Mechanical arm-based pipeline bending process adaptive feeding and discharging system

Abstract

The application discloses a pipeline bending process adaptive feeding and discharging system based on a mechanical arm, which comprises a binocular vision detection system, the binocular vision detection system is used for recognizing the posture and position of the pipeline; a mechanical arm, the mechanical arm is used for driving a flexible clamping mechanism to move; the flexible clamping mechanism, the flexible clamping mechanism is arranged on the mechanical arm, and the flexible clamping mechanism is used for clamping the pipe; a mechanical arm end force control system, the mechanical arm end force control system is used for touching the pipe and inhibiting pipe vibration; a controller, the controller is used for controlling the binocular vision detection system, the mechanical arm, the flexible clamping mechanism, the mechanical arm end force control system and used for calculating the position of a grabbing point, determining a space posture and performing trajectory planning. The pipeline bending process mechanical arm has the advantages of flexibility, adaptive feeding and discharging, reduction of clamp replacement and path planning time waste, and great improvement of production efficiency in a multi-variety and small-batch production mode.

Description

Technical Field

[0001] This invention relates to the field of mechanical manufacturing and loading / unloading systems, and more specifically, to an adaptive loading / unloading system for pipe bending processes based on a robotic arm. Background Technology

[0002] Currently, automated pipe loading and unloading processes primarily involve equipping robotic arms with fixed-size rigid fixtures and employing traditional on-site teaching or offline programming path planning methods. However, given the diverse pipe types, small batch sizes, complex spatial structures, and high precision requirements of aerospace products, fixed-size rigid fixtures are incompatible with various pipe specifications, necessitating frequent fixture changes. Furthermore, pipe bending results in some vibration due to inertia, requiring precise clamping maneuvers. Additionally, traditional path planning methods require manual programming for each pipe, leading to low efficiency and high labor costs.

[0003] Currently, the robotic arm's loading and unloading path planning in the pipe bending process mainly relies on on-site manual teaching. Given the characteristics of aerospace products—various pipe types, small batches, complex spatial structures, and high precision requirements—the following problems exist: ① On-site manual teaching causes production line shutdowns and requires skilled operators; ② Path teaching is more suitable for large-volume, simple, repetitive product production, and is difficult to meet the current demand for multi-variety, small-batch production; ③ For products with complex shapes and large dimensions, the teaching programming process is complex, time-consuming, inefficient, and labor-intensive; ④ Human error during teaching can lead to collisions between the robot and products / equipment, causing equipment damage and even threatening personnel safety; ⑤ Manual teaching relies mainly on visual inspection, which introduces errors and deviations from the target; ⑥ After pipe bending, there is vibration, requiring a waiting period before clamping.

[0004] In summary, at least one of the following technical problems exists:

[0005] On-site manual teaching can cause the production line to stop, and it also requires certain qualifications from the operators.

[0006] Path teaching is more suitable for mass production of simple, repetitive products, and it is difficult to meet the current production needs of multiple varieties and small batches.

[0007] For products with complex shapes and large dimensions, the teaching and programming process is complicated, time-consuming, has low production efficiency, and high labor intensity for personnel.

[0008] During the teaching process, human error may occur, leading to collisions between the robot and products or equipment, which could damage the equipment or even threaten the safety of personnel.

[0009] The manual teaching process relies mainly on visual inspection, which can introduce certain errors and cause deviations from the target.

[0010] After the pipe is bent, there will be some vibration, and you need to wait for a certain period of time before clamping. Summary of the Invention

[0011] The main objective of this invention is to provide an adaptive loading and unloading system for pipe bending based on a robotic arm. This system addresses several technical issues in existing technologies, including: manual teaching processes causing production line stoppages and requiring skilled operators; path teaching being more suitable for large-volume, simple, repetitive product production, which is insufficient for meeting the current demand for multi-variety, small-batch production; complex programming processes for products with large shapes and dimensions, resulting in low production efficiency and high labor intensity; human error during teaching leading to collisions between the robot and products / equipment, causing equipment damage and even threatening personnel safety; visual inspection-based manual teaching, which introduces errors and deviations from the target; and the need for a waiting period after pipe bending to allow for clamping.

[0012] To achieve the above objectives, according to one aspect of the present invention, an adaptive loading and unloading system for pipe bending processes based on a robotic arm is provided, comprising:

[0013] A binocular vision inspection system, used to identify the posture and position of pipelines;

[0014] A robotic arm, used to move a flexible gripping mechanism;

[0015] A flexible clamping mechanism is mounted on a robotic arm and is used to clamp pipe fittings.

[0016] A force control system at the end of a robotic arm, used to lightly touch the pipe fitting and suppress pipe vibration;

[0017] The controller is used to control the binocular vision detection system, the robotic arm, the flexible gripping mechanism, the robotic arm end force control system, and to calculate the position of the gripping point to determine the spatial posture and perform trajectory planning.

[0018] Preferably, the flexible clamping mechanism employs a flexible gripper that inflates and bends to adaptively cover the target object for grasping; and deflates and deforms in the opposite direction for placement or pre-grabbing.

[0019] Preferably, the controller includes a central processing unit, a binocular vision detection system control module, a robotic arm control module, a flexible gripping mechanism control module, a robotic arm end effector force control system control module, a data calculation module, and a planning module. The central processing unit is connected to the binocular vision detection system control module, the robotic arm control module, the flexible gripping mechanism control module, the robotic arm end effector force control system control module, the data calculation module, and the planning module, respectively.

[0020] Preferably, the binocular vision detection system control module is connected to the binocular vision detection system, the robotic arm control module is connected to the robotic arm, the flexible clamping mechanism control module is connected to the flexible clamping mechanism, and the robotic arm end effector force control system control module is connected to the robotic arm end effector force control system.

[0021] Preferably, the binocular vision detection system is used to acquire pipe size, depth information and surface features, and to collect information on the surrounding working environment and the distribution of obstacles. The binocular vision detection system transmits the collected information to the controller, and the controller performs spatial modeling of the working environment.

[0022] Preferably, the data calculation module of the controller is used to calculate the gripping position, establish a digital model of the pipe fitting through pre-processed data, clarify the position of the centroid of the pipeline, and make adaptive adjustments based on the position of the centroid.

[0023] Preferably, the controller filters the image data using depth information, saving only features within a threshold range and filtering out complex background information. It performs binarization and image filtering preprocessing on the bent pipe image, and uses a contour extraction algorithm to extract the contour features of the bent pipe. Then, based on the captured partial features and dimensions, it compares them with the bent pipe model in the digital model, maps the calculated gripping point position to the robot's base coordinate system, and determines its spatial pose information.

[0024] Preferably, the controller performs spatial modeling of the working environment based on robot accessibility, obstacle avoidance constraints, motion smoothness and efficiency, obtains obstacle distribution information, and generates the robot's motion trajectory in combination with a trajectory generation algorithm.

[0025] Preferably, the controller establishes a digital model of the pipeline using pre-processed data. Taking the center P0 of the straight pipe as the dividing point, the straight pipe is divided into left and right sides. Combining the known pipeline model processing data information, the controller calculates the centroid positions P1 and P2 of the bends on both sides of the center point P0 of the straight pipe. Points P1 and P2 are connected in space, and the midpoint P3 of the line connecting them is taken as the centroid of the bend. A plane perpendicular to the line connecting P1 and P2 is drawn through point P3. This plane will intersect the bend at point P4, which can be used as the bend gripping point. When the gripping point falls exactly at the bend, the robot cannot stably grip the bend. The gripping point needs to be translated to both sides along the axial direction of the bend. The translation distance is the sum of the bend radius of the pipeline and the width of the gripper.

[0026] Preferably, the end effector force control system is installed on the side of the robotic arm and adopts a structure with a force sensor mounted on a flexible material. The end effector force control system uses the sensor's detection information as a force feedback signal and then converts it into a control signal for the robot. When executing force control motion commands, the controller adjusts the robot's motion state and trajectory according to the magnitude and direction of the force detected by the force sensor, so that the robotic arm can suppress vibrations in the pipeline and avoid rigid collisions with the pipeline.

[0027] The technical solution of this invention has the following technical effects:

[0028] This invention enables flexible and automated loading and unloading of a robotic arm during pipe bending. The adaptive pipe loading and unloading system includes a flexible gripping mechanism at the end of the robotic arm, spatial pose mapping and trajectory planning for the gripping position, gripping position calculation, and a vibration suppression system for the gripping process. The flexible gripping mechanism employs a pneumatic flexible gripper. Inflation achieves bending deformation, adaptively enveloping the target object to complete the gripping action; deflating achieves reverse deformation to complete the placement action. Spatial pose mapping and trajectory planning for the gripping position utilizes a binocular vision detection system at the robot's end effector to simultaneously detect pipe dimensions, extract depth information and surface features, spatially model the working environment, obtain obstacle distribution information, and generate the robot's motion trajectory using algorithms. Gripping position calculation uses pre-processed data to establish a digital model of the pipe, clarifying the pipe's center of gravity position, and making adaptive adjustments based on the center of gravity position. The vibration suppression method for the gripping process uses a force control system at the end effector to lightly touch the pipe, quickly stopping pipe vibration. It enables the robotic arm to be flexible and adaptive in the pipe bending process, reducing the time wasted on fixture replacement and path planning, greatly improving production efficiency in multi-variety, small-batch production modes, and has wide applicability. Attached Figure Description

[0029] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0030] Figure 1 A schematic diagram of the spatial visual recognition system for the bending gripping point pose of the adaptive loading and unloading system for pipe bending process based on a robotic arm according to the present invention is shown.

[0031] Figure 2 It shows Figure 1 The gripping position calculation view of the adaptive loading and unloading system based on a robotic arm in the pipe bending process;

[0032] Figure 3 It shows Figure 1 A position / force hybrid control view of an adaptive loading and unloading system for pipe bending based on a robotic arm;

[0033] Figure 4 It shows Figure 1 A view of the flexible gripper structure of an adaptive loading and unloading system for pipe bending based on a robotic arm.

[0034] The above figures include the following reference numerals:

[0035] 1. Flexible gripper; 2. Bend; 3. Binocular vision inspection system. Detailed Implementation

[0036] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0037] like Figures 1 to 4 As shown, this embodiment of the invention provides an adaptive loading and unloading system for pipe bending processes based on a robotic arm, comprising: a binocular vision detection system 3 for identifying the posture and position of the pipe; a robotic arm for moving a flexible clamping mechanism; a flexible clamping mechanism mounted on the robotic arm for clamping the pipe; a robotic arm end effector force control system for lightly touching the pipe and suppressing pipe vibration; and a controller for controlling the binocular vision detection system 3, the robotic arm, the flexible clamping mechanism, the robotic arm end effector force control system, and for calculating the gripping point position to determine the spatial posture and performing trajectory planning.

[0038] In this embodiment, a binocular vision inspection system 3 is used to identify the posture and position of the pipeline. A control module for the binocular vision inspection system 3 is connected to the system, as are a robotic arm control module, a flexible clamping mechanism control module, and a robotic arm end effector force control system module. The binocular vision inspection system 3 is used to acquire pipe dimensions, depth information, and surface features, and to collect information about the surrounding working environment, including obstacle distribution. The system transmits the collected information to a controller, which then performs spatial modeling of the working environment.

[0039] In this embodiment, the robotic arm is used to drive the flexible clamping mechanism to move;

[0040] In this embodiment, a flexible clamping mechanism is mounted on the robotic arm and is used to clamp the pipe fitting. The flexible clamping mechanism uses a flexible gripper 1 to be activated, which inflates and bends to adaptively cover the target object and perform a gripping action. It then deflates and deforms in the opposite direction to place the object.

[0041] In this embodiment, the end effector force control system is used to lightly touch the pipe and suppress its vibration. The end effector force control system is installed on the side of the robotic arm and adopts a structure with a force sensor mounted on a flexible material. The end effector force control system uses the sensor's detection information as a force feedback signal and then converts it into a control signal for the robot. When executing force control motion commands, the controller adjusts the robot's motion state and trajectory according to the magnitude and direction of the force detected by the force sensor, so that the robotic arm can suppress vibration of the pipe and avoid rigid collision with the pipe.

[0042] In this embodiment, the controller is used to control the binocular vision detection system 3, the robotic arm, the flexible gripping mechanism, the robotic arm end effector force control system, and to calculate the gripping point position to determine the spatial posture and perform trajectory planning. The controller includes a central processing unit (CPU), a control module for the binocular vision detection system 3, a control module for the robotic arm, a control module for the flexible gripping mechanism, a control module for the robotic arm end effector force control system, a data calculation module, and a planning module. The CPU is connected to each of these modules. The controller's data calculation module calculates the gripping position, establishes a digital model of the pipe fitting using pre-processed data, clarifies the pipe's centroid position, and makes adaptive adjustments based on the centroid position. The controller filters image data using depth information, saving only features within a threshold range and filtering out complex background information. It performs binarization and image filtering preprocessing on the image of bend 2, and uses a contour extraction algorithm to extract the contour features of bend 2. Then, based on the captured partial features and dimensions, it compares them with the digital model of bend 2, mapping the calculated gripping point position onto the robot's base coordinate system to determine its spatial pose information. The controller performs spatial modeling of the working environment based on robot reachability, obstacle avoidance constraints, motion smoothness, and efficiency, acquiring obstacle distribution information, and combining this with a trajectory generation algorithm to generate the robot's motion trajectory. The controller establishes a digital model of the pipeline using pre-processed data. Taking the center P0 of the straight pipe as the dividing point, the straight pipe is divided into left and right sides. Combining the known pipeline model processing data, the controller calculates the centroid positions P1 and P2 of the bend 2 on both sides of the center point P0 of the straight pipe. Points P1 and P2 are connected in space, and the midpoint P3 of the line connecting them is taken as the centroid of the bend 2. A plane perpendicular to the line connecting P1 and P2 is drawn through point P3. This plane will intersect the bend 2 at point P4, which can be used as the gripping point of the bend 2. When the gripping point falls exactly on the bend, the robot cannot stably grip the bend 2. The gripping point needs to be translated to both sides along the axial direction of the bend 2. The translation distance is the sum of the bend radius of the pipeline and the width of the gripper.

[0043] Specifically, a binocular vision inspection system 3 is established at the robot's end effector to simultaneously perform dimensional detection, depth information extraction, and surface feature extraction of the bent pipe 2, such as... Figure 1As shown. Since the position and height of the clamping device for bending pipe 2 are fixed, the robot always takes pictures of bending pipe 2 with a fixed posture and the same shooting height. Then, the image data is filtered using depth information, saving only features within a threshold range and filtering out complex background information. Preprocessing operations such as binarization and image filtering are performed on the images of bending pipe 2. A contour extraction algorithm is used to extract the contour features of bending pipe 2. Then, based on the captured partial features and dimensions, the model of bending pipe 2 in the digital model is compared, and the calculated gripping point position is mapped to the robot's base coordinate system to determine its spatial pose information. Trajectory planning method: Spatial modeling of the working environment is performed to obtain obstacle distribution information, and the robot's motion trajectory is generated using an algorithm. During the planning process, factors such as robot accessibility, obstacle avoidance constraints, motion smoothness, and efficiency are comprehensively considered to ensure the optimization and safety of the motion path.

[0044] Specifically, the gripping position calculation involves establishing a digital model of the pipeline using pre-processed data. The straight pipe center P0 is used as the dividing point, dividing the straight pipe into left and right sides. Combined with known pipeline model processing data, the centroid positions P1 and P2 of the bend 2 on both sides of P0 are calculated. Points P1 and P2 are connected in space, and the midpoint P3 of this line is taken as the centroid of bend 2. A plane perpendicular to the line connecting P1 and P2 is drawn through point P3. This plane intersects bend 2 at point P4, which can then be used as the gripping point for bend 2. Figure 2 As shown. When the gripping point falls exactly on the bend, the robot cannot stably grip the bend 2. Therefore, the gripping point needs to be translated to both sides along the axial direction of the bend 2. The translation distance is the sum of the bend radius of the pipe and the width of the gripper.

[0045] Specifically, regarding vibration suppression during the grasping process, after the pipe is bent, it will continue to vibrate due to bending inertia. Therefore, vibration suppression is needed before the robotic arm grasps the pipe to ensure stable grasping. A force control system will be installed on the robotic arm for vibration suppression. The vibration suppression mechanism is installed on the side of the robotic arm, using a flexible material to mount force sensors. The sensor's detection information is used as a force feedback signal, which is then converted into a control signal for the robot. When executing force-controlled motion commands, the robot controller can adjust the robot's motion state and trajectory based on the magnitude and direction of the force detected by the force sensor, enabling the robotic arm to suppress vibrations in the pipe without rigidly colliding with it. The robot force control strategy adopts a position / force hybrid control method. Based on the Jacobian matrix in the Cartesian coordinate system, the force and position in the robot's workspace are distributed to each joint of the robot. Its control structure is as follows: Figure 4 As shown, by selecting matrix C, the robot's motion space is decomposed into a force control space and a position control space. Only one control mode is assigned to each joint, realizing separate force and position control of the entire robot. The two control loops operate independently without interfering with each other.

[0046] Specifically, the flexible clamping mechanism can be selected according to the pipe specifications, using flexible grippers of appropriate size 1 rack products, as shown in the schematic diagram. Figure 4 As shown, this innovative flexible gripper mimics a starfish-like design and simulates human hand grasping. The flexible gripper 1 has a special air bladder structure inside, which inflates to achieve bending deformation, adaptively enveloping the target object to complete the grasping action; deflating it achieves reverse deformation to complete the placement action. It has good compatibility, supports a wide variety of items, and its gentle movements make it suitable for grasping most objects. The design of a robotic arm-based adaptive loading and unloading system addresses the low efficiency problem of traditional robotic arm automatic loading and unloading in multi-variety, small-batch production. The robotic arm-based adaptive loading and unloading system design includes flexible gripping at the robotic arm end effector, spatial pose mapping and trajectory planning for the grasping position, grasping position calculation, and vibration suppression during the grasping process, achieving adaptive control throughout the entire process.

[0047] As can be seen from the above description, the embodiments of the present invention achieve the following technical effects:

[0048] This invention enables flexible and automated loading and unloading of a robotic arm during pipe bending. The adaptive pipe loading and unloading system includes a flexible gripping mechanism at the end of the robotic arm, spatial pose mapping and trajectory planning for the gripping position, gripping position calculation, and a vibration suppression system for the gripping process. The flexible gripping mechanism employs a pneumatic flexible gripper 1, which achieves bending deformation through inflation, adaptively enveloping the target object to complete the gripping action; it also achieves reverse deformation through deflating to complete the placement action. Spatial pose mapping and trajectory planning for the gripping position utilizes a binocular vision detection system 3 established at the robot's end effector to simultaneously detect the dimensions, depth information, and surface features of the bent pipe 2, perform spatial modeling of the working environment, obtain obstacle distribution information, and generate the robot's motion trajectory using algorithms. Gripping position calculation uses pre-processed data to establish a digital model of the pipe, clarify the pipe's center of gravity position, and make adaptive adjustments based on the center of gravity position. The vibration suppression method for the gripping process uses a force control system at the end effector of the robotic arm to lightly touch the pipe, quickly stopping pipe vibration. It enables the robotic arm to be flexible and adaptive in the pipe bending process, reducing the time wasted on fixture replacement and path planning, greatly improving production efficiency in multi-variety, small-batch production modes, and has wide applicability.

[0049] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An adaptive loading and unloading system for pipe bending process based on a robotic arm, characterized in that, include: A binocular vision inspection system, used to identify pipeline posture and position; A robotic arm, used to move a flexible gripping mechanism; A flexible clamping mechanism is mounted on a robotic arm and is used to clamp pipe fittings. A force control system at the end of a robotic arm is used to lightly touch the pipe fitting and suppress vibration of the pipe fitting after bending. The controller controls the binocular vision detection system, robotic arm, flexible gripping mechanism, and robotic arm end effector force control system. It also calculates the gripping point position to determine spatial posture and performs trajectory planning. The controller establishes a digital model of the pipeline using pre-processed data. Using the center P0 of the straight pipe as the dividing point, it divides the straight pipe into left and right sides. Combining this with known pipeline model processing data, it calculates the centroid positions P1 and P2 of the bends on both sides of the center P0. It connects P1 and P2 in space and takes the midpoint P3 of this line as the centroid of the bend. A plane perpendicular to the line connecting P1 and P2 is drawn through P3. This plane intersects the bend at point P4, which serves as the gripping point. When the gripping point falls precisely at the bend, the robot cannot stably grip the bend. Therefore, the gripping point is translated laterally along the axial direction of the bend, with the translation distance being the sum of the bend radius and the gripper width. The end effector force control system is installed on the side of the robotic arm and uses a flexible material to mount force sensors. The end effector force control system uses the sensor's detection information as a force feedback signal and then converts it into a control signal for the robot. When executing force control motion commands, the controller adjusts the robot's motion state and trajectory according to the magnitude and direction of the force detected by the force control sensor, so that the robotic arm can suppress vibrations in the pipeline and avoid rigid collisions with the pipeline.

2. The adaptive loading and unloading system for pipe bending process based on a robotic arm as described in claim 1, characterized in that, The flexible clamping mechanism employs a flexible gripper that inflates and bends to adaptively envelop the target object for grasping; it then deflates and deforms in the opposite direction to place the object.

3. The adaptive loading and unloading system for pipe bending process based on a robotic arm as described in claim 1, characterized in that, The controller includes a central processing unit, a binocular vision detection system control module, a robotic arm control module, a flexible gripping mechanism control module, a robotic arm end effector force control system control module, a data calculation module, and a planning module. The central processing unit is connected to the binocular vision detection system control module, the robotic arm control module, the flexible gripping mechanism control module, the robotic arm end effector force control system control module, the data calculation module, and the planning module, respectively.

4. The adaptive loading and unloading system for pipe bending process based on a robotic arm as described in claim 3, characterized in that, The binocular vision detection system control module is connected to the binocular vision detection system, the robotic arm control module is connected to the robotic arm, the flexible clamping mechanism control module is connected to the flexible clamping mechanism, and the robotic arm end effector force control system control module is connected to the robotic arm end effector force control system.

5. The adaptive loading and unloading system for pipe bending process based on a robotic arm as described in claim 1, characterized in that, The binocular vision inspection system is used to acquire pipe size, depth information and surface features, and to collect information on the surrounding working environment and the distribution of obstacles. The binocular vision inspection system transmits the collected information to the controller, which performs spatial modeling of the working environment.

6. The adaptive loading and unloading system for pipe bending process based on a robotic arm as described in claim 1, characterized in that, The controller's data calculation module is used to calculate the gripping position, establish a digital model of the pipe fitting through pre-processed data, clarify the position of the pipe's centroid, and make adaptive adjustments based on the centroid position.

7. The adaptive loading and unloading system for pipe bending process based on a robotic arm as described in claim 1, characterized in that, The controller filters image data using depth information, saving only features within a threshold range and filtering out complex background information. It performs binarization and image filtering preprocessing on the bent pipe image, and uses a contour extraction algorithm to extract the contour features of the bent pipe. Then, based on the captured partial features and dimensions, it compares them with the bent pipe model in the digital model, maps the calculated gripping point position to the robot's base coordinate system, and determines its spatial pose information.

8. The adaptive loading and unloading system for pipe bending process based on a robotic arm as described in claim 1, characterized in that, The controller performs spatial modeling of the working environment based on robot accessibility, obstacle avoidance constraints, motion smoothness and efficiency, obtains obstacle distribution information, and generates the robot's motion trajectory in combination with trajectory generation algorithms.

Images