Bending machine control system and method

Abstract

The invention discloses a bending machine control system and method, the control system comprises a bending machine, an acquisition processing module, a preset process module and a control module, and an upper mold and a lower mold are both associated with corresponding mold information identifiers; the collecting and processing module is used for collecting mold information identification, real-time images and down force, the preset procedure module comprises a preset procedure configuration unit for a user to input multiple pieces of bending procedure information, and the multiple pieces of bending procedure information comprise procedure numbers, bending positions, mold information, bending pressure preset numerical values and procedure standard images. The control module comprises an image recognition unit, an image comparison unit and a control unit, mold information, a workpiece image and bending pressure are collected in real time through the collecting and processing module, the mold information and the workpiece state are automatically compared in combination with standardized configuration of the preset process module, and automatic control over the bending process is achieved; the bending machine has the advantages of improving the bending efficiency, reducing operation errors and improving the workpiece bending precision and consistency.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical processing equipment control, and particularly to a bending machine control system and method. Background Art

[0002] A bending machine is a machine that can bend thin plates. Generally, it is provided with an upper die holder, a lower die holder, a backstop, and a downward pressing assembly for driving the upper die holder to move. Most of the downward pressing assemblies use hydraulic equipment. The upper die holder and the lower die holder are both detachably provided with an upper die and a lower die. By replacing different upper and lower dies, different shapes can be bent. Currently, in order to facilitate operation, multiple different dies can generally be placed on both the upper die holder and the lower die holder.

[0003] The existing bending machines have obvious deficiencies in terms of bending efficiency. Especially when processing workpieces that require multiple bends, the operator needs to frequently change the dies and is prone to placing the workpiece in the wrong die position, resulting in workpiece scrapping or die damage. In addition, since different dies are required for multiple bending processes, the operator must repeatedly confirm whether each process corresponds to the die during the bending operation, which not only increases the complexity of the operation but also significantly reduces the bending efficiency. Manual operation is not only error-prone but also difficult to ensure the bending accuracy and consistency of the workpiece. Therefore, the prior art urgently needs a bending machine control system that can automatically identify the dies, monitor the bending process in real time, and automatically adjust the bending parameters to improve the bending efficiency and accuracy and reduce the risk of operation errors and die damage. Summary of the Invention

[0004] The purpose of the present application is to provide a bending machine control system and control method, which have the advantages of improving the bending efficiency, reducing operation errors, and enhancing the bending accuracy and consistency of the workpiece.

[0005] To solve the above technical problems, the present invention is solved by the following technical solutions: A bending machine control system includes a bending machine, on which an upper die holder, a lower die holder, and a downward pressing assembly for driving the upper die holder to move towards the lower die holder to bend a workpiece are provided. An upper die is installed on the upper die holder, and a lower die is installed on the lower die holder. It further includes an acquisition and processing module, a preset process module, and a control module. Corresponding die information identifiers are associated with both the upper die and the lower die; the acquisition and processing module is connected to the bending machine and is used to acquire the die information identifiers of the current upper die and lower die of the bending machine, the real-time image in the working state of the bending machine, and the downward pressure of the downward pressing assembly, so as to obtain the working die information, the real-time image, and the bending pressure value respectively; the preset process module includes a preset process configuration unit, and the preset process configuration unit is used for the user to input a number of bending process information. The number of bending process information all includes a process number, a bending position, die information, a preset value of the bending pressure, and a process standard image. The control module includes: an image recognition unit, connected to the acquisition and processing module and the preset process module, for identifying the information codes of the upper die and the lower die by the acquisition and processing module and comparing the die information in the bending process information; an image comparison unit, connected to the preset process module and the acquisition and processing module, for comparing the workpiece image after the completion of the process intercepted from the real-time image with the process standard image in the bending process information; a control unit, connected to the preset process module, for driving the bending machine to bend the workpiece according to the preset process configuration unit according to the preset process.

[0006] The present invention is further configured as: The acquisition and processing module includes a workpiece position acquisition unit, a die identifier acquisition unit, and a pressure acquisition unit. A backstop is provided on the bending machine. The workpiece position acquisition unit includes a workpiece movement acquisition subunit and a position-in-place acquisition subunit. The workpiece movement acquisition subunit is used for acquiring the movement position of the workpiece and generating a real-time image of the workpiece; the position-in-place acquisition subunit is used for acquiring whether the workpiece is in contact with the backstop and generating workpiece position-in-place information; the die identifier acquisition unit is used for identifying the die information identifiers of the upper die and the lower die and generating the working die information; the pressure acquisition unit is used for acquiring the downward pressure of the downward pressing assembly on the workpiece to obtain the bending pressure value.

[0007] The present invention is further configured as: The bending machine is equipped with a robotic arm, and the robotic arm is provided with a mechanical gripper for picking up the workpiece. The workpiece movement acquisition subunit is arranged on the bending machine, and the die identifier acquisition unit is arranged on the mechanical gripper.

[0008] The present invention is further configured such that: the preset process module further includes a manual process input unit, and the manual process input unit is used for a user to manually operate the robotic arm. The manual process input unit includes a manual trajectory generation sub-unit, a manual process confirmation sub-unit, and a manual process configuration sub-unit. The manual trajectory generation sub-unit is used to generate a workpiece movement trajectory based on the edge of the workpiece in the real-time image of the workpiece, and generate a gripper movement trajectory according to the state of the workpiece bending that drives the gripper to move when the robotic arm drives the mechanical gripper and the bending machine to bend the workpiece; the manual process confirmation sub-unit is used for the user to confirm the workpiece movement trajectory and the gripper movement trajectory; the manual process configuration sub-unit generates a number of bending process information according to the trajectory optimization sub-unit and the manual process confirmation unit. The bending process information includes the bending position, the process standard image, the die information generated according to the acquisition and processing module, as well as the automatically generated process number and the preset bending pressure value generated according to the bending pressure value.

[0009] The present invention is further configured such that: the manual process input unit further includes a coordinate generation sub-unit and a trajectory optimization sub-unit. The coordinate generation sub-unit is connected to the manual trajectory generation sub-unit and generates a starting coordinate point, a transfer coordinate point, a bending coordinate point, and an ending coordinate point according to the manual trajectory generation sub-unit. The trajectory optimization sub-unit is connected to the coordinate generation sub-unit and generates the shortest optimized path in sequence according to the starting coordinate point, the transfer coordinate point, the bending coordinate point, and the ending coordinate point.

[0010] The present invention is further configured such that: the ending coordinate point is the starting coordinate point of the next process.

[0011] The present invention is further configured such that: the workpiece position acquisition unit further includes a workpiece bending acquisition sub-unit. The workpiece bending acquisition sub-unit is arranged on the side of the workpiece to detect the front view of the workpiece. The workpiece bending acquisition sub-unit is used to acquire the bending state of the workpiece to generate a real-time front view image. The image recognition unit is used to extract key frames from the real-time front view image. When the workpiece in-place information is generated, the image recognition unit generates the in-place key frame. When the pressure acquisition unit acquires a stable bending pressure value, the image recognition unit generates the bending-in-place key frame; The image comparison unit compares the key frames generated by the image recognition with the process standard images of the corresponding bending processes.

[0012] The present invention is further configured such that: the coordinate generation sub-unit generates an avoidance area based on the front view images of the upper die and the lower die on the real-time front view image of the workpiece bending acquisition sub-unit. The coordinate generation sub-unit generates avoidance coordinate points according to the avoidance area. The avoidance coordinate points are arranged between the starting coordinate point and the bending coordinate point, and the avoidance coordinate points are also arranged between the bending coordinate point and the ending coordinate point.

[0013] A bending machine control method is applied to the bending machine control system. The bending machine control method includes: Step S1, a preset process module presets the bending process of the workpiece. The preset process module is used for the user to input a number of bending process information, and the number of bending process information all includes a process number, a bending position, die information, a preset value of bending pressure, and a process standard image; Step S2, the control unit controls the bending machine and the robotic arm to bend the workpiece according to the process based on the preset bending process information; The die identification acquisition unit is used to identify the die information identification of the upper die and the lower die, and generate working die information; The workpiece position acquisition unit is used to generate a real-time image of the bending state of the workpiece; The pressure acquisition unit is used to acquire the downward pressure of the downward pressing component on the workpiece to obtain the bending pressure value; Step S4, the image recognition unit recognizes the information codes of the upper die and the lower die according to the acquisition and processing module and compares the die information in the bending process information; The image recognition unit is used to extract key frames from the real-time image, and the image comparison sub-unit compares the key frames generated by the image recognition with the process standard images of the corresponding bending processes to detect whether the workpieces in each bending process meet the standards.

[0014] The present invention is further configured as: Step S1 further includes a manual process input unit. The manual process input unit includes a manual trajectory generation sub-unit, a coordinate generation sub-unit, a trajectory optimization sub-unit, a manual process confirmation sub-unit, and a manual process configuration sub-unit. The manual trajectory generation sub-unit is used to generate the workpiece movement trajectory and the bending state jaw movement trajectory of the mechanical jaw when the bending machine bends the workpiece according to the edge of the workpiece in the real-time image; the coordinate generation sub-unit is connected to the manual trajectory generation sub-unit and generates a starting coordinate point, a transfer coordinate point, a bending coordinate point, and an end coordinate point according to the manual trajectory generation sub-unit. The trajectory optimization sub-unit is connected to the coordinate generation sub-unit and sequentially generates the shortest optimized path according to the starting coordinate point, the transfer coordinate point, the bending coordinate point, and the end coordinate point. The manual process confirmation sub-unit is used for the user to confirm the workpiece movement trajectory and the jaw movement trajectory; the manual process configuration sub-unit generates a number of bending process information according to the trajectory optimization sub-unit and the manual process confirmation unit. The bending process information includes the bending position, the process standard image, the die information generated according to the acquisition and processing module, as well as the automatically generated process number and the preset value of the bending pressure generated according to the bending pressure value.

[0015] A bending machine control system and control method provided by the present application collect die information, workpiece images, and bending pressure in real time through a collection and processing module. Combining with the standardized configuration of a preset process module, it automatically compares the die information and the workpiece status to achieve automatic control of the bending process, having the advantages of improving bending efficiency, reducing operation errors, and enhancing the bending accuracy and consistency of workpieces. Description of the Drawings

[0016] Figure 1 It is a schematic structural diagram of the bending machine control system in the embodiment. Detailed Embodiment

[0017] The present invention will be further described in detail below in conjunction with the drawings and embodiments.

[0018] Embodiment: A bending machine control system includes a bending machine, a collection and processing module, a preset process module, and a control module. The bending machine is configured with a backstop, an upper die holder, a lower die holder, and a downward pressing component for driving the upper die holder to move towards the lower die holder. The upper die is detachably installed on the upper die holder, and the lower die is detachably installed on the lower die holder. Both the upper die and the lower die are associated with corresponding die information identifiers. The collection and processing module obtains the working die information by identifying the die information identifiers, and at the same time collects the bending pressure value and real-time images. The preset process module stores process information including process numbers, die information, preset bending pressure values, and process standard images. The control module compares the die information through image recognition, verifies the workpiece shape through image comparison, and finally controls the bending action according to the preset process.

[0019] The die information identifier refers to an identifiable coding carrier attached to the die surface, which can be realized by a style or radio frequency chip with an identifying function such as a shape or pattern or text or two-dimensional code label or bar code label or a combination of the former several. The die model parameters are read through an optical scanning or radio frequency identification device. Among them, the process standard image refers to a standard morphological diagram of the workpiece after bending generated by three-dimensional modeling, and can be a two-dimensional reference contour diagram generated by converting CAD drawings, which is used for pixel-level comparison with the real-time bending image. Among them, the image recognition unit refers to a visual processor with a pattern matching function, which uses a convolutional neural network algorithm to extract and decode the die identification characters. Among them, the control unit refers to an industrial controller with a logical judgment function, which uses PLC programming to realize the execution of the process sequence and abnormal interruption control.

[0020] The operator pre-enters the die model, bending position and pressure parameters corresponding to each process in the preset process module. When the bending machine is started, the acquisition and processing module first scans the information identifiers of the upper and lower dies to verify whether the current die meets the requirements of the first process. During the bending process, the pressure acquisition unit monitors the output value of the hydraulic system in real time. When the pressure value exceeds the preset range, an alarm is triggered. The image recognition unit captures the workpiece contour image after each bending and performs geometric feature matching with the process standard image to confirm the bending angle and position accuracy. If it is detected that the die model does not match or the workpiece shape deviates, the control unit immediately stops the execution of subsequent processes.

[0021] Traditional bending operations rely on the operator to visually check the die model and the information on the process card, which poses a risk of human error. In the existing technology, the bending pressure is only displayed as a process parameter and is not in closed-loop control with the process requirements. This solution eliminates the error probability in the manual verification link through the automatic verification mechanism of die information identifiers. By dynamically associating the pressure monitoring data with the process preset values, real-time calibration of processing parameters is achieved. The image comparison unit replaces the traditional manual caliper measurement method, increasing the bending quality inspection efficiency by more than three times.

[0022] This application effectively avoids the problem of batch scrapping of workpieces caused by incorrect die installation. The automated multi-dimensional verification mechanism during the process execution improves the connection efficiency of multiple bending processes by more than 40%. The dual monitoring of pressure parameters and image data controls the bending angle error within ±0.5 degrees. The whole-process traceability function of die information identifiers provides a data support basis for process optimization.

[0023] The acquisition and processing module includes a workpiece position acquisition unit, a die identifier acquisition unit, and a pressure acquisition unit. The bending machine is provided with a stop for limiting the position of the workpiece. The workpiece position acquisition unit includes a workpiece movement acquisition sub-unit and a in-place acquisition sub-unit. The workpiece movement acquisition sub-unit is used to acquire the movement position of the workpiece and generate a real-time image of the workpiece. The in-place acquisition sub-unit is used to acquire whether the workpiece is in contact with the stop and generate workpiece in-place information. The die identifier acquisition unit is used to identify the die information identifiers of the upper die and the lower die and generate working die information. The pressure acquisition unit is used to acquire the downward pressure of the pressing component on the workpiece to obtain the bending pressure value.

[0024] The workpiece position acquisition unit refers to a detection device composed of a vision sensor and a displacement sensor, which is realized by using an industrial camera in cooperation with a laser rangefinder and is used to obtain the position coordinates of the workpiece on the bending machine workbench in real time. The die identifier acquisition unit refers to a scanning device with image recognition function, which is realized by using a two-dimensional code scanner or a radio frequency identification reader and is used to automatically obtain die identity data. The pressure acquisition unit refers to a pressure sensor installed on the hydraulic cylinder, which is realized by using a strain type pressure sensor and is used to continuously monitor the pressure fluctuation during the bending process.

[0025] The workpiece movement acquisition sub-unit captures the movement trajectory of the workpiece through the vision system, generates a real-time image data stream containing position coordinates. The in-place acquisition sub-unit detects the fitting state of the workpiece and the backing through a contact sensor, and triggers an in-place signal when the workpiece fully contacts the positioning reference surface. The mold identification acquisition unit scans the preset coding identification on the mold surface and automatically matches the mold parameters required for the current process. The pressure acquisition unit records the pressure curve of the pressing component in real time and generates a pressure value corresponding to the process requirements. When each unit works together, after the workpiece is clamped by the robotic arm to the backing for positioning, the system automatically verifies the mold matching degree and monitors the bending pressure to ensure that each bending action conforms to the preset process parameters.

[0026] Existing bending operations rely on manual visual inspection of the mold matching degree and the workpiece positioning status, which has a risk of misjudgment. This solution realizes triple verification of mold information identification, workpiece positioning detection, and pressure monitoring through an automated acquisition system, eliminating human operation errors. For example, the mold identification acquisition unit replaces the step of manually checking the mold number, and the in-place acquisition sub-unit replaces visual observation of positioning through physical contact detection.

[0027] This application effectively prevents workpiece scrapping and equipment damage caused by mold mismatch, solves the problem of low efficiency caused by frequent mold switching in multi-process bending operations, ensures the accuracy consistency of each bending action by real-time collecting pressure data and positioning information, and is particularly suitable for complex workpiece processing scenarios that require multiple mold replacements.

[0028] The bending machine is equipped with a robotic arm, and the robotic arm is provided with mechanical grippers for picking up the workpiece. The workpiece movement acquisition sub-unit is set on the bending machine, and the mold identification acquisition unit is set on the mechanical grippers. The robotic arm refers to an automated operating device with multi-degree-of-freedom movement ability, realized by a six-axis industrial robot, and is used to perform the actions of grasping, moving, and positioning the workpiece. The mechanical grippers refer to the clamping mechanism installed at the end of the robotic arm, realized by pneumatic or electric-driven parallel grippers, and are used to stably grasp the workpiece and adjust the posture in cooperation with the bending process. The workpiece movement acquisition sub-unit refers to a detection device for obtaining the position change of the workpiece, realized by a vision sensor or a laser displacement sensor, and generates a movement trajectory by real-time collecting workpiece images or coordinate data. The mold identification acquisition unit refers to a reading device for identifying the mold identity information, realized by an RFID reader or a QR code scanner, and directly reads the coding labels attached to the upper and lower mold surfaces through the sensors installed on the mechanical grippers.

[0029] After the robotic arm grasps the workpiece with the mechanical gripper, the workpiece movement acquisition sub-unit continuously tracks the movement path of the workpiece during the bending process, and generates position data in real time to determine whether the process requirements are met. When the mechanical gripper touches the die, the die identification acquisition unit automatically reads the encoding information of the current upper and lower dies, and compares it with the die data in the preset process. When the workpiece is transported to the target die area, the system ensures the alignment of the workpiece and the die through the precise positioning of the robotic arm. At the same time, the die matching is quickly verified through the identification acquisition unit integrated on the mechanical gripper, avoiding die misuse or positioning deviation caused by manual operation.

[0030] The existing bending operations rely on manual handling and visual verification of die information, resulting in low positioning efficiency and high risk of misoperation. However, this solution realizes the automated tracking of the workpiece movement trajectory and the real-time verification of die information through the coordinated control of the robotic arm and the identification acquisition unit, eliminating the manual intervention link, significantly improving the coherence and accuracy of multi-process bending operations, optimizing the positioning accuracy of the workpiece during the bending process and the die matching verification efficiency. The mechanical gripper completes the die identity recognition while grasping the workpiece, avoiding workpiece scrapping or equipment damage caused by incorrect die replacement. In addition, the real-time acquisition of the workpiece movement state provides a data basis for the path planning of subsequent processes, ensuring the stable execution of complex bending processes.

[0031] The manual process input unit includes a manual trajectory generation sub-unit, a manual process confirmation sub-unit, and a manual process configuration sub-unit. The manual trajectory generation sub-unit is used to generate the workpiece movement trajectory based on the edge of the workpiece in the real-time image of the workpiece, and generate the gripper movement trajectory according to the movement state of the gripper driven by the robotic arm and the bending of the workpiece by the bending machine. The manual process confirmation sub-unit is used for the user to confirm the workpiece movement trajectory and the gripper movement trajectory. The manual process configuration sub-unit generates several bending process information according to the trajectory optimization sub-unit and the manual process confirmation unit. The bending process information includes the bending position, the process standard image, the die information generated according to the acquisition and processing module, as well as the automatically generated process number and the preset bending pressure value generated according to the bending pressure value.

[0032] The manual trajectory generation subunit refers to a module that generates the movement path of the robotic gripper based on the edge contour of the workpiece in the real-time image. It is implemented by combining an image edge detection algorithm with the kinematic model of the robotic arm. By identifying the edge contour of the workpiece and calculating the gripping position of the gripper, an initial trajectory is generated. The manual process confirmation subunit refers to an interactive interface for operators to review the trajectory generated by the system. It is implemented by displaying the trajectory path through a graphical interface and setting a confirmation button to ensure that the trajectory meets the actual bending requirements. The manual process configuration subunit refers to a module that converts the confirmed trajectory into executable process parameters. It can bind the trajectory coordinates with the preset die information and pressure values, and automatically generate a process number to form a complete process instruction. During the operation, when a manual bending process needs to be input, after the robotic arm grips the workpiece, the real-time image of the workpiece is transmitted to the manual trajectory generation subunit. The edge contour of the workpiece is identified and the corresponding gripper movement trajectory is generated. At the same time, the change in the gripper position caused by the deformation of the workpiece during the bending process of the robotic arm is also recorded to form the gripper movement trajectory in the bending state. The operator views the visualization graphics of the two trajectories through the manual process confirmation subunit. If there is path interference or position deviation, the trajectory points can be manually adjusted. The confirmed trajectory data is input into the manual process configuration subunit. This subunit extracts the bending position coordinates in the trajectory, combines the current die identification information and the downward pressure value, and automatically generates a complete process information including the process number, die information, pressure preset value, and standard image, converting the manual operation process into a standardized process instruction to achieve the rapid configuration of complex bending processes. The operator only needs to visually confirm the trajectory generated by the system without manually calculating the path or recording parameters, which not only reduces the operation difficulty but also improves the consistency and reliability of multi-process bending operations.

[0033] The manual process input unit further includes a coordinate generation subunit and a trajectory optimization subunit. The coordinate generation subunit is connected to the manual trajectory generation subunit and generates a starting coordinate point, a transfer coordinate point, a bending coordinate point, and an end coordinate point according to the manual trajectory generation subunit. The trajectory optimization subunit is connected to the coordinate generation subunit and generates the shortest optimized path in sequence according to the starting coordinate point, the transfer coordinate point, the bending coordinate point, and the end coordinate point. The coordinate generation subunit refers to a module that generates discrete space coordinates based on the edge data of the real-time image of the workpiece, which can be realized by combining a three-dimensional modeling algorithm with a vision recognition technology and is used to discretize the moving trajectory of the workpiece into a quantifiable sequence of coordinate points. The trajectory optimization subunit refers to an operation module that performs path planning on the discrete coordinate points, which can be realized by the Dijkstra algorithm and generates the moving path of the robotic arm by calculating the shortest distance between adjacent coordinate points. The starting coordinate point refers to the grasping position of the mechanical gripper when the workpiece is initially positioned, and the specific coordinates can be determined by identifying the edge contour of the workpiece. The transfer coordinate point refers to the transition position that the robotic arm needs to pass through during the movement, which can be generated within the safe area by a spatial obstacle avoidance algorithm. The bending coordinate point refers to the fixed position of the mechanical gripper when the pressing component performs the bending action on the workpiece, which is determined according to the geometric relationship between the mold and the contact surface of the workpiece. The end coordinate point refers to the release position of the workpiece after a single bending process is completed, which can be dynamically associated with the starting coordinate point of the next process. The shortest optimized path refers to the motion trajectory that minimizes the total moving distance of the robotic arm under the condition of meeting the avoidance constraint conditions, and eliminates redundant moving nodes through a path optimization algorithm. When the user manually operates the robotic arm to complete a demonstration action, the coordinate generation subunit captures the motion trajectory of the mechanical gripper in real time and decomposes it into four types of coordinate points: starting, transfer, bending, and end. After receiving the discrete coordinate points, the trajectory optimization subunit first establishes a spatial coordinate system model, then uses a path search algorithm to calculate the connection method between each coordinate point, and finally generates a collision-free moving path including the avoidance area. By converting the demonstration action into a standardized coordinate sequence, the subsequent automatic bending process can accurately reproduce the manual demonstration path, and at the same time eliminate the ineffective travel generated during the movement of the robotic arm. Compared with the traditional bending machine that relies on the operator's experience to manually control the moving path of the robotic arm, there are problems such as poor path repeatability and easy collision risk. This solution converts the manual demonstration action into a quantifiable optimal motion trajectory through coordinate point discretization and path optimization algorithm, not only retains the operator's process experience, but also ensures the accuracy and safety of the moving path through mathematical modeling, realizing the efficient conversion from the demonstration action to the automatic process, effectively reducing the ineffective moving time of the robotic arm in the multi-process bending operation, and at the same time avoiding the interference risk between the mold and the workpiece through path optimization. The standardized processing of the coordinate point sequence makes the connection between different processes closer, and the dynamic association between the end coordinate and the starting point of the next process further shortens the process switching time and improves the overall efficiency of the complex workpiece bending operation.

[0034] The end coordinate point is the starting coordinate point of the next process. The end coordinate point refers to the final position coordinate where the robotic arm moves to after the completion of the current bending process. It can be achieved by using a vision positioning system or an encoder to locate the spatial coordinates of the mechanical gripper, and is used to mark the end position of the current process. Among them, the starting coordinate point refers to the initial position coordinate of the robotic arm at the start of the next bending process, which can be determined through a workpiece edge recognition algorithm combined with a coordinate mapping relationship, and is used to mark the starting operation position of the next process. Among them, process connection refers to the coordinate point association relationship between adjacent processes, which is achieved by automatically setting the end coordinate point of the current process as the starting coordinate point of the next process, and is used to eliminate the path redundancy during process switching. After completing a bending process, the robotic arm moves the workpiece to the end coordinate point according to the preset path. At this time, this coordinate point is automatically recorded as the starting coordinate point of the next process. When generating the process path, the trajectory optimization sub-unit matches the end coordinate point with the starting coordinate point of the next process. When the two coincide, the system automatically omits the repeated positioning steps in the intermediate movement path. Thus, the robotic arm can directly start the clamping or movement action of the next process from the end coordinate point of the current process without returning to the initial position or making additional adjustments.

[0035] After the bending machine completes a single process, it is necessary to reposition the workpiece to a fixed starting point, resulting in an invalid movement path for the robotic arm, increasing the operation time. This solution enables the path planning between processes to form a continuous trajectory through a coordinate point dynamic association mechanism, effectively shortening the movement distance of the robotic arm, enabling seamless connection of multiple bending processes, reducing the idle travel time of the robotic arm during process switching, avoiding cumulative errors caused by repeated positioning, and at the same time reducing the risk of the workpiece shifting or colliding during transfer, thereby improving the operation efficiency and accuracy of multi-process bending of complex workpieces.

[0036] The bending machine control system also includes a workpiece bending acquisition sub-unit. The workpiece bending acquisition sub-unit is arranged on the side of the workpiece to detect the front view of the workpiece. The workpiece bending acquisition sub-unit is used to collect the bending state of the workpiece to generate a real-time front view image. The image recognition unit is used to extract key frames from the real-time front view image. When the workpiece in-place information is generated, the image recognition unit generates an in-place key frame. When the pressure acquisition unit collects a stable bending pressure value, the image recognition unit generates a bending-in-place key frame. The image comparison unit compares the key frames generated by the image recognition with the process standard images corresponding to the bending processes.

[0037] The workpiece bending acquisition sub-unit refers to the image acquisition device set on the side of the workpiece, which can be implemented by an industrial camera or a laser scanner. Its function is to perform real-time image acquisition on the front view during the workpiece bending process and capture the deformation state of the workpiece. Key frame extraction refers to extracting representative static pictures from the continuously acquired image sequence, which is achieved through time interval sampling or dynamic change detection algorithms, and is used to reduce the amount of data processing and improve the comparison efficiency. The in-place key frame refers to the image data at the moment when the workpiece abuts against the backstop, which is synchronously intercepted through the trigger signal of the in-place acquisition sub-unit and is used to verify the accuracy of workpiece positioning. The key frame of bending in place refers to the image data when the downward pressure reaches the preset value and remains stable, which is achieved by combining the threshold judgment of the pressure sensor and the image acquisition timestamp, and is used to confirm the completion degree of the bending action. Image comparison refers to performing pixel-level matching between the key frame and the preset process standard image, and gray-scale comparison or edge feature matching algorithms can be used to detect the deviation between the bending shape of the workpiece and the process requirements.

[0038] The workpiece bending acquisition sub-unit is installed on the side of the workpiece to obtain the real-time image stream of the front view including the upper die, the lower die and the workpiece. When the workpiece moves to contact the backstop, the in-place acquisition sub-unit sends a trigger signal, and the image recognition unit immediately intercepts the current frame as the in-place key frame. During the process of the downward pressing component applying the bending force, the pressure acquisition unit continuously monitors the pressure value. When the fluctuation range of the pressure value is less than the threshold within the preset time, it is determined that the bending is in place, and the front view at this moment is synchronously intercepted as the key frame of bending in place. The two frames of images are respectively compared with the process standard images of the corresponding processes. If the workpiece positioning error of the in-place key frame exceeds the allowable range, an alarm is triggered and the process is paused; if the bending angle or shape of the key frame of bending in place does not match the standard image enough, it is prompted to adjust the die or correct the pressure. In the traditional bending process, the operator relies on visual inspection of the workpiece position and bending result, which has subjective judgment errors and cannot monitor the correlation between the pressure change and the workpiece deformation in real time. There is a lack of an automated image capture mechanism for key nodes during the bending process in the prior art, and it is difficult to synchronously verify the workpiece state during the process execution stage.

[0039] This application realizes the closed-loop control of the bending process, automatically triggers image acquisition and comparison when the workpiece positioning is completed and the bending action ends, ensures that the execution result of each process step meets the preset standards, and effectively avoids workpiece scrapping caused by die misalignment or pressure deviation through the automated comparison between the key frame and the standard image. At the same time, it reduces the operation frequency of manual re-inspection and improves the coherence and reliability of the multi-process bending operation.

[0040] The coordinate generation subunit generates an avoidance area for the front views of the upper die and the lower die based on the real-time front view image of the workpiece bending acquisition subunit. The coordinate generation subunit generates avoidance coordinate points according to the avoidance area. The avoidance coordinate points are set between the starting coordinate point and the bending coordinate point, and the avoidance coordinate points are also set between the bending coordinate point and the end coordinate point. The avoidance area refers to the spatial area formed by analyzing the side structure of the die through the real-time front view image, which is used to avoid interference between the movement path of the robotic arm and the die. It is generated by using an image processing algorithm to identify the side contour of the die and calculate the safety distance. The avoidance coordinate points refer to the temporary positioning points set in the movement path of the robotic arm, which are used to bypass the side area of the die and are realized by inserting intermediate points between the starting coordinate point and the bending coordinate point and between the bending coordinate point and the end coordinate point through a path planning algorithm.

[0041] During the process of the robotic arm gripping and moving the workpiece, the coordinate generation subunit extracts the side contour information of the die based on the real-time front view image and generates the boundary of the avoidance area. Subsequently, the path planning module inserts avoidance coordinate points into the straight-line path from the starting coordinate point to the bending coordinate point, enabling the robotic arm to move along the optimized broken-line path to ensure that the gripper and the workpiece maintain a safe distance from the side of the die. After the bending operation is completed, when the robotic arm returns from the bending coordinate point, the avoidance coordinate points guide the gripper to bypass the die area again to avoid collisions on the return path.

[0042] When the traditional bending machine plans the movement path of the robotic arm, it does not consider the spatial occupation of the side structure of the die, resulting in the gripper or the workpiece being prone to rubbing against the die during the movement process. This solution generates a dynamic avoidance area through real-time images and sets avoidance coordinate points at key path segments, enabling the movement trajectory of the robotic arm to actively bypass the dangerous area of the die, effectively reducing the risk of equipment collision.

[0043] This application solves the problem of interference between the robotic arm path and the die in multi-process bending operations, avoids die damage or workpiece scrapping caused by improper path planning, and at the same time reduces the frequency of manual intervention to adjust the path, improving the automation level and operation safety of multi-process continuous bending.

[0044] A bending machine control method is applied to the above-mentioned bending machine control system. The control method of the bending machine includes: Step S1, the preset process module presets the bending process of the workpiece. The preset process module is used for the user to input several bending process information, and several bending process information all includes process number, bending position, die information, preset value of bending pressure, and process standard image; The manual process input unit includes a manual trajectory generation subunit, a coordinate generation subunit, a trajectory optimization subunit, a manual process confirmation subunit, and a manual process configuration subunit; The manual trajectory generation subunit is used to generate the workpiece movement trajectory based on the edge of the workpiece in the real-time image and the bending state jaw movement trajectory of the mechanical jaw when the bending machine bends the workpiece; The coordinate generation subunit is connected to the manual trajectory generation subunit and generates the starting coordinate point, transfer coordinate point, bending coordinate point, and ending coordinate point according to the manual trajectory generation subunit; The trajectory optimization subunit is connected to the coordinate generation subunit and generates the shortest optimized path according to the starting coordinate point, transfer coordinate point, bending coordinate point, and ending coordinate point in sequence.

[0045] The manual process confirmation subunit is used for the user to confirm the workpiece movement trajectory and the jaw movement trajectory; The manual process configuration subunit generates several bending process information according to the trajectory optimization subunit and the manual process confirmation unit. The bending process information includes the bending position, process standard image, die information generated according to the acquisition and processing module, as well as the automatically generated process number and the preset bending pressure value generated according to the bending pressure value; Step S2, the control unit controls the bending machine and the robotic arm to bend the workpiece according to the preset bending process information; Step S3, the die identification acquisition unit is used to identify the die information identification of the upper die and the lower die and generate the working die information; The workpiece position acquisition unit is used to generate a real-time image of the bending state of the workpiece; The pressure acquisition unit is used to acquire the downward pressure of the downward pressing component on the workpiece to obtain the bending pressure value; Step S4, the image recognition unit recognizes the information codes of the upper die and the lower die according to the acquisition and processing module and compares the die information in the bending process information; The image recognition unit is used to extract key frames from the real-time image, and the image comparison subunit compares the key frames generated by the image recognition with the process standard images of the corresponding bending processes to detect whether the workpieces in each bending process meet the standards.

[0046] The control method defines the execution parameters of each bending process through preset process information. During the bending process, it synchronously collects die information, workpiece position images, and pressure data, and realizes the closed-loop verification of the process execution status through image recognition and comparison. For example, in step S4, the system extracts the key frames when the workpiece arrives and performs contour matching with the process standard images. If the deviation exceeds the threshold, an alarm or correction instruction is triggered to avoid quality problems caused by die misalignment or incorrect workpiece placement. In addition, the pressure acquisition unit continuously monitors the lower pressure data. If the actual pressure deviates from the preset value, the output of the hydraulic system can be dynamically adjusted to ensure bending accuracy. The manual process input unit generates the workpiece movement path and the movement trajectory of the mechanical gripper according to the real-time image through the manual trajectory generation subunit. The trajectory optimization subunit generates the shortest path based on the starting coordinate point, the transfer coordinate point, and the ending coordinate point, reducing the idle travel time of the robotic arm. The coordinate generation subunit can also set avoidance coordinate points according to the avoidance area to prevent the mechanical gripper from interfering with the die.

[0047] Traditional bending processes rely on manual verification of die information and visual inspection of workpiece status, which have the defects of low efficiency and high error rates. This method realizes the real-time verification of process execution parameters through automatic identification of die information identification and image comparison technology, eliminating the risk of human operation errors. At the same time, the comparison mechanism based on key frames can quickly locate the bending deviation of the workpiece, avoiding material waste caused by process errors, solving the problems of die matching errors, workpiece placement deviation, and inaccurate pressure control in multiple bending processes, improving the consistency and yield rate of the bending process. Through the synergistic effect of preset process parameters and real-time data collection, it realizes the automatic monitoring and closed-loop control of the bending process, especially suitable for the continuous bending processing scenario of complex workpieces.

Claims

1. A bending machine control system, including a bending machine, wherein an upper die holder, a lower die holder and a downward pressing assembly for driving the upper die holder to move towards the lower die holder to bend a workpiece are arranged on the bending machine. An upper die is installed on the upper die holder, and a lower die is installed on the lower die holder. It is characterized in that, It further includes a collection and processing module, a preset process module, and a control module. The upper die and the lower die are both associated with corresponding die information identifiers. The collection and processing module is connected to the bending machine and is used to collect the die information identifiers of the current upper die and lower die of the bending machine, the real-time image in the working state of the bending machine, and the downward pressure of the downward pressing component, so as to obtain the working die information, the real-time image, and the bending pressure value respectively. The preset process module includes a preset process configuration unit, and the preset process configuration unit is used for the user to input several bending process information. The several bending process information all includes a process number, a bending position, die information, a preset value of the bending pressure, and a process standard image. The control module includes: An image recognition unit, connected to the collection and processing module and the preset process module, is used to recognize the information codes of the upper die and the lower die by the collection and processing module and compare the die information in the bending process information. An image comparison unit, connected to the preset process module and the collection and processing module, is used to compare the workpiece image after the process is completed intercepted from the real-time image with the process standard image in the bending process information. A control unit, connected to the preset process module, is used to drive the bending machine to bend the workpiece according to the preset process by the preset process configuration unit.

2. The bending machine control system according to claim 1, characterized in that, The collection and processing module includes a workpiece position collection unit, a die identifier collection unit, and a pressure collection unit. The bending machine is provided with a backstop for limiting the workpiece. The workpiece position collection unit includes a workpiece movement collection sub-unit and a in-place collection sub-unit. The workpiece movement collection sub-unit is used to collect the movement position of the workpiece and generate a real-time image of the workpiece. The in-place collection sub-unit is used to collect whether the workpiece is in contact with the backstop and generate workpiece in-place information. The die identifier collection unit is used to identify the die information identifiers of the upper die and the lower die and generate working die information. The pressure collection unit is used to collect the downward pressure of the downward pressing component on the workpiece to obtain the bending pressure value.

3. The bending machine control system according to claim 2, wherein, The bending machine is configured with a robotic arm, and the robotic arm is provided with a mechanical gripper for picking up the workpiece. The workpiece movement collection sub-unit is arranged on the bending machine, and the die identifier collection unit is arranged on the mechanical gripper.

4. The bending machine control system according to claim 3, characterized in that, The preset process module further includes a manual process input unit, and the manual process input unit is used for the user to manually operate the robotic arm. The manual process input unit includes a manual trajectory generation sub-unit, a manual process confirmation sub-unit, and a manual process configuration sub-unit. The manual trajectory generation sub-unit is used to generate a workpiece movement trajectory according to the edge of the workpiece in the real-time image of the workpiece, and generate a gripper movement trajectory according to the state of the workpiece bending and driving the gripper to move when the robotic arm drives the mechanical gripper and the bending machine to bend the workpiece. The manual process confirmation sub-unit is used for the user to confirm the workpiece movement trajectory and the gripper movement trajectory. The manual process configuration subunit generates a number of the bending process information according to the trajectory optimization subunit and the manual process confirmation unit. The bending process information includes the bending position, the process standard image, the die information generated according to the acquisition and processing module, as well as the automatically generated process number and the bending pressure preset value generated according to the bending pressure value.

5. The bending machine control system according to claim 4, characterized in that, The manual process input unit further includes a coordinate generation subunit and a trajectory optimization subunit; The coordinate generation subunit is connected to the manual trajectory generation subunit and generates a starting coordinate point, a transfer coordinate point, a bending coordinate point, and an ending coordinate point according to the manual trajectory generation subunit; The trajectory optimization subunit is connected to the coordinate generation subunit and sequentially generates the shortest optimized path according to the starting coordinate point, the transfer coordinate point, the bending coordinate point, and the ending coordinate point.

6. The bending machine control system according to claim 5, characterized in that, The ending coordinate point is the starting coordinate point of the next process.

7. A bending machine control system according to claim 5, characterized in that, The workpiece position acquisition unit further includes a workpiece bending acquisition subunit. The workpiece bending acquisition subunit is arranged on the side of the workpiece to detect the front view of the workpiece. The workpiece bending acquisition subunit is used to acquire the bending state of the workpiece to generate a real-time front view image. The image recognition unit is used to extract key frames from the real-time front view image. When the workpiece in-place information is generated, the image recognition unit generates the in-place key frame. When the pressure acquisition unit acquires that the bending pressure value is stable, the image recognition unit generates the bending-in-place key frame. The image comparison unit compares the key frames generated by the image recognition with the process standard images of the corresponding bending processes.

8. A bending machine control system according to claim 7, characterized in that, The coordinate generation subunit generates an avoidance area based on the front view images of the upper die and the lower die on the real-time front view image of the workpiece bending acquisition subunit. The coordinate generation subunit generates avoidance coordinate points according to the avoidance area. The avoidance coordinate points are set between the starting coordinate point and the bending coordinate point, and the avoidance coordinate points are also set between the bending coordinate point and the ending coordinate point.

9. A bending machine control method, characterized in that, Applied to the bending machine control system according to any one of claims 1-8, the bending machine control method includes: Step S1, the preset process module presets the bending process of the workpiece. The preset process module is used for the user to input a number of bending process information. A number of the bending process information all includes a process number, a bending position, die information, a bending pressure preset value, and a process standard image; Step S2, the control unit controls the bending machine and the robotic arm to bend the workpiece according to the process according to the preset bending process information; Step S3, the die identification acquisition unit is used to identify the die information identifiers of the upper die and the lower die and generate the working die information; The workpiece position acquisition unit is used to generate a real-time image of the bending state of the workpiece; The pressure acquisition unit is used to acquire the downward pressure of the pressing component on the workpiece to obtain the bending pressure value; Step S4, the image recognition unit identifies the information codes of the upper die and the lower die according to the acquisition and processing module and compares with the die information in the bending process information; The image recognition unit is used to extract key frames from real-time images. The image comparison sub-unit compares the key frames generated by the image recognition with the process standard images corresponding to the bending processes, and is used to detect whether the workpieces in each bending process meet the standards.

10. A bending machine control method according to claim 9, characterized in that, Step S1 further includes a manual process input unit, which includes a manual trajectory generation sub-unit, a coordinate generation sub-unit, a trajectory optimization sub-unit, a manual process confirmation sub-unit, and a manual process configuration sub-unit; The manual trajectory generation sub-unit is used to generate the workpiece movement trajectory and the gripper movement trajectory of the bending state of the mechanical gripper when the bending machine bends the workpiece according to the edge of the workpiece in the real-time image; The coordinate generation sub-unit is connected to the manual trajectory generation sub-unit, and generates a starting coordinate point, a transfer coordinate point, a bending coordinate point, and an end coordinate point according to the manual trajectory generation sub-unit; The trajectory optimization sub-unit is connected to the coordinate generation sub-unit, and sequentially generates the shortest optimized path according to the starting coordinate point, the transfer coordinate point, the bending coordinate point, and the end coordinate point; The manual process confirmation sub-unit is used for the user to confirm the workpiece movement trajectory and the gripper movement trajectory; The manual process configuration sub-unit generates a number of bending process information according to the trajectory optimization sub-unit and the manual process confirmation unit. The bending process information includes the bending position, the process standard image, the die information generated according to the acquisition and processing module, as well as the automatically generated process number and the preset bending pressure value generated according to the bending pressure value.

Images