robot arm so that the 2D positioning identification plate is within the field of view of the 2D camera and in the middle position, record the total adjustment amount ΔT2D1 of the robot arm TCP relative to the loading and unloading posture at this time, and record the TCP coordinates at this time as ZTCP2D. The 2D camera takes a picture of the 2D positioning identification plate to obtain the posture Z2D1 of the center point of the 2D positioning identification plate in the 2D camera stereo coordinate system; obtain the first loading and unloading posture T2D1 = ΔT2D1-Z2D1 relative to the center of the 2D positioning identification plate on the first CNC in the 2D camera stereo coordinate system, and record this value for use in the loading and unloading stage.

5. A composite robot CNC loading and unloading method according to claim 4, characterized in that: It also includes the deployment stage of the nth CNC machine tool, n>1; the composite robot remains stable and motionless at the door of the nth CNC, and refers to the ZTCP3D recorded in the first deployment, adjusts the 3D camera shooting position, takes a picture of the loading seat, uses the first deployment to obtain the matching template of the loading seat, and obtains the current position T3Dn of the center point of the loading seat in the 3D camera stereo coordinate system; at the same time, refers to the ZTCP2D recorded in the first deployment, adjusts the 2D camera shooting position, records the adjustment amount ΔT2Dn relative to the robot arm TCP this time, and the 2D camera fixed at the end of the robot arm takes a picture of the 2D positioning identification plate to obtain the position Z2Dn of the center point of the 2D positioning identification plate in the 2D camera stereo coordinate system; obtains the nth loading and unloading posture T2Dn=ΔT2Dn+(T3Dn-T3D)-Z2Dn in the 2D camera stereo coordinate system relative to the center of the 2D positioning identification plate on the nth CNC, and records this value for use in the loading and unloading stage.

6. A composite robot CNC loading and unloading method according to claim 5, characterized in that: The loading and unloading stages include the following: Only use 2D camera, no 3D camera; The mobile chassis of the scheduling composite robot remains stable and motionless at the door of the mth CNC, m ≥ 1; refer to the ZTCP2D recorded in the first deployment, adjust the end of the robot arm to the 2D camera photo position, and the 2D camera fixed at the end of the robot arm takes a photo of the 2D positioning identification plate fixed on the mth CNC, and obtains the position Z2Dm of the center point of the 2D positioning identification plate in the 2D camera stereo coordinate system at the moment of this photo; read the mth loading and unloading position T2Dm relative to the center of the 2D positioning identification plate on the mth CNC recorded in the 2D camera stereo coordinate system; calculate this loading task, the relative increment that TCP needs to adjust when it arrives at the mth loading and unloading position is W2Dm = Z2Dm + T2Dm.

7. A composite robot CNC loading and unloading method, applicable to the composite robot CNC loading and unloading device according to any one of claims 1 to 3, characterized in that: The first CNC deployment phase includes the following steps: Step 1: Get the loading and unloading posture of TCP through teaching; Step 2: Adjust the end position of the robot arm to the position for the 3D camera to take photos of the loading seat, record the adjustment amount ΔT3D1 for use in step 3, and record the coordinates ZTCP3D of the TCP for use when the 3D camera takes photos and positions other CNC machine tools during deployment; calculate and record the position T3D of the loading and unloading position in the 3D camera coordinate system; Step 3: Adjust the end position of the robot arm again to the 3D camera's photo pose for the 3D positioning plate, and record the TCP coordinate ZTCP2D for use when the 3D camera is taking photos and positioning during the loading and unloading stage; record it in the 3D camera's stereo coordinate system, relative to the center of the 3D positioning plate on the first CNC, the first optimal loading and unloading pose T2D1. The step 2 specifically includes the following: Adjust the robot arm posture so that the loading seat is within the field of view of the 3D camera and in the middle position, record the relative adjustment amount ΔT3D1 of TCP and the TCP coordinate ZTCP3D at this time; take a picture of the loading seat with a high-precision 3D camera fixed at the end of the robot arm, obtain the matching template of the loading seat and the posture T3D1 of the center point of the loading seat in the 3D camera stereo coordinate system; according to the adjustment amount of the posture of the end of the robot arm, calculate the best loading and unloading posture in the 3D camera coordinate system T3D = T3D1-ΔT3D1; The step three specifically includes the following: Adjust the end of the robot arm so that the 3D positioning identification plate is within the field of view of the 3D camera and in the middle position, record the total adjustment amount ΔT2D1 of the robot arm TCP relative to the optimal loading and unloading posture at this time, and record the TCP coordinates at this time as ZTCP2D. The 3D camera takes a picture of the 3D positioning identification plate to obtain the posture Z2D1 of the center point of the 3D positioning identification plate in the 3D camera stereo coordinate system; obtain the first optimal loading and unloading posture T2D1 = ΔT2D1-Z2D1 relative to the center of the 3D positioning identification plate on the first CNC in the 3D camera stereo coordinate system, and record this value for use in the loading and unloading stage.

8. A composite robot CNC loading and unloading method according to claim 8, characterized in that: The loading and unloading stages include the following: It also includes the deployment stage of the nth CNC machine tool, n>1; the composite robot remains stable and motionless at the door of the nth CNC, and refers to the ZTCP3D recorded in the first deployment, adjusts the 3D camera shooting position, takes a picture of the loading seat, uses the first deployment to obtain the matching template of the loading seat, and obtains the current position T3Dn of the center point of the loading seat in the 3D camera stereo coordinate system; at the same time, refers to the ZTCP2D recorded in the first deployment, adjusts the 3D camera shooting position, records the adjustment amount ΔT2Dn relative to the robot arm TCP, and the 3D camera fixed at the end of the robot arm takes a picture of the 3D positioning identification plate to obtain the position Z2Dn of the center point of the 3D positioning identification plate in the 3D camera stereo coordinate system; obtains the nth loading and unloading posture T2Dn=ΔT2Dn+(T3Dn-T3D)-Z2Dn in the 3D camera stereo coordinate system relative to the center of the 3D positioning identification plate on the nth CNC, and records this value for use in the loading and unloading stage.

9. A composite robot CNC loading and unloading method according to claim 8, characterized in that: The loading and unloading stages include the following: The mobile chassis of the scheduling composite robot remains stable and motionless at the door of the mth CNC, m≥1; refer to the ZTCP2D recorded in the first deployment, adjust the end of the robotic arm to the 3D camera shooting position, and the 3D camera fixed at the end of the robotic arm takes a picture of the 3D positioning identification plate fixed on the mth CNC, and obtains the position Z2Dm of the center point of the 3D positioning identification plate in the 3D camera stereo coordinate system at the moment of this photo shooting; read the mth loading and unloading position T2Dm relative to the center of the 3D positioning identification plate on the mth CNC recorded in the 3D camera stereo coordinate system; calculate this loading task, the relative increment that TCP needs to adjust to reach the mth optimal loading and unloading position is W2Dm=Z2Dm+T2Dm.

10. A composite robot CNC loading and unloading method according to any one of claims 5 or 8, characterized in that: In the first CNC machine tool deployment stage and the nth CNC machine tool deployment stage, after determining the TCP loading and unloading posture, the position of the current composite robot mobile chassis in the laser SLAM navigation system is calibrated and recorded.

Citation Information

Patent Citations

  • Robot 3D vision hand-eye calibration method

    CN112123329A

  • Automatic feeding and discharging method and system on assembly line based on machine vision

    CN112456128A

  • Visual positioning method of high-environment self-adaptive mobile composite robot

    CN119260801A

  • Cell -phone shell CNC goes up unloading system

    CN207900753U

  • Humanoid inspection operation method and system for semantic intelligent substation robot

    WO2022021739A1

Cited By

  • Robot CNC feeding and discharging method and device

    CN120606284A

  • Industrial robot control method and system based on deep reinforcement learning

    CN121374604A

  • Industrial robot control method and system based on deep reinforcement learning

    CN121374604B