Intelligent material taking method and system for injection mold

Abstract

The invention discloses an intelligent material taking method and system for an injection mold, and relates to the technical field of intelligent material taking methods, the relative positions of a plurality of suction cups in a material taking support arm are determined according to a plurality of material taking areas and the spatial position of the material taking support arm, and the accuracy of the relative positions of the plurality of suction cups is ensured. Precise adsorption of the multiple suction cups to the product is achieved. Therefore, the adsorption force of the multiple suction cups is determined according to the demolding force of the injection mold on the product, the relative positions of the multiple suction cups and the curved surface coefficients of the multiple material taking areas; an intelligent material taking system of the product is determined according to the moving tracks of the multiple suction cups, the posture parameters of the product and the adsorption force of the multiple suction cups, and intelligent control measures of the product in the material taking process are generated in the intelligent material taking system so that the multiple suction cups can intelligently adsorb the product, the disengagement of the product in the material taking process can be overcome, and the product taking efficiency can be improved. And the intelligent material taking effect of the injection mold is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent material taking methods, and in particular to an intelligent material taking method and system for an injection mold. Background Art

[0002] With the development of technology, injection molds are gradually applied to people's lives and some plastic products are formed during the injection process. The work of injection molds covers the injection stage, the mold release stage, and the detachment stage of plastic products. In the prior art, during the control of the mold release stage of plastic products, multiple suction cups in the material taking arm adsorb the product along a preset adsorption force, and each suction cup uses the same adsorption force for the product, which cannot realize the intelligent adsorption of the product by multiple suction cups, resulting in the event that the product still detaches during the material taking process. Summary of the Invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art, and the present invention provides an intelligent material taking method and system for an injection mold.

[0004] An embodiment of the present invention provides an intelligent material taking method for an injection mold, including:

[0005] Collect an open mold image of the injection mold;

[0006] Determine multiple material taking areas according to the open mold image of the injection mold and the product model formed by the injection mold;

[0007] With the material taking arm in the open mold space of the injection mold, determine the relative positions of multiple suction cups in the material taking arm according to the multiple material taking areas and the spatial position of the material taking arm;

[0008] Determine the adsorption forces of multiple suction cups according to the demolding force of the injection mold on the product, the relative positions of multiple suction cups, and the surface coefficients of multiple material taking areas;

[0009] Determine an intelligent material taking system for the product according to the movement trajectories of multiple suction cups, the attitude parameters of the product, and the adsorption forces of multiple suction cups, and generate intelligent control measures for the product during the material taking process in the intelligent material taking system.

[0010] An embodiment of the present invention provides an intelligent material taking system for an injection mold. The intelligent material taking system for an injection mold is applied to the above-mentioned intelligent material taking method for an injection mold. The intelligent material taking system for an injection mold includes:

[0011] A collection module, configured to collect an open mold image of the injection mold;

[0012] A material taking area module, configured to determine multiple material taking areas according to the open mold image of the injection mold and the product model formed by the injection mold;

[0013] A position module, which is used for the material taking arm to be in the mold opening space of the injection mold, and determines the relative positions of multiple suction cups in the material taking arm according to multiple material taking areas and the spatial position of the material taking arm;

[0014] An adsorption force module, which is used to determine the adsorption forces of multiple suction cups according to the demolding force of the injection mold on the product, the relative positions of multiple suction cups, and the surface coefficients of multiple material taking areas;

[0015] An intelligent material taking module, which is used to determine an intelligent material taking system for the product according to the movement trajectories of multiple suction cups, the attitude parameters of the product, and the adsorption forces of multiple suction cups, and generate intelligent control measures for the product during the material taking process in the intelligent material taking system.

[0016] Compared with the prior art, the beneficial effects of the present invention are:

[0017] In the embodiment of the present invention, by the method in the embodiment of the present invention, an opening mold image of the injection mold is collected; multiple material taking areas are determined according to the opening mold image of the injection mold and the product model formed by the injection mold; the material taking arm is in the mold opening space of the injection mold, and the relative positions of multiple suction cups in the material taking arm are determined according to multiple material taking areas and the spatial position of the material taking arm, which comprehensively considers multiple material taking areas and the spatial position of the material taking arm, ensures the accuracy of the relative positions of multiple suction cups, and realizes the precise adsorption of multiple suction cups on the product.

[0018] Therefore, the adsorption forces of multiple suction cups are determined according to the demolding force of the injection mold on the product, the relative positions of multiple suction cups, and the surface coefficients of multiple material taking areas; an intelligent material taking system for the product is determined according to the movement trajectories of multiple suction cups, the attitude parameters of the product, and the adsorption forces of multiple suction cups, and intelligent control measures for the product during the material taking process are generated in the intelligent material taking system, so as to facilitate the intelligent adsorption of multiple suction cups on the product, overcome the detachment of the product during the material taking process, and realize the intelligent material taking effect of the injection mold. Description of the Drawings

[0019] Figure 1 is a schematic flow chart of the intelligent material taking method for the injection mold in the embodiment of the present invention;

[0020] Figure 2 is a schematic flow chart of step S11 in the intelligent material taking method for the injection mold in the embodiment of the present invention;

[0021] Figure 3 is a schematic flow chart of step S12 in the intelligent material taking method for the injection mold in the embodiment of the present invention;

[0022] Figure 4It is a schematic flowchart of step S13 in the intelligent material taking method of the injection mold in the embodiment of the present invention;

[0023] Figure 5 It is a schematic flowchart of step S14 in the intelligent material taking method of the injection mold in the embodiment of the present invention;

[0024] Figure 6 It is a schematic flowchart of step S15 in the intelligent material taking method of the injection mold in the embodiment of the present invention;

[0025] Figure 7 It is a schematic diagram of the structural composition of the intelligent material taking system of the injection mold in the embodiment of the present invention. Detailed implementation manners

[0026] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention.

[0027] Please refer to Figures 1 to 7 , an intelligent material taking method for an injection mold, which is applied to the intelligent material taking scenario of the injection mold; the intelligent material taking method for the injection mold includes:

[0028] Step S11: Collect the mold opening image of the injection mold;

[0029] Step S12: Determine multiple material taking areas according to the mold opening image of the injection mold and the product model formed by the injection mold;

[0030] Step S13: The material taking arm is in the mold opening space of the injection mold, and determine the relative positions of multiple suction cups in the material taking arm according to the multiple material taking areas and the spatial position of the material taking arm;

[0031] Step S14: Determine the adsorption forces of multiple suction cups according to the demolding force of the injection mold on the product, the relative positions of multiple suction cups, and the surface coefficients of multiple material taking areas;

[0032] Step S15: Determine the intelligent material taking system of the product according to the movement trajectories of multiple suction cups, the attitude parameters of the product, and the adsorption forces of multiple suction cups, and generate intelligent control measures for the product during the material taking process in the intelligent material taking system;

[0033] Refer to Figure 2 , in step S11, collect the mold opening image of the injection mold;

[0034] In the specific implementation process of the present invention, the specific steps are:

[0035] S111: When the injection mold is in the mold opening state, collect the mold opening space of the injection mold, and determine the product area of the injection mold according to the mold opening space of the injection mold and the cavity position of the injection mold;

[0036] S112: Trigger the responses of multiple cameras according to the product area of the injection mold and the mold opening direction of the injection mold, construct a shooting system for the injection mold based on the multiple cameras, and determine the mold opening image of the injection mold according to the shooting of the product area of the injection mold by the shooting system. At this time, the multiple cameras are distributed around the injection mold.

[0037] In the embodiment of the present application, when the injection mold is in the mold opening state, collect the mold opening space of the injection mold, and determine the product area of the injection mold according to the mold opening space of the injection mold and the cavity position of the injection mold, and introduce the product area of the injection mold.

[0038] At this time, through the signal of the sensor or the injection molding machine control system, confirm that the injection mold has been fully opened to the preset mold opening position, so as to control whether the injection mold is in the mold opening state; use a three-dimensional scanning device (such as a laser scanner, a structured light scanner or a stereo vision system) to scan the mold opening space of the injection mold to obtain the three-dimensional point cloud data inside the injection mold; perform preprocessing operations such as denoising, smoothing and registration on the collected three-dimensional point cloud data to generate a high-precision three-dimensional model.

[0039] In the three-dimensional model, identify the cavity position of the injection mold through an image processing algorithm; the cavity is the cavity part in the mold for forming the product; according to the shape, size and position of the cavity, combined with the CAD model of the product design, determine the product area in the injection mold.

[0040] Specifically, assume that we are dealing with an injection mold for an automotive interior part; after the mold is fully opened, we use a laser scanner to scan the inside of the mold from multiple angles; after the scanning is completed, we obtain a three-dimensional point cloud data set containing the shape and size information inside the mold; through data processing software, we convert these point cloud data into a high-precision three-dimensional model for subsequent analysis.

[0041] The cavity area of the mold is identified in the three-dimensional model; by comparing the CAD model of the product design, we find that the shape and size of the cavity are exactly the same as the product design; therefore, we can determine that the product area in the mold is the area where the cavity is located; in order to more accurately extract the product area, we use three-dimensional modeling software to perform a cutting operation on the three-dimensional model of the mold to obtain a three-dimensional model fragment that is consistent with the product design shape, and this three-dimensional model fragment is the product area we want to determine.

[0042] Therefore, trigger the responses of multiple cameras according to the product area of the injection mold and the mold opening direction of the injection mold, construct a shooting system for the injection mold based on the multiple cameras, and determine the mold opening image of the injection mold according to the shooting of the product area of the injection mold by the shooting system. At this time, the multiple cameras are distributed around the injection mold, ensuring the accuracy of the mold opening image of the injection mold.

[0043] At this time, trigger the responses of multiple cameras according to the product area of the injection mold and the mold opening direction of the injection mold, introducing the responses of multiple cameras. At the same time, according to the product area determined in step S111 and the mold opening direction of the injection mold, analyze the optimal layout and shooting angles of the cameras; this usually needs to consider the shape, size, material of the product and the structural characteristics of the injection mold; according to the analysis results, configure the trigger mechanism of the cameras; this can be achieved through hardware triggers (such as photoelectric switches, proximity sensors) or software commands (such as signals from the injection molding machine control system); when the injection mold reaches the mold opening position and the product area is clearly visible, trigger the cameras to take pictures.

[0044] Determine the installation positions and angles of the cameras according to the size of the injection mold and the field of view of the cameras; ensure that the cameras can clearly capture the details of the product area; connect the cameras to an image acquisition card or an image processor, and configure the parameters of the cameras (such as exposure time, white balance, frame rate, etc.) to ensure the shooting quality; at the same time, configure the parameters of the image acquisition software, such as the image storage path, file name format, etc.; if multiple cameras are used, it is necessary to ensure that they can shoot synchronously to avoid image inconsistency problems caused by shooting time differences; this can be achieved by using synchronous triggers or software synchronization functions.

[0045] When the cameras are triggered, they will shoot the product area inside the injection mold; the image acquisition software will receive and store these images; preprocess the acquired images, such as denoising, enhancing contrast, cropping, etc., to improve the image quality; if multiple cameras are used, it is also necessary to stitch or fuse the images to generate a complete three-dimensional view or panoramic image; determine the mold opening image of the injection mold according to the preprocessed image data.

[0046] Specifically, taking an injection mold for a mobile phone case as an example, we know that the product area is a flat shell with complex curved surfaces; the mold opening direction of the injection mold is vertically upward; therefore, we decide to install four high-definition cameras around the injection mold, and each camera points to the inside of the injection mold to ensure that a complete view of the product area can be captured; we configure a photoelectric switch as a trigger, and when the injection mold is fully opened and the mobile phone case is exposed, the photoelectric switch is blocked, triggering the cameras to take pictures.

[0047] Four cameras with 1080P resolution and wide-angle view were selected and installed at the four corners of the injection mold; by adjusting the installation angle and focal length of the cameras, we ensured that each camera could clearly capture the complete view of the mobile phone case; we connected the cameras to a high-performance image acquisition computer and configured image acquisition software, setting appropriate shooting parameters and storage paths; to ensure that the four cameras could shoot synchronously, we used a synchronization trigger to trigger their shooting.

[0048] When the injection mold was fully opened and the cameras were triggered, the four cameras simultaneously captured the view of the mobile phone case; the image acquisition software received and stored these images; then, we preprocessed the images using image processing software, such as denoising and enhancing contrast; since we used four cameras, we stitched the four views together to generate a complete three-dimensional view; finally, we determined the mold opening image of the injection mold based on this three-dimensional view and used it in the subsequent intelligent material taking method.

[0049] Reference Figure 3 , in step S12, multiple material taking areas are determined according to the mold opening image of the injection mold and the product model formed by the injection mold;

[0050] In the specific implementation process of the present invention, the specific steps are as follows:

[0051] S121: Generate multiple functional areas according to the division of the mold opening image of the injection mold, determine the product area based on the screening of the functional areas, determine multiple product contours according to the preprocessing of the product area, and mark the central positions of the multiple product contours;

[0052] S122: Determine the product model formed by the injection mold according to the database of the injection mold and the mold number information of the injection mold, determine multiple theoretical material taking positions of the product model according to the detection of the product model formed by the injection mold, and determine multiple material taking areas according to the multiple theoretical material taking positions, the central positions of the multiple product contours, and the product area.

[0053] In the embodiment of the present application, generating multiple functional areas according to the division of the mold opening image of the injection mold, determining the product area based on the screening of the functional areas, determining multiple product contours according to the preprocessing of the product area, and marking the central positions of the multiple product contours ensure the accuracy of the multiple product contours.

[0054] At this time, the opening image of the injection mold is introduced and the opening image of the injection mold is divided to facilitate the generation of multiple functional areas. At the same time, the opening image of the injection mold is carefully analyzed so that the image can be segmented into different functional areas according to features such as pixel information, color, and texture of the image. These functional areas may include the product area, the waste area, the mold frame, the positioning hole area, etc. The segmentation method can be based on image processing algorithms such as threshold segmentation, edge detection, region growing, etc. Optionally, use image processing software (such as OpenCV) to load the opening image; apply a suitable segmentation algorithm to segment the image into different areas according to the gray value, color or texture difference of the image; label and classify the segmented areas, such as the product area, the waste area, etc.

[0055] After multiple functional areas are segmented, a series of screening conditions are used to determine the product area. These screening conditions may include the size, shape, position of the area, and the degree of matching with the mold design drawing, etc. Through screening, we can exclude non-product areas such as the waste area and the mold frame, so as to accurately determine the product area. Optionally, set the screening conditions according to the mold design drawing and product specifications, such as the size range and shape characteristics of the product area; screen each segmented area to determine whether it meets the conditions of the product area; mark the areas that meet the conditions as the product area.

[0056] After the product area is determined, we need to preprocess the product area to improve the image quality and extract the contour of the product. The preprocessing may include denoising, enhancing contrast, morphological processing, etc. The contour extraction can use edge detection algorithms such as Canny edge detection, etc. Optionally, preprocess the product area, such as using a Gaussian filter to remove noise and using histogram equalization to enhance contrast; apply an edge detection algorithm to extract the contour of the product; post-process the extracted contour, such as removing burrs and filling holes, etc., to obtain a more accurate contour.

[0057] After the product contour is extracted, calculate the center position or centroid position of each contour and mark it. This helps the subsequent material taking operation because the material taking tools such as the robotic arm or suction cup need to accurately locate to the center position of the product. Optionally, calculate the center point or centroid position of each extracted contour; mark the center position of each contour on the image, and marking methods such as dots or crosshairs can be used; save or transmit the marked image to the subsequent material taking system.

[0058] Specifically, assume that we are processing the opening image of a mobile phone case mold for injection molding; use OpenCV to load the opening image; apply the threshold segmentation algorithm to segment the image into different areas; divide the image into the product area, the waste area, and the mold frame area according to the size, shape, and position of the area.

[0059] Set screening conditions according to the design drawing and specifications of the mobile phone case, such as the size range and shape characteristics of the product area; screen each divided area to determine the product areas that meet the conditions; at the same time, preprocess the product areas, such as removing noise and enhancing contrast; apply the Canny edge detector to extract the contour of the mobile phone case; post-process the extracted contour, such as removing burrs and filling holes.

[0060] Calculate the center point or centroid position of each contour; mark the center position of each mobile phone case contour on the image; transmit the marked image to the subsequent material picking system so that material picking tools such as robotic arms or suction cups can accurately locate the center position of the mobile phone case; through the above steps, we can accurately determine the product area, product contour and the center position of the contour in the injection mold opening image, providing key data support for subsequent intelligent material picking operations.

[0061] Therefore, determine the product model formed by the injection mold according to the database of the injection mold and the mold number information of the injection mold, determine multiple theoretical material picking positions of the product model according to the detection of the product model formed by the injection mold, and determine multiple material picking areas according to the multiple theoretical material picking positions, the center positions of multiple product contours and the product area, taking into account the overall consideration of multiple theoretical material picking positions, the center positions of multiple product contours and the product area, ensuring the accuracy of multiple material picking areas.

[0062] At this time, access the database of the injection mold, which usually contains key data such as the mold number, corresponding product model, product design drawing, material information and production parameters of the mold; by inputting the mold number information, we can retrieve the product model corresponding to the mold in the database; optionally, open the injection mold database; input the mold number information to ensure the uniqueness and accuracy of the number; retrieve the product model data matching the input number in the injection mold database; load and display the retrieved product model for subsequent analysis.

[0063] After determining the product model, determine multiple theoretical material picking positions according to factors such as the shape, size, material and production requirements of the product; the theoretical material picking positions are usually parts of the product that are easy to grasp, not easy to damage and can meet the production rhythm; when determining the theoretical material picking positions, factors such as the center of gravity of the product, the size and shape of the grasping tool and the layout of the production line need to be considered; optionally, analyze the three-dimensional data of the product model to understand the shape and size of the product; initially determine several possible material picking positions according to the center of gravity and production requirements of the product; use simulation software or simulation tools to simulate the situation of the grasping tool grasping the product at different positions, and evaluate the stability and efficiency of the grasping; finally determine multiple theoretical material picking positions according to the simulation results.

[0064] After determining the theoretical material taking position, map the theoretical material taking position onto the mold opening image of the injection mold, and combine the center position of the product contour and the product area information to determine the actual material taking area; the material taking area is usually near the product contour and is part that is easily accessible to material taking tools such as robotic arms or suction cups; when determining the material taking area, factors such as the accessibility of the material taking tool, the grasping accuracy, and the production rhythm need to be considered. Optionally, map the theoretical material taking position onto the mold opening image of the injection mold to ensure the accuracy of the position; combine the center position of the product contour and the product area information to determine the actual material taking area corresponding to each theoretical material taking position; consider the accessibility and grasping accuracy of the material taking tool, and fine-tune the material taking area to ensure the smooth progress of the material taking operation; save or transmit the determined material taking area information to the subsequent material taking system so that material taking tools such as robotic arms or suction cups can accurately position and perform the material taking operation.

[0065] Specifically, assume that we are dealing with a mobile phone case mold for injection molding, and the mold number is XYZ123; open the injection mold database; input the mold number XYZ123; retrieve the mobile phone case product model that matches the mold number in the injection mold database; load and display the three-dimensional model of the mobile phone case.

[0066] Analyze the three-dimensional model of the mobile phone case to understand its shape and size; based on the center of gravity of the mobile phone case and production requirements, initially determine several possible material taking positions, such as the top, bottom, or side of the mobile phone case; use simulation software to simulate the situation of the grasping tool grasping the mobile phone case at different positions, and evaluate the stability and efficiency of the grasping; according to the simulation results, finally determine three theoretical material taking positions: the center of the top of the mobile phone case, and both sides of the bottom.

[0067] Map the theoretical material taking position onto the mold opening image of the injection mold; combine the center position of the mobile phone case contour and the product area information to determine the actual material taking area corresponding to each theoretical material taking position; for example, the material taking area at the center of the top may be a circular area, and the material taking areas on both sides of the bottom may be two rectangular areas.

[0068] Consider the accessibility and grasping accuracy of the material taking tool, and fine-tune the material taking area; for example, ensure that the diameter of the circular area is slightly larger than the diameter of the grasping tool to ensure stable grasping; save and transmit the determined material taking area information to the subsequent material taking system so that the robotic arm can accurately position and perform the material taking operation; through the above steps, we can accurately determine the product model, theoretical material taking position, and actual material taking area formed by the injection mold, providing key data support for subsequent intelligent material taking operations.

[0069] Reference Figure 4, in step S13, the material picking arm is within the mold opening space of the injection mold, and the relative positions of multiple suction cups in the material picking arm are determined according to multiple material picking areas and the spatial position of the material picking arm;

[0070] In the specific implementation process of the present invention, the specific steps are as follows:

[0071] S131: When the injection mold is in the mold opening state, determine the movement path of the material picking arm relative to the mold opening space according to the mold opening space of the injection mold and the current position of the material picking arm. The material picking arm moves along this movement path and enters the mold opening space of the injection mold;

[0072] S132: Collect the spatial position of the material picking arm, determine multiple current adsorption areas according to the spatial position of the material picking arm and the distribution positions of multiple suction cups relative to the material picking arm, and trigger the position adjustment of multiple suction cups according to the matching of multiple current adsorption areas and multiple material picking areas to determine the relative positions of multiple suction cups in the material picking arm.

[0073] In the embodiment of the present application, when the injection mold is in the mold opening state, determine the movement path of the material picking arm relative to the mold opening space according to the mold opening space of the injection mold and the current position of the material picking arm. The material picking arm moves along this movement path and enters the mold opening space of the injection mold, realizing the control of the material picking arm within the mold opening space.

[0074] At this time, it is confirmed that the injection mold has completed the injection process and is already in the mold opening state; the mold opening state means that the upper mold and the lower mold of the mold have been separated, exposing the molded product; this is usually detected by sensors or vision systems inside the mold to ensure that the mold has been fully opened and there is no residual pressure or locking force.

[0075] Once it is confirmed that the mold is in the mold opening state, the mold opening space of the mold needs to be determined next; the mold opening space refers to the internal space formed after the mold is opened, and its size and shape depend on the design of the mold and the size of the product; the purpose of determining the mold opening space is to ensure that the material picking arm can safely enter and contact the molded product. Optionally, use measuring tools (such as laser rangefinders or 3D scanners) to measure the internal space dimensions of the mold after it is opened.

[0076] After determining the mold opening space, it is necessary to know the current position of the material picking arm (usually a robotic arm or a similar device); this can be achieved through sensors or positioning systems installed on the material picking arm; determining the current position of the material picking arm is for planning the best path for it to move into the mold opening space; optionally, use sensors installed on the material picking arm (such as position sensors or encoders) to obtain the current position of the material picking arm; or, if the material picking arm is connected to the control system, the current position of the material picking arm can be directly obtained from the control system.

[0077] After knowing the mold opening space and the current position of the material picking arm, the next step is to determine the movement path of the material picking arm relative to the mold opening space; this path should take into account multiple factors, such as avoiding collisions with the mold or other equipment, minimizing movement time and distance, and ensuring that the material picking arm can accurately reach and contact the molded product; the determination of the movement path is usually achieved through path planning algorithms or robotic kinematics algorithms; optionally, use path planning algorithms or robotic kinematics algorithms to calculate the best movement path of the material picking arm from the current position to the mold opening space; consider the shape, size, and position of the mold opening space, as well as the kinematic limitations of the material picking arm (such as joint angles, speed limits, etc.); output the instruction of the movement path to the control system of the material picking arm so that it can move along the calculated path.

[0078] Control the material picking arm to move along the calculated movement path until it completely enters the mold opening space of the injection mold and is ready for the next material picking operation; this is usually achieved by the control system sending a movement instruction to the material picking arm, and the instruction contains parameters such as movement speed, acceleration, position, etc.; optionally, the control system receives the instruction of the movement path and converts it into a format that the material picking arm can understand (such as joint angles, motor current, etc.); the control system sends a movement instruction to the material picking arm to control its movement along the calculated path; monitor the movement process of the material picking arm to ensure that it does not collide with the mold or other equipment and accurately reaches the expected position.

[0079] Specifically, assume that we are dealing with an injection molding process for producing a mobile phone case, where the mold number is XYZ123 and the molded product is mobile phone case A; check the sensor signals inside the mold to confirm that mold XYZ123 has been opened; observe the opening of the mold to confirm that it is large enough to accommodate the entry of the material picking arm.

[0080] Use a laser rangefinder to measure the internal space dimensions of the mold after mold opening to obtain the dimension information of length x width x height; use a position sensor installed on the material picking arm to obtain its current position information, and further confirm that the material picking arm is currently in the standby position beside the mold; use a path planning algorithm to calculate the optimal movement path of the material picking arm from the standby position to the mold opening space; consider the shape and size of the mold, as well as the kinematic limitations of the material picking arm; output the instruction of the movement path to the control system of the material picking arm.

[0081] The control system receives the instruction of the movement path and converts it into a format that the material picking arm can understand; the control system sends a movement instruction to the material picking arm to control its movement along the calculated path; monitor the movement process of the material picking arm to ensure that it does not collide with the mold or other equipment and accurately reaches the expected position in the mold opening space; through the above steps, we can ensure that the material picking arm can safely and accurately enter the mold opening space of the injection mold, providing a solid foundation for the subsequent material picking operation.

[0082] Therefore, collect the spatial position of the material picking arm, determine multiple current adsorption areas according to the spatial position of the material picking arm and the distribution positions of multiple suction cups relative to the material picking arm, and trigger the position adjustment of multiple suction cups according to the matching of multiple current adsorption areas and multiple material picking areas to determine the relative positions of multiple suction cups in the material picking arm, which is compatible with the overall consideration of multiple material picking areas and the spatial position of the material picking arm, ensures the accuracy of the relative positions of multiple suction cups, and realizes the accurate adsorption of the product by multiple suction cups.

[0083] At this time, use a sensor or a vision system to collect the accurate position of the material picking arm (usually a robotic arm or a similar device with multiple suction cups installed thereon for adsorbing the molded product) in the three-dimensional space; this usually involves determining the coordinates of the end effector of the material picking arm (i.e., the part where the suction cups are installed), as well as possible pose information (such as rotation angle); the purpose of collecting the spatial position is to determine the adsorption area of the suction cups subsequently; optionally, use a position sensor installed on the material picking arm (such as a laser rangefinder, radar, inertial navigation system or vision positioning system) to obtain its spatial position information; send the collected position information to the control system for subsequent processing.

[0084] Determine the current adsorption area according to the distribution position of the suction cups relative to the material picking arm; this usually involves converting the positions of the suction cups from the coordinate system of the material picking arm to the global coordinate system (or the coordinate system of the mold), and considering the size and shape of the suction cups to determine the adsorption area that can be covered; optionally, calculate the positions of each suction cup in the global coordinate system according to the spatial position of the material picking arm and the distribution position of the suction cups on the material picking arm (which is usually determined during the design phase); consider the size and shape of the suction cups (such as circular, square, etc.) to determine the adsorption area that each suction cup can cover; merge the adsorption areas of all suction cups to obtain the current adsorption area.

[0085] Match the currently determined adsorption area with the material picking area determined in step S122 before; this usually involves comparing the positions, shapes, and sizes of the adsorption area and the material picking area to determine their corresponding relationships; optionally, use image processing algorithms or spatial geometry algorithms to compare the adsorption area and the material picking area; calculate the overlap degree, distance, or other metric indicators between the adsorption area and the material picking area; determine which adsorption areas match which material picking areas according to the comparison results.

[0086] If there is a mismatch between the adsorption area and the material picking area (such as position offset, shape mismatch, etc.), it is necessary to trigger the position adjustment mechanism of the suction cups to adjust the positions of the suction cups so that they can more accurately cover the expected material picking area; this usually involves controlling the movement of the material picking arm to change the positions and postures of the suction cups; optionally, calculate the positions and postures of the suction cups that need to be adjusted according to the matching results of the adsorption area and the material picking area; send the adjustment instructions to the control system of the material picking arm; the control system controls the movement of the material picking arm according to the adjustment instructions to adjust the positions and postures of the suction cups; monitor the adjustment process to ensure that the suction cups accurately reach the expected positions and postures.

[0087] After adjusting the positions of the suction cups, confirm the final positions of all suction cups relative to the material picking arm; this usually involves collecting the spatial position information of the material picking arm (including the suction cups) again and verifying whether the suction cups have accurately reached the expected material picking area; optionally, use the same sensors or vision systems as in step one to collect the final spatial position information of the material picking arm and its suction cups; compare the collected position information with the position information of the expected material picking area; if all suction cups have accurately reached the expected material picking area, confirm that the position adjustment is completed.

[0088] Specifically, assume that we are dealing with an injection molding production process for mobile phone cases, where the mold number is XYZ123, the molded product is mobile phone case A, and four suction cups are installed on the material picking arm to adsorb the mobile phone cases; use the vision positioning system installed on the material picking arm to collect its spatial position information; send the collected position information to the control system.

[0089] Calculate the position of each suction cup in the global coordinate system according to the spatial position of the material-taking support arm and the distribution position of the suction cups on the material-taking support arm; consider the circular shape and size of the suction cups to determine the adsorption area that each suction cup can cover; merge the adsorption areas of all suction cups to obtain the current adsorption area.

[0090] Use an image processing algorithm to compare the current adsorption area with the material-taking area of mobile phone case A determined in step S122; find that there is a slight positional offset between the current adsorption area and the material-taking area; calculate the position and posture of the suction cups that need to be adjusted according to the matching result between the current adsorption area and the material-taking area; send the adjustment instruction to the control system of the material-taking support arm; the control system controls the material-taking support arm to make fine adjustments to adjust the position and posture of the suction cups; monitor the adjustment process to ensure that the suction cups accurately reach the position near the expected material-taking area.

[0091] Use the vision positioning system again to collect the final spatial position information of the material-taking support arm and its suction cups; compare the collected position information with the expected material-taking area position information; confirm that all suction cups have accurately reached within the expected material-taking area and the position adjustment is completed; through the above steps, we can ensure that the suction cups on the material-taking support arm can accurately cover the expected material-taking area, providing a solid foundation for subsequent material-taking operations.

[0092] Reference Figure 5 , in step S14, determine the adsorption force of multiple suction cups according to the demolding force of the injection mold on the product, the relative positions of multiple suction cups, and the surface coefficients of multiple material-taking areas;

[0093] In the specific implementation process of the present invention, the specific steps are as follows:

[0094] S141: Determine multiple demolding positions according to the database of the injection mold and the arrangement positions of multiple products, and determine the demolding force of the injection mold on the product according to multiple demolding positions, the materials of multiple products, and the contact areas of multiple products relative to the injection mold;

[0095] S142: Determine the surface coefficients of multiple material-taking areas based on the surface detection of multiple material-taking areas, determine the first parameter according to the demolding force of the injection mold on the product and the relative positions of multiple suction cups, determine the second parameter according to the relative positions of multiple suction cups and the surface coefficients of multiple material-taking areas, and determine the adsorption force of multiple suction cups according to the first parameter, the second parameter, and the adsorption mapping relationship to prompt multiple suction cups to adsorb the corresponding products.

[0096] In the embodiments of the present application, multiple demolding positions are determined based on the database of the injection mold and the arrangement positions of multiple products, and the demolding force of the injection mold on the products is determined according to the multiple demolding positions, the materials of the multiple products, and the contact areas of the multiple products relative to the injection mold, ensuring the accuracy of the demolding force of the injection mold on the products.

[0097] At this time, the database of the injection mold and the arrangement positions of multiple products are introduced. The database of the injection mold is a database containing detailed mold information, usually including key information such as the geometric shape, size, material, cooling system layout, number and layout of cavities, etc.; this information is crucial for determining the demolding positions.

[0098] In the mold design stage, the arrangement positions of the products have already been determined; this involves the specific placement methods of the products in the mold cavities, including directions, angles, and relative positions, etc.; this information is usually recorded in the mold design drawings or databases.

[0099] Combining the mold database and the arrangement positions of the products, the demolding positions of each product in the mold can be accurately calculated; the demolding position refers to the specific position where the product separates from the mold, usually located at the opening of the mold or on the parting surface. Optionally, consult the database of the injection mold to obtain information such as the geometric shape, size, and cavity layout of the mold; determine the arrangement positions of the products in the mold according to the information in the product design drawings or databases; use CAD software or mold design software to calculate the demolding positions of each product by combining the mold database and the product arrangement positions.

[0100] Multiple demolding positions, the materials of multiple products, and the contact areas of multiple products relative to the injection mold are introduced. Different demolding positions will affect the force required for demolding; for example, products located at the edge of the mold may require a greater demolding force to overcome the edge effect and frictional resistance; at the same time, the frictional forces between products of different materials and the mold are different; for example, some plastic materials will adhere tightly to the mold after cooling and require a greater demolding force to separate; in addition, the larger the contact area between the product and the mold, the greater the demolding force usually required; this is because the increase in the contact area will lead to an increase in frictional resistance. The demolding force of the injection mold on the products is determined according to the multiple demolding positions, the materials of the multiple products, and the contact areas of the multiple products relative to the injection mold. Optionally, combine information such as demolding positions, product materials, and contact areas, and use an empirical formula to calculate the demolding force required for each mobile phone case; assume the empirical formula is F = kSμ, where F is the demolding force, k is a constant (preset value), S is the contact area, and μ is the friction coefficient (determined according to the friction characteristics of ABS plastic and the mold); substitute the known values for calculation to obtain that the demolding force required for each product is approximately XX Newtons.

[0101] Therefore, the surface coefficients of multiple material taking areas are determined based on the surface detection of multiple material taking areas. The first parameter is determined according to the demolding force of the injection mold on the product and the relative positions of multiple suction cups. The second parameter is determined according to the relative positions of multiple suction cups and the surface coefficients of multiple material taking areas. And the adsorption forces of multiple suction cups are determined according to the first parameter, the second parameter and the adsorption mapping relationship, so as to prompt multiple suction cups to adsorb the corresponding products, ensuring the accuracy of the adsorption forces of multiple suction cups.

[0102] At this time, the surface coefficients of multiple material taking areas are determined based on the surface detection of multiple material taking areas, and the surface coefficients of multiple material taking areas are introduced. At the same time, a high-precision measuring device (such as a laser scanner, a structured light sensor or a three-dimensional vision system) is used to scan the material taking areas to obtain their three-dimensional shape data; these data are usually presented in the form of point clouds or mesh models and can accurately reflect the surface characteristics of the material taking areas;

[0103] The surface coefficients are calculated based on the obtained three-dimensional shape data; the surface coefficient is a quantitative index used to describe the characteristics such as the complexity of the surface and the curvature change. Optionally, a laser scanner or other high-precision measuring device is used to comprehensively scan the material taking areas; the obtained three-dimensional shape data is imported into the surface coefficient calculation software for processing and analysis; according to the custom script built in the surface coefficient calculation software, the surface coefficients of the material taking areas are calculated.

[0104] Regarding the first parameter, the first parameter is determined according to the demolding force of the injection mold on the product and the relative positions of multiple suction cups. At this time, the demolding force of the injection mold on the product is one of the important factors affecting the adsorption force of the suction cup; the greater the demolding force, it means that the greater the resistance the product may encounter when separating from the mold, so the suction cup needs a greater adsorption force to ensure stable grasping.

[0105] The relative positions of the suction cups on the material taking support arm will also affect their adsorption forces; for example, the suction cups located at the edge of the product may need a greater adsorption force to overcome the edge effect, while the suction cups located at the center of the product may be subjected to less mechanical challenges. The demolding force of the injection mold on the product and the relative positions of multiple suction cups are input into a preset parameter training model, and the first parameter is output by the preset parameter training model. The first parameter is usually a quantitative value used to describe the required adsorption force level of the suction cup under specific positions and demolding force conditions.

[0106] For the second parameter, the second parameter is determined according to the relative positions of multiple suction cups and the surface coefficients of multiple material picking areas. At this time, the relative positions of multiple suction cups are introduced. For example, the interaction between adjacent suction cups, the distance between the suction cup and the edge of the material picking area, etc. may all affect its adsorption effect. The surface coefficient of the material picking area also has an important impact on the adsorption force of the suction cup. The more complex the surface, the greater the challenges that the suction cup may encounter during the adsorption process. Therefore, a greater adsorption force is required to ensure stable grasping.

[0107] The relative positions of multiple suction cups and the surface coefficients of multiple material picking areas are input into a preset parameter training model, and the second parameter is output by the preset parameter training model. The second parameter is also a quantization value, which is used to describe the adsorption force level required by the suction cup under specific surface and suction cup layout conditions.

[0108] The adsorption mapping relationship refers to the one-to-one correspondence between the suction cup and the material picking area. This relationship has been determined during the design stage of the material picking arm and is recorded in relevant drawings or databases. Considering the first parameter, the second parameter, and the adsorption mapping relationship comprehensively, the adsorption force required for each suction cup is calculated through an optimization method. This adsorption force should be large enough to ensure stable grasping of the product, but should not be too large to avoid damaging the product or increasing energy consumption. Optionally, the adsorption mapping relationship covers the corresponding relationship between the first parameter, the second parameter, and the adsorption force, and the adsorption force is determined by the matching of the first parameter, the second parameter, and the adsorption mapping relationship.

[0109] Specifically, assume that we are dealing with an injection molding production process of mobile phone cases, where the mold can produce two mobile phone cases simultaneously. Four suction cups are installed on the material picking arm, which are respectively used to adsorb the four corners of the two mobile phone cases. The material of the mobile phone case is ABS plastic, which has certain elasticity and adhesiveness. The material picking area is scanned using a laser scanner to obtain the three-dimensional shape data of the mobile phone case. The scanned data is processed and analyzed in software to calculate the surface coefficient of the material picking area. Assume that the surface coefficients of the two mobile phone cases are 0.8 and 0.9 respectively (the larger the value, the more complex the surface).

[0110] Refer to the demolding force data of the injection mold and find that the demolding forces of the two mobile phone cases are 50N and 60N respectively. Analyze the relative positions of the suction cups on the material picking arm and find that suction cup 1 and suction cup 2 are respectively located at two diagonal positions of mobile phone case 1, and suction cup 3 and suction cup 4 are located at two diagonal positions of mobile phone case 2. Combining the demolding force and the suction cup position information, calculate the first parameter. Assume that the first parameters obtained by the algorithm are 70N and 80N respectively (corresponding to the two mobile phone cases).

[0111] Analyze the relative position relationship between the suction cups again and find that the distance between adjacent suction cups is appropriate and will not affect each other; combining the surface coefficient information (0.8 and 0.9), use an algorithm to calculate the second parameter; assume that the second parameters obtained by the algorithm are 75N and 85N (corresponding to the material-taking areas of two mobile phone cases).

[0112] Consult the adsorption mapping relationship data to determine the material-taking area corresponding to each suction cup; combining the first parameters (70N and 80N), the second parameters (75N and 85N) and the adsorption mapping relationship information, use an optimization method to calculate the adsorption force of each suction cup; assume that the finally obtained adsorption forces are 72N, 78N, 82N and 87N (corresponding to suction cups 1, 2, 3, and 4 respectively); input the calculation results into the control system of the material-taking support arm to adjust the adsorption force of the suction cup during actual production; in this way, it can be ensured that the suction cup can stably grasp the mobile phone case during the adsorption process without damaging the product or increasing unnecessary energy consumption.

[0113] In an embodiment of the present application, use a high-precision measuring device (such as a laser scanner) to scan the material-taking area to obtain its three-dimensional shape data; based on the scan data, analyze and calculate the surface coefficient of each material-taking area through an algorithm; the surface coefficient reflects the complexity of the surface, such as curvature changes, concavity and convexity, etc.; the surface coefficient matching table is shown in Table 1:

[0114] Table 1 Coefficient matching table

[0115] Feeding area number Surface coefficient A1 0.75 A2 0.85 B1 0.65 B2 0.90

[0116] Consult the demolding force data of the injection mold to understand the demolding requirements of each product; analyze the relative position of the suction cup on the material-taking support arm and consider its influence on the adsorption force; the first parameter matching table is shown in Table 2:

[0117] Table 2 First parameter matching table

[0118] Product number Demolding force (N) Suction cup position First parameter (N) P1 50 Near the center 60 P2 60 Near the edge 70

[0119] Assume that suction cup 1 is located near the center position, corresponding to the material-taking area A1 with a surface coefficient of 0.75; suction cup position score (assuming a full score of 10 points): 8 points for the near center position; surface coefficient score (assuming a full score of 10 points, the closer the surface coefficient is to 1, the higher the score): 0.75 corresponds to 7.5 points;

[0120] Second parameter calculation: Second parameter = (suction cup position score × suction cup position weight + surface coefficient score × surface coefficient weight) × conversion coefficient (the conversion coefficient is used to convert the score into an actual adsorption force value); assume that the conversion coefficient is 10N / point, then the second parameter = (8 × 0.6 + 7.5 × 0.4) × 10 = 78N (this value is an example).

[0121] According to the adsorption mapping relationship, determine the material taking area and product corresponding to each suction cup; combine the first parameter, the second parameter and the adsorption mapping relationship to calculate the final adsorption force of each suction cup.

[0122] The adsorption force matching table is shown in Table 3:

[0123] Table 3 Adsorption Force Matching Table

[0124]

[0125] Reference Figure 6 , in step S15, determine the intelligent material taking system of the product according to the movement trajectories of multiple suction cups, the attitude parameters of the product, and the adsorption forces of multiple suction cups, and generate intelligent control measures for the product during the material taking process in the intelligent material taking system.

[0126] In the specific implementation process of the present invention, the specific steps are as follows:

[0127] S151: When the product is adsorbed by multiple suction cups, determine the movement trajectories of multiple suction cups according to the current positions of multiple suction cups and the discharging position of the injection mold, mark multiple trajectory nodes in the movement trajectory, and determine the corresponding attitude parameters of the product based on multiple trajectory nodes and the corresponding product movement images.

[0128] S152: Determine the first training combination according to multiple trajectory nodes and the adsorption forces of multiple suction cups, determine the second training combination according to multiple trajectory nodes and the corresponding attitude parameters, and determine the intelligent material taking system of the product through the learning and training of the first training combination and the second training combination.

[0129] S153: In the intelligent material taking system of the product, determine the abnormal information of multiple suction cups relative to the product according to the adsorption forces of multiple suction cups at the same time node, and determine the corresponding intelligent control measures according to the abnormal information, the corresponding suction cup and the control mapping relationship.

[0130] In the embodiment of the present application, when the product is adsorbed by multiple suction cups, determine the movement trajectories of multiple suction cups according to the current positions of multiple suction cups and the discharging position of the injection mold, mark multiple trajectory nodes in the movement trajectory, and determine the corresponding attitude parameters of the product based on multiple trajectory nodes and the corresponding product movement images, and introduce the attitude parameters of the product.

[0131] At this time, the system needs to obtain the current accurate position of each suction cup; this is usually achieved through sensors integrated on the robot or the material taking mechanism, such as position encoders, laser rangefinders or vision positioning systems; the system needs to analyze the discharging position of the injection mold; this usually involves a deep understanding of the mold structure and the discharging mechanism to ensure that the suction cup can accurately reach and adsorb the product.

[0132] Based on the current position of the suction cup and the product discharge position of the injection mold, the system uses path planning algorithms (such as A* algorithm, Dijkstra algorithm, or genetic algorithm) to calculate the optimal movement trajectory; this trajectory should ensure that the suction cup can reach and adsorb the product smoothly and quickly, while avoiding collisions with other equipment on the production line; to improve efficiency and reduce energy consumption, the system also performs smoothing and optimization processing on the calculated movement trajectory.

[0133] On the determined movement trajectory, the system needs to select a series of key trajectory nodes; these trajectory nodes can be key points, turning points, target points, or areas that require special attention on the path; each selected trajectory node will be marked and stored in the system; parameters such as the position, speed, and acceleration of these nodes will be recorded for subsequent analysis and use; the system will also establish the relationships between the trajectory nodes, such as the distance between adjacent nodes and the direction change; this information is crucial for determining the product attitude parameters and formulating intelligent control measures in the future.

[0134] During the process of the suction cup moving and adsorbing the product, the system needs to use visual sensors or cameras to capture the moving images of the product; these images should contain sufficient information to identify the attitude of the product; the captured images will be analyzed through image processing algorithms (such as edge detection, feature extraction, image matching, etc.); these algorithms can identify the key features in the images and calculate the attitude parameters of the product; based on the results of image processing, the system can determine the attitude parameters of the product; these parameters may include the tilt angle, rotation direction, position offset, etc. of the product; these parameters are crucial for formulating intelligent control measures in the future; the determined attitude parameters will be stored in the system and updated continuously with the movement and adsorption process of the product; optionally, use visual sensors or cameras to capture the moving images of the product; perform image processing and analysis on the captured images to identify the key features; determine the attitude parameters of the product based on the results of image processing; store the attitude parameters in the system and update them with the process.

[0135] Specifically, assume that on an automated injection production line, the system needs to adsorb and move an injection-molded mobile phone case; the system first obtains the current position of the suction cup and analyzes the product discharge position of the injection mold; then, it uses a path planning algorithm to calculate an optimal movement trajectory that starts from the current position of the suction cup, passes through a series of key points, and finally reaches the product discharge position of the mobile phone case; to improve efficiency, the system also performs smoothing and optimization processing on the trajectory.

[0136] On the determined moving trajectory, the system selects several key trajectory nodes, including the starting point, turning points (such as points for avoiding other devices), and the target point (the discharging position of the phone case); the parameters of each trajectory node, such as position, speed, and acceleration, are recorded, and the relationships between the trajectory nodes are established.

[0137] During the process of the suction cup moving and adsorbing the phone case, the system uses a vision sensor to capture the moving images of the phone case; then, image processing and analysis are performed on the captured images to identify the key features of the phone case (such as edges, corner points, etc.); based on these features, the system calculates the attitude parameters of the phone case, including the tilt angle, rotation direction, and position offset; these parameters are stored in the system and are continuously updated during the movement and adsorption process of the phone case; by implementing step S151, the system can accurately plan the moving trajectory of the suction cup, mark the key trajectory nodes, and determine the attitude parameters of the product, which provide an important basis for formulating subsequent intelligent control measures.

[0138] Furthermore, a first training combination is determined according to multiple trajectory nodes and the adsorption forces of multiple suction cups, a second training combination is determined according to multiple trajectory nodes and the corresponding attitude parameters, and the intelligent material taking system of the product is determined through the learning and training of the first training combination and the second training combination, ensuring the accuracy of the intelligent material taking system of the product.

[0139] At this time, the system needs to collect the adsorption force data of multiple suction cups at multiple trajectory nodes and corresponding time points; these data can be obtained in real time through sensors integrated on the robot or the material taking mechanism; the collected data may contain noise or outliers, so preprocessing is required; this includes steps such as data cleaning (removing outliers, filling in missing values, etc.), data normalization (scaling the data to the same range), and data transformation (such as logarithmic transformation, square root transformation, etc.).

[0140] The system needs to extract features from the preprocessed data; these features may include the position, speed, and acceleration of the trajectory nodes, as well as the adsorption force of the suction cup, etc.; these features will be used for subsequent machine learning model training; based on the extracted features, the system constructs a first training combination; this combination will contain multiple samples, each sample consisting of a feature vector and a corresponding label (such as adsorption stability index, energy consumption index, etc.); these samples will be used to train the machine learning model to predict or optimize the adsorption force of the suction cup;

[0141] For the second training combination, determine the second training combination based on multiple trajectory nodes and corresponding pose parameters. At this time, similar to the first training combination, the system needs to collect data on the pose parameters of the product at multiple trajectory nodes and corresponding time points; this data can also be obtained in real time through sensors or vision systems and preprocessed; extract features from the preprocessed data, which may include the position and direction changes of the trajectory nodes, as well as pose parameters such as the tilt angle and rotation direction of the product. Based on the extracted features, the system constructs the second training combination; this combination also contains multiple samples, each sample consisting of a feature vector and a corresponding label (such as pose stability index, product positioning accuracy, etc.); these samples will be used to train a machine learning model to predict or adjust the pose of the product.

[0142] Furthermore, select a suitable machine learning algorithm according to the specific nature of the problem and the characteristics of the data; these algorithms may include regression algorithms (such as linear regression, ridge regression, etc.), classification algorithms (such as logistic regression, support vector machine, etc.), clustering algorithms (such as K-means, DBSCAN, etc.) or reinforcement learning algorithms, etc.;

[0143] Use the data in the first training combination and the second training combination to train the machine learning model; during the training process, the system needs to continuously adjust the parameters of the model to minimize the prediction error or maximize the performance index; after training, the system needs to verify and optimize the model; this includes using strategies such as cross-validation and holdout method to evaluate the performance of the model and optimize the model according to the evaluation results; based on the trained machine learning model, the system can determine the intelligent material picking system; this system can automatically adjust the movement trajectory, adsorption force and pose adjustment strategy of the suction cup according to different products and production conditions to achieve an efficient and stable material picking process. Optionally, the system selects a suitable machine learning algorithm (such as a regression algorithm or a reinforcement learning algorithm) and uses the data in the first training combination and the second training combination to train the model; during the training process, the system continuously adjusts the parameters of the model to minimize the prediction error; after training, the system verifies and optimizes the model to ensure that the model has good generalization ability and stability; finally, based on the trained model, the system determines the intelligent material picking system; this intelligent material picking system can automatically adjust the movement trajectory, adsorption force and pose adjustment strategy of the suction cup according to different mobile phone case types and production conditions to achieve an efficient and stable material picking process.

[0144] Specifically, assume that on an automated injection molding production line, the system needs to construct an intelligent material picking system to adsorb and move the injection-molded mobile phone cases.

[0145] Determine the first training set: The system first collected the adsorption force data of the suction cups at multiple trajectory nodes; this data included the adsorption force values of the suction cups at different positions and different speeds; then, the system preprocessed and extracted features from the data, extracting features such as the position, speed, acceleration of the trajectory nodes and the adsorption force of the suction cups; next, the system constructed the first training set, where each sample consisted of a feature vector composed of these features and a corresponding adsorption stability index label.

[0146] Determine the second training set: The system also collected the attitude parameter data of the mobile phone case at multiple trajectory nodes; this data included the tilt angle, rotation direction, etc. of the mobile phone case at different positions and different attitudes; the system preprocessed, extracted features from, and enhanced the data, extracting features such as the position, direction change of the trajectory nodes and the attitude parameters of the mobile phone case, and introducing derivative features such as the rate of change of speed and the rate of change of acceleration; then, the system constructed the second training set, where each sample consisted of a feature vector composed of these features and a corresponding attitude stability index label.

[0147] The system selected a suitable machine learning algorithm (such as a regression algorithm or a reinforcement learning algorithm) and used the data in the first training set and the second training set to train the model; during the training process, the system continuously adjusted the parameters of the model to minimize the prediction error; after the training was completed, the system verified and optimized the model to ensure that the model had good generalization ability and stability; finally, based on the trained model, the system determined the intelligent material taking system; this system could automatically adjust the movement trajectory, adsorption force, and attitude adjustment strategy of the suction cups according to different mobile phone case types and production conditions to achieve an efficient and stable material taking process.

[0148] Therefore, in the intelligent material taking system of the product, the abnormal information of multiple suction cups relative to the product is determined according to the adsorption force of multiple suction cups at the same time node, and the corresponding intelligent control measures are determined according to this abnormal information, the corresponding suction cups, and the control mapping relationship, so as to enable multiple suction cups to perform intelligent adsorption on the product to overcome the detachment of the product during the material taking process and achieve the intelligent material taking effect of the injection mold.

[0149] At this time, in the intelligent material taking system of the product, the system needs to collect the adsorption force data of multiple suction cups at the same time node in real time; this data is usually obtained through high-precision sensors integrated on the robot or the material taking mechanism; in order to judge whether the adsorption force of the suction cup is normal, the system needs to preset a reasonable adsorption force range or threshold; this range or threshold can be adjusted according to factors such as product type, production conditions, and suction cup type.

[0150] Compare the adsorption force data of each suction cup collected with a preset threshold value; if the adsorption force of a certain suction cup exceeds the normal range, it is determined as abnormal; for the detected abnormal suction cup, the system needs to record its abnormal information, including the time of abnormality occurrence, the number of the abnormal suction cup, the specific value of the adsorption force, etc.

[0151] In the intelligent material handling system, it is necessary to establish a mapping relationship between abnormal information and intelligent control measures in advance; this mapping relationship can be customized according to factors such as product type, production requirements, safety specifications, etc.; when the system detects abnormal information, it will automatically match the pre-established control mapping relationship to find the corresponding intelligent control measure; once the intelligent control measure is determined, the system will convey the measure to the robot or the material handling mechanism through a control signal to perform the corresponding operation; during the execution of the control measure, the system needs to continuously monitor the adsorption force of the suction cup and the state of the product to ensure the effectiveness of the control measure; if the control measure fails to achieve the expected effect, the system may need to adjust the measure or trigger a further emergency response. Optionally, establish a mapping relationship between abnormal information and intelligent control measures; automatically match control measures when detecting abnormal information; execute control measures and ensure the effect through monitoring; adjust measures or trigger an emergency response according to the monitoring results.

[0152] Specifically, assume that on an automated injection molding production line, the system is performing an intelligent material handling task for mobile phone cases; the system collects the adsorption force data of four suction cups at the same time node in real time; after comparing with the preset adsorption force threshold value, the system finds that the adsorption force of the third suction cup is significantly lower than the normal range, so it is determined as abnormal; the system records this abnormal information, including the time of abnormality occurrence, the number of the third suction cup, and the specific value of the adsorption force.

[0153] According to the pre-established control mapping relationship, the system finds the intelligent control measures corresponding to the abnormal adsorption force of the third suction cup; these measures include steps such as pausing the material handling task, reducing the moving speed of the robotic arm where the suction cup is located, and increasing the adsorption force of the suction cup; the system conveys these measures to the robot through a control signal to perform the corresponding operation.

[0154] During the execution of the control measure, the system continuously monitors the adsorption force of the third suction cup and the state of the mobile phone case; after a period of time, the system finds that the adsorption force of this suction cup gradually returns to normal, and the position and posture of the mobile phone case remain stable; this indicates that the control measure is effective, and the system continues to perform the subsequent material handling task; if the monitoring results show that the control measure fails to achieve the expected effect, the system may adjust the measure, such as further increasing the adsorption force or triggering an emergency response such as an emergency stop.

[0155] In an embodiment of the present application, it is assumed that there are four suction cups (A, B, C, D) in the system, and each suction cup has a preset adsorption force range; the following is an abnormal information matching table, as shown in Table IV:

[0156] Table IV Abnormal Information Matching Table

[0157] Abnormal information Suction cup number Control measures Suction force too low A Increase the suction force of suction cup A Suction force too high B Reduce the suction force of suction cup B Suction force fluctuates too much C Pause feeding and check suction cup C and connecting parts No abnormality - Continue with the feeding task Abnormal suction forces of multiple suction cups simultaneously A, B, C Emergency stop and comprehensively check the feeding system

[0158] In this example, when the system detects that the adsorption force of suction cup A is too low, it will automatically match the control measure of "increasing the adsorption force of suction cup A" and perform the corresponding operation; in this way, the system can more flexibly take appropriate intelligent control measures according to the abnormal situation of the adsorption force of the suction cups to ensure the smooth progress of the material taking task.

[0159] Please refer to Figure 7 , Figure 7 which is a schematic structural composition diagram of the intelligent material taking system of the injection mold in the embodiment of the present invention; the intelligent material taking system of the injection mold includes:

[0160] The acquisition module 21 is used to acquire the mold opening image of the injection mold;

[0161] The material taking area module 22 is used to determine a plurality of material taking areas according to the mold opening image of the injection mold and the product model formed by the injection mold;

[0162] The position module 23 is used to determine the relative positions of a plurality of suction cups in the material taking arm according to a plurality of material taking areas and the spatial position of the material taking arm when the material taking arm is in the mold opening space of the injection mold;

[0163] The adsorption force module 24 is used to determine the adsorption forces of a plurality of suction cups according to the demolding force of the injection mold on the product, the relative positions of the plurality of suction cups, and the surface coefficients of the plurality of material taking areas;

[0164] The intelligent material taking module 25 is used to determine the intelligent material taking system of the product according to the movement trajectories of the plurality of suction cups, the attitude parameters of the product, and the adsorption forces of the plurality of suction cups, and generate intelligent control measures for the product during the material taking process in the intelligent material taking system.

[0165] Arbitrary combinations of the technical features of the above embodiments are made. For the sake of brevity of description, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

Claims

1. An intelligent material taking method for an injection mold, characterized in that, Including: Collect the mold opening image of the injection mold; Determine multiple material taking areas according to the mold opening image of the injection mold and the product model formed by the injection mold; The material taking arm is in the mold opening space of the injection mold. Determine the relative positions of multiple suction cups in the material taking arm according to the multiple material taking areas and the spatial position of the material taking arm; Determine the adsorption forces of multiple suction cups according to the demolding force of the product by the injection mold, the relative positions of multiple suction cups, and the surface coefficients of multiple material taking areas; Determine the intelligent material taking system of the product according to the moving trajectories of multiple suction cups, the attitude parameters of the product, and the adsorption forces of multiple suction cups, and generate intelligent control measures for the product during the material taking process in the intelligent material taking system.

2. The intelligent material taking method of the injection mold according to claim 1, characterized in that The collecting the mold opening image of the injection mold includes: When the injection mold is in the mold opening state, collect the mold opening space of the injection mold, and determine the product area of the injection mold according to the mold opening space of the injection mold and the cavity position of the injection mold; Trigger the responses of multiple cameras according to the product area of the injection mold and the mold opening direction of the injection mold, construct a shooting system of the injection mold based on the multiple cameras, and determine the mold opening image of the injection mold according to the shooting of the product area of the injection mold by the shooting system. At this time, multiple cameras are distributed around the injection mold.

3. The intelligent material taking method of the injection mold according to claim 1, characterized in that The determining multiple material taking areas according to the mold opening image of the injection mold and the product model formed by the injection mold includes: Generate multiple functional areas according to the division of the mold opening image of the injection mold, determine the product area based on the screening of the functional areas, determine multiple product contours according to the preprocessing of the product area, and mark the central positions of the multiple product contours; Determine the product model formed by the injection mold according to the database of the injection mold and the mold number information of the injection mold, determine multiple theoretical material taking positions of the product model according to the detection of the product model formed by the injection mold, and determine multiple material taking areas according to the multiple theoretical material taking positions, the central positions of the multiple product contours, and the product area.

4. The intelligent material taking method of the injection mold according to claim 1, wherein, The material taking arm is in the mold opening space of the injection mold. The determining the relative positions of multiple suction cups in the material taking arm according to the multiple material taking areas and the spatial position of the material taking arm includes: When the injection mold is in the mold opening state, determine the moving path of the material taking arm relative to the mold opening space according to the mold opening space of the injection mold and the current position of the material taking arm. The material taking arm moves along the moving path and enters the mold opening space of the injection mold.

5. The intelligent material taking method of the injection mold according to claim 4, characterized in that, The material taking arm is in the mold opening space of the injection mold. The determining the relative positions of multiple suction cups in the material taking arm according to the multiple material taking areas and the spatial position of the material taking arm further includes: Collect the spatial position of the material taking arm, determine multiple current adsorption areas according to the spatial position of the material taking arm and the distribution positions of multiple suction cups relative to the material taking arm, and trigger the position adjustment of multiple suction cups according to the matching of the multiple current adsorption areas and the multiple material taking areas to determine the relative positions of multiple suction cups in the material taking arm.

6. The intelligent material taking method of the injection mold according to claim 1, wherein, The determining the adsorption forces of multiple suction cups according to the demolding force of the product by the injection mold, the relative positions of multiple suction cups, and the surface coefficients of multiple material taking areas includes: Determine multiple demolding positions based on the database of the injection mold and the arrangement positions of multiple products, and determine the demolding force of the injection mold on the products according to the multiple demolding positions, the materials of the multiple products, and the contact areas of the multiple products with respect to the injection mold.

7. The intelligent material taking method of the injection mold according to claim 6, characterized in that The determination of the adsorption forces of multiple suction cups according to the demolding force of the injection mold on the products, the relative positions of the multiple suction cups, and the surface coefficients of the multiple material-taking areas further includes: Determine the surface coefficients of the multiple material-taking areas based on the surface detection of the multiple material-taking areas, determine a first parameter according to the demolding force of the injection mold on the products and the relative positions of the multiple suction cups, determine a second parameter according to the relative positions of the multiple suction cups and the surface coefficients of the multiple material-taking areas, and determine the adsorption forces of the multiple suction cups according to the first parameter, the second parameter, and the adsorption mapping relationship, so as to prompt the multiple suction cups to adsorb the corresponding products.

8. The intelligent material taking method of the injection mold according to claim 1, characterized in that The determination of the intelligent material-taking system of the product according to the movement trajectories of the multiple suction cups, the attitude parameters of the product, and the adsorption forces of the multiple suction cups, and the generation of intelligent control measures for the product during the material-taking process in the intelligent material-taking system includes: When the product is adsorbed by the multiple suction cups, determine the movement trajectories of the multiple suction cups according to the current positions of the multiple suction cups and the discharging position of the injection mold, mark multiple trajectory nodes in the movement trajectories, and determine the corresponding attitude parameters of the product based on the multiple trajectory nodes and the corresponding product movement images; Determine a first training combination according to the multiple trajectory nodes and the adsorption forces of the multiple suction cups, determine a second training combination according to the multiple trajectory nodes and the corresponding attitude parameters, and determine the intelligent material-taking system of the product through the learning and training of the first training combination and the second training combination.

9. The intelligent material taking method of the injection mold according to claim 8, characterized in that, The determination of the intelligent material-taking system of the product according to the movement trajectories of the multiple suction cups, the attitude parameters of the product, and the adsorption forces of the multiple suction cups, and the generation of intelligent control measures for the product during the material-taking process in the intelligent material-taking system further includes: In the intelligent material-taking system of the product, determine the abnormal information of the multiple suction cups with respect to the product according to the adsorption forces of the multiple suction cups at the same time node, and determine the corresponding intelligent control measures according to the abnormal information, the corresponding suction cup, and the control mapping relationship.

10. An intelligent material taking system for an injection mold, characterized in that, The intelligent material-taking system of the injection mold is applied to the intelligent material-taking method of the injection mold as described in any one of claims 1-9. The intelligent material-taking system of the injection mold includes: An acquisition module, configured to acquire the mold-opening image of the injection mold; A material-taking area module, configured to determine multiple material-taking areas according to the mold-opening image of the injection mold and the product model formed by the injection mold; A position module, configured to determine the relative positions of multiple suction cups in the material-taking arm according to the multiple material-taking areas and the spatial position of the material-taking arm when the material-taking arm is in the mold-opening space of the injection mold; An adsorption force module, configured to determine the adsorption forces of the multiple suction cups according to the demolding force of the injection mold on the products, the relative positions of the multiple suction cups, and the surface coefficients of the multiple material-taking areas; The intelligent material taking module is used to determine the intelligent material taking system of the product according to the movement trajectories of multiple suction cups, the attitude parameters of the product, and the adsorption forces of the multiple suction cups, and generate intelligent control measures for the product during the material taking process in the intelligent material taking system.

Images