Multi-station self-adaptive material taking manipulator of injection molding machine

Abstract

The application discloses a multi-station self-adaptive material taking manipulator of an injection molding machine, comprising a six-axis driving system and a material taking system. The material taking system comprises a flexible strip and a plurality of suction cups. The flexible strip is composed of a flexible outer wall and a magneto-rheological fluid, and can freely bend when not powered. After being powered, the magneto-rheological fluid solidifies to lock the shape. Each suction cup is integrated with a spherical hinge structure, a magneto-rheological damper, a linear servo motor and a pressure sensor to form an active compliant unit. The flexible strip realizes quick switching between flexible fitting and rigid support through the magneto-rheological fluid. The active compliant unit independently regulates and controls the posture of the suction cup to accurately adapt to the local characteristics of the curved surface. In combination with target function optimization and real-time pressure feedback, the application ensures uniform distribution of the grabbing force, avoids damage to the workpiece, and improves the grabbing stability of the complex curved surface product and the multi-station operation efficiency.

Description

Technical Field

[0001] This invention relates to industrial robots and robotic gripping technology, and more particularly to a multi-station adaptive material handling robotic arm for injection molding machines. Background Technology

[0002] In the field of industrial automation, injection-molded products typically require multi-station material handling and transport using robotic arms. Traditional robotic arms generally employ rigid grippers as material handling devices, which are simple in structure and suitable for workpieces with regular geometric shapes. However, with the increasing demand for more complex injection-molded products (such as curved surfaces and thin-walled parts), existing technologies have the following drawbacks:

[0003] Rigid grippers are difficult to adapt to the complex shapes of curved injection molded products. Especially when gripping workpieces with large changes in surface curvature, uneven distribution of clamping force may lead to excessive local stress, causing scratches, deformation or even damage to the product.

[0004] Traditional robotic arms rely on overall posture adjustment (such as six-axis linkage) to match the shape of the workpiece, but their response speed and flexibility are limited, making it difficult to compensate for small deviations in local curved surfaces in real time.

[0005] Existing robotic arms lack the ability to actively adjust the pressure distribution during the gripping process, resulting in decreased gripping stability when dealing with injection molded products of different sizes or materials, and making them prone to slippage or clamping failure.

[0006] Furthermore, traditional robotic arm gripping systems typically use fixed suction cups or rigid grippers for material handling. However, their fixed structural rigidity prevents them from dynamically adjusting the suction cup posture or gripping force according to the workpiece's curved surface shape, resulting in a low success rate. This is especially problematic in multi-station production scenarios where robotic arms need to frequently switch gripping positions, which is quite cumbersome. Summary of the Invention

[0007] Based on the above-mentioned technical problems, this invention proposes a multi-station adaptive material handling robot for injection molding machines.

[0008] The technical solution of this invention is implemented as follows:

[0009] A multi-station adaptive material handling robot for injection molding machines includes a six-axis drive system and a material handling system. The six-axis drive system drives the material handling system through multiple degrees of freedom.

[0010] The material handling system is characterized by comprising a flexible strip, with multiple suction cups disposed at the bottom of the flexible strip, and the flexible strip consisting of a flexible outer wall and a magnetorheological fluid.

[0011] Among them, multiple suction cups integrate ball joint structures, magnetorheological dampers, linear servo motors, and pressure sensors at the bottom of the flexible strip, forming an active compliance unit.

[0012] The specific steps are as follows:

[0013] Step 1: Target detection, identifying the gripping surface of the curved injection molded product;

[0014] Step 2: Calculate the initial gripping posture based on the detected curved injection molded product;

[0015] Step 3: Define the suction cup that is completely in contact with the gripping surface of the curved injection molded product as the core suction cup, and define the suction cup that is not completely in contact with the gripping surface of the curved injection molded product as the non-core suction cup.

[0016] Step 4: Adjust the posture of non-core suction cups to optimize gripping stability;

[0017] Step 5: Evaluate the gripping strength of all suction cups to determine the probability of successful gripping.

[0018] In this invention, the gripping surface of the curved injection molded product is scanned to obtain the gripping curved surface S.

[0019] Establish an objective function that includes the distance errors from all suction cups to the gripping surface S, as well as the angular errors between the suction cup axis direction and the normal vector of the gripping surface S.

[0020] The objective function is:

[0021]

[0022] Where the number of suction cups is N, and T represents the transformation matrix of multiple degrees of freedom of all suction cups;

[0023] Distance error term: p i This represents the local position coordinates of the i-th suction cup. This means that the local position coordinates of the suction cup are transformed into the global position coordinates of the suction cup using the transformation matrix T;

[0024] This represents the distance function from the point to the grasping surface S. Represents the weighting coefficient of the distance term;

[0025] Angle error term: m is the projection point q i The surface normal vector at point q i for The projection point on the grasping surface S is determined by the distance function. The nearest surface point obtained through calculation;

[0026] n i This represents the local axial direction of the i-th suction cup, which is multiplied by T to obtain the axial direction.

[0027] The measure of angular error is the dot product of two vectors, equal to... , Let be the angle between the two vectors.

[0028] In this invention, Ceres / PyTorch is used to construct residuals, automatically differentiate and optimize the transformation matrix T, and then output the optimal transformation matrix T. Based on the optimal transformation matrix T, the initial grasping posture is output.

[0029] In this invention, a pressure threshold Q is set. th For each suction cup, measure its pressure sensor reading Q. i ,

[0030] If Q i Greater than Q th If Q is the core suction cup, then it is determined to be the core suction cup; if Q i Less than Q th If so, it is determined to be a non-core suction cup.

[0031] In this invention, the linear servo motor adjusts the displacement along the axial direction. Until Q i Approaching Q th Set pressure threshold,

[0032] The calculation formula is as follows:

[0033]

[0034] Indicates pressure deviation, where K p K i K d This indicates the adjustment parameters, where the tilt angle of the suction cup is adjusted through a ball joint structure during the adjustment of the linear servo motor.

[0035] In this invention, the material handling system includes a mounting shaft that engages with the wrist arm.

[0036] An opening and closing motor is provided on the side of the mounting shaft, and a first transmission gear is provided on the output shaft of the opening and closing motor. The first transmission gear meshes with a second transmission gear.

[0037] The second transmission gear is rotatably connected to the bottom of the mounting shaft, and the mounting shaft passes through the second transmission gear and is fixedly connected to the cage.

[0038] In this invention, a worm gear is fixedly connected to the middle of the cage, and the worm gear is fixedly connected to the second transmission gear and rotates with the second transmission gear. Both sides of the cage are provided with turbines, which mesh with the worm gear.

[0039] In this invention, the turbine is connected to one end of the first connecting rod, the other end of the first connecting rod is hinged to one end of the second connecting rod, and a gripper is installed on the end of the second connecting rod away from the first connecting rod.

[0040] The bottom of the cage is connected to one end of the third link, and the other end of the third link is hinged to the middle of the second link.

[0041] In this invention, the flexible strip is disposed at the bottom of the third link.

[0042] In this invention, the six-axis drive system includes a base, a circumferential rotation axis, a first drive motor, a lower swing arm, a second drive motor, an upper swing arm, a third drive motor, and a wrist arm.

[0043] The multi-station adaptive material handling robot for injection molding machines that implements the present invention has the following beneficial effects:

[0044] The material handling system integrates flexible strips and magnetorheological fluid. By controlling the rheological properties of the magnetorheological fluid through an external magnetic field, the flexible strips can bend freely to fit the curved surface when no power is applied, and instantly lock their shape after power is applied, providing rigid support.

[0045] Each suction cup is equipped with an active compliance unit, which can independently adjust the suction cup tilt angle and axial position to achieve dynamic compensation for local features of curved surfaces, significantly improving the fitting accuracy.

[0046] Geometric data of the curved injection-molded part is acquired through laser or image scanning. Combined with an objective function optimization algorithm, the optimal initial posture of the suction cups is calculated to maximize the number of core suction cups. Real-time monitoring of suction cup pressure sensor data dynamically adjusts the clamping force of non-core suction cups, ensuring uniform pressure distribution during gripping and preventing workpiece damage. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the multi-station adaptive material handling robot for injection molding machines according to the present invention;

[0048] Figure 2 This is another structural schematic diagram of the multi-station adaptive material handling robot for injection molding machines according to the present invention;

[0049] Figure 3 This is a schematic diagram of the material injection and dispensing system of the present invention;

[0050] Figure 4 This is a schematic diagram of the material injection and dispensing system of the present invention from another angle;

[0051] Figure 5 This is a partial structural diagram of the material injection and dispensing system of the present invention;

[0052] Figure 6 This is a schematic diagram of the state of the material injection and dispensing system of the present invention;

[0053] Figure 7 This is a schematic diagram of another state of the material injection and dispensing system of the present invention;

[0054] Figure 8 This is a flowchart of the material injection and dispensing system of the present invention;

[0055] Figure 9 This is a logic block diagram of the material injection and dispensing system of the present invention.

[0056] The reference numerals in the attached figures are as follows: 10-six-axis drive system, 101-base, 102-circumferential rotation axis, 103-first drive motor, 104-lower swing arm, 105-second drive motor, 106-upper swing arm, 107-third drive motor, 108-wrist arm, 20-material handling system, 201-mounting shaft, 201A-cage, 202-opening and closing motor, 203-first transmission gear, 204-second transmission gear, 205-worm gear, 206-turbine, 207-first connecting rod, 208-second connecting rod, 209-third connecting rod, 210-gripper, 211-flexible strip, 211A-flexible outer wall, 211B-magnetorheological fluid, 212-suction cup. Detailed Implementation

[0057] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0058] Example 1

[0059] Reference Figures 1 to 2 As shown, this embodiment provides a multi-station adaptive material handling robot for injection molding machines, including a six-axis drive system 10 and a material handling system 20.

[0060] The six-axis drive system 10 includes a base 101, and a circumferential rotation axis 102 is mounted on top of the base 101. A first drive motor 103 is disposed between the base 101 and the circumferential rotation axis 102. The first drive motor 103 is used to drive the circumferential rotation axis 102 to rotate in the J1 direction.

[0061] A lower swing arm 104 is provided at the upper end of the circumferential rotation shaft 102, and a second drive motor 105 is installed between the lower swing arm 104 and the circumferential rotation shaft 102. The second drive motor 105 is used to control the lower swing arm 104 to rotate in the J2 direction.

[0062] An upper swing arm 106 is mounted on one end of the lower swing arm 104, and a third drive motor 107 is disposed between the upper swing arm 106 and the lower swing arm 104. The third drive motor 107 is used to control the upper swing arm 106 to rotate in the J3 direction. The upper swing arm 106 is a telescopic shaft used to control its extension and retraction in the J4 direction. A wrist arm 108 is disposed on one end of the upper swing arm 106, and the wrist arm 108 rotates in the J5 direction. A material handling system 20 is disposed on the end of the wrist arm 108 away from the upper swing arm 106. This material handling system 20 can rotate in the J6 direction and is used to pick up the product molded by the injection molding machine and transport it to a designated position.

[0063] In this embodiment, the six-axis drive system 10 and the material handling system 20 work together to achieve automatic gripping, posture adjustment and precise placement of injection molded products at multiple stations.

[0064] Its working principle is as follows: The first drive motor 103 drives the circumferential rotation shaft 102 to rotate around the vertical axis (J1 direction) of the base 101, realizing the overall rotation of the robot in the horizontal plane, adjusting the horizontal orientation of the robot, and covering the material handling needs of multiple workstations.

[0065] The second drive motor 105 drives the lower swing arm 104 to swing up and down around the hinge point (J2 direction) of the circumferential rotation axis 102, controlling the initial pitch angle of the robot in the vertical plane and expanding the working range.

[0066] The third drive motor 107 drives the upper swing arm 106 to swing around the hinge point (J3 direction) of the lower swing arm 104, further adjusting the extension angle of the robot arm, and working with the lower swing arm 104 to achieve precise spatial positioning.

[0067] The upper swing arm 106 serves as a telescopic axis, capable of extending and retracting along its axis (J4 direction). It integrates a telescopic mechanism, typically a linear motor or hydraulic cylinder. This directly alters the effective length of the robotic arm, adapting to different material handling and placement needs at varying distances.

[0068] The wrist arm 108 rotates about the end axis (J5 direction) of the upper swing arm 106. The drive battery for controlling the wrist arm 108 is not shown in the diagram. The posture of the material handling system 20 is adjusted to ensure that the gripping angle matches the product to be molded by the injection molding machine.

[0069] The material handling system 20 rotates around the end axis (J6 direction) of the wrist arm 108 to complete the final posture adjustment of the product and ensure that it can be accurately placed in the designated position.

[0070] The six-axis drive system 10 utilizes the coordinated control of the six degrees of freedom. Axes J1-J3 collaboratively determine the robot's large-scale spatial position, axis J4 fine-tunes the distance, and axes J5-J6 adjust the posture of the end effector 20 to ensure gripping accuracy. After the end effector 20 grips the injection-molded product, the axes move in opposite directions to transport the product to the designated location.

[0071] Reference Figures 3 to 4 As shown, the material handling system 20 includes a mounting shaft 201, which is engaged with the wrist arm 108. An opening / closing motor 202 is mounted on the side of the mounting shaft 201. A first transmission gear 203 is mounted on the output shaft of the opening / closing motor 202, and the first transmission gear 203 meshes with a second transmission gear 204. The second transmission gear 204 is rotatably connected to the bottom of the mounting shaft 201, and the mounting shaft 201 passes through the second transmission gear 204 and is fixedly connected to the retainer 201A.

[0072] A worm gear 205 is fixedly connected to the middle of the cage 201A. The worm gear 205 is fixedly connected to the second transmission gear 204 and rotates with the second transmission gear 204. A turbine 206 is provided on both sides of the cage 201A, and the turbine 206 meshes with the worm gear 205.

[0073] The turbine 206 is connected to one end of the first connecting rod 207, and the other end of the first connecting rod 207 is hinged to one end of the second connecting rod 208. A gripper 210 is installed on the end of the second connecting rod 208 away from the first connecting rod 207. The bottom of the retainer 201A is connected to one end of the third connecting rod 209, and the other end of the third connecting rod 209 is hinged to the middle of the second connecting rod 208.

[0074] In this embodiment, the material handling system 20 controls the opening and closing action of the gripper 210 by the opening and closing motor 202, thereby completing the gripping and releasing of materials.

[0075] After the opening and closing motor 202 starts, its output shaft drives the first transmission gear 203 to rotate. The first transmission gear 203 meshes with the second transmission gear 204, transmitting power to the second transmission gear 204. Since the second transmission gear 204 is rotatably connected to the bottom of the mounting shaft 201, the mounting shaft 201 itself does not rotate, but the rotation of the second transmission gear 204 will drive the worm gear 205 fixedly connected to it to rotate synchronously.

[0076] The rotation of the worm gear 205 drives the turbines 206 on both sides to rotate. The meshing of the turbines 206 and the worm gear 205 converts the rotational motion into the lateral oscillation of the turbines 206, thereby driving the first connecting rod 207 connected to the turbines 206 to move. The other end of the first connecting rod 207 is hinged to the second connecting rod 208, and the oscillation of the turbines 206 is transmitted to the second connecting rod 208 through the first connecting rod 207. At the same time, the third connecting rod 209 at the bottom of the cage 201A is hinged to the middle of the second connecting rod 208, playing a role in stabilization and auxiliary transmission.

[0077] The movement of the second link 208 ultimately drives the gripper 210 to complete the opening and closing action. When the turbine 206 drives the first link 207 to retract inward, the second link 208 pushes the gripper 210 to close, realizing the gripping of materials; conversely, when the turbine 206 drives the first link 207 to extend outward, the gripper 210 opens and releases the materials.

[0078] However, the grippers 210 of the material handling system 20 still have shortcomings when gripping curved injection molded products. The main issues are: the rigid grippers 210 have difficulty conforming to the curved surface; if the surface shape of the injection molded product is complex or its dimensions vary significantly, the grippers may not be able to fully conform, leading to unstable gripping or excessive localized force, causing scratches or deformation of the product. Uneven pressure distribution during gripping of curved surfaces may also cause indentations or damage to the product surface.

[0079] Injection-molded products have low rigidity, especially thin-walled parts. If the clamping force applied by the grippers is too large, it may cause product deformation or dimensional deviations. If the clamping force is insufficient, curved products may slip during handling, affecting the stability of automated production.

[0080] Example 2

[0081] Based on the above embodiments, and in light of the aforementioned problems, this embodiment provides further disclosure.

[0082] Refer again Figures 3 to 9 As shown, at least two third connecting rods 209 have flexible strips 211 at their bottom, and multiple suction cups 212 are provided at the bottom of the flexible strips 211. The flexible strips 211 consist of a flexible outer wall 211A and a magnetorheological fluid 211B. Magnetorheological fluid (MR fluid) exhibits rheological properties that can change rapidly and reversibly under the influence of an external magnetic field. After applying a magnetic field, the magnetorheological fluid can change from a liquid state to a semi-solid state, and quickly return to a liquid state after the magnetic field is removed. The performance change is completely reversible, with no residual magnetization or permanent structural alteration.

[0083] When handling curved injection molded products, the material handling system 20 removes the grippers 210 and uses only the multiple suction cups 212 at the bottom of the flexible strip 211 to grip the curved injection molded products. Before gripping, the magnetorheological fluid 211B inside the flexible strip 211 is not subjected to a magnetic field and can be bent at will.

[0084] Specifically, refer again Figures 6 to 7 As shown, each suction cup 212 integrates a ball joint structure, a magnetorheological damper, a linear servo motor, and a pressure sensor at the bottom of the flexible strip 211, forming an active compliant unit. The ball joint structure allows the suction cup 212 to tilt freely within a ±25° range, adapting to sudden changes in local curvature. The magnetorheological damper adjusts the damping characteristics in real time by regulating the magnetic field strength, achieving soft contact buffering and preventing damage to the workpiece from hard collisions. The linear servo motor, in conjunction with the pressure sensor, dynamically adjusts the axial position of the suction cup based on pressure sensor feedback, ensuring that each suction cup can conform to the curved surface of the workpiece.

[0085] In this embodiment, refer to Figures 8 to 9 As shown, the specific steps of a multi-station adaptive material handling robot for injection molding machines include:

[0086] Step 1: Target detection, identifying the gripping surface of the curved injection molded product;

[0087] Step 2: Calculate the initial gripping posture based on the detected curved injection molded product;

[0088] The initial gripping posture is calculated and then attached to the gripping surface of the curved injection molded product. However, since the gripping surface of the curved injection molded product is often a complex surface, not all suction cups can be completely attached to the gripping surface of the curved injection molded product.

[0089] Step 3: Define the suction cup that is completely in contact with the gripping surface of the curved injection molded product as the core suction cup, and define the suction cup that is not completely in contact with the gripping surface of the curved injection molded product as the non-core suction cup.

[0090] Step 4: Adjust the posture of non-core suction cups to optimize gripping stability;

[0091] Step 5: Evaluate the gripping strength of all suction cups to determine the probability of successful gripping.

[0092] In this embodiment, the gripping surface of the curved injection molded product is scanned to obtain the gripping curved surface S. The scanning method can be either laser scanning or image scanning.

[0093] First, define an objective function that includes the distance error of all suction cups to the gripping surface S and the angular error between the suction cup axis direction and the normal vector of the gripping surface S. Adjust the overall error to a minimum to ensure that each suction cup fits the gripping surface S as accurately as possible. In other words, obtain as many core suction cups as possible.

[0094] The objective function is:

[0095]

[0096] The number of suction cups is N, and T represents the transformation matrix of all suction cups with multiple degrees of freedom.

[0097] Distance error term: p i This represents the local position coordinates of the i-th suction cup. This means that the local position coordinates of the suction cup are transformed into the global position coordinates of the suction cup through the transformation matrix T.

[0098] This represents the distance function from the point to the grasping surface S. This represents the weighting coefficient of the distance term.

[0099] Angle error term: m is the projection point q i The surface normal vector at point q i for The projection point on the grasping surface S is determined by the distance function. The nearest surface point obtained through calculation.

[0100] n i The direction of the local axis of the i-th suction cup is represented by T, and the axis direction is obtained by multiplying it by T.

[0101] The measure of angular error is the dot product of two vectors, equal to... , Let be the angle between two vectors; when The larger the angle, the greater the error.

[0102] In this embodiment, the distance error term is used to make the positions of multiple suction cups as close as possible to the gripping surface in the global coordinate system, minimizing the square of the distance error; the angle error term is used to make the axial directions of multiple suction cups as aligned as possible with the normal vector of the curved surface.

[0103] Furthermore, Ceres / PyTorch is used to construct residuals, and the transformation matrix T is automatically differentiated and optimized. This process is repeated 3-4 times until convergence, and the optimal transformation matrix T is output, thus ensuring that all suction cups simultaneously satisfy the following conditions: they are as close as possible to the workpiece surface, and the axial direction is aligned with the surface normal as much as possible.

[0104] Based on the optimal transformation matrix T, output the initial grasping posture.

[0105] In this embodiment, the distinction between the core suction cup and the non-core suction cup is determined by the pressure value measured by a pressure sensor. Each suction cup integrates a pressure sensor to monitor the contact pressure between the suction cup and the gripping surface in real time. The pressure value is higher when the suction cups are fully adhered; the pressure value is lower when they are not fully adhered, possibly due to gaps or partial contact.

[0106] Set a pressure threshold Q th For each suction cup, measure its pressure sensor reading Q. i If Q i Greater than Q th If Q is the core suction cup, then it is determined to be the core suction cup; if Q i Less than Q th If so, it is determined to be a non-core suction cup.

[0107] In this embodiment, the attitude adjustment of the non-core suction cup is performed. The displacement is adjusted axially by a linear servo motor. Until Q i Approaching Q th Set the pressure threshold.

[0108] The calculation formula is as follows:

[0109]

[0110] Indicates pressure deviation, where K p K i K d This indicates parameter adjustment. Specifically, during the adjustment of the linear servo motor, the tilt angle of the suction cup is adjusted via a ball joint structure.

[0111] Furthermore, the magnetorheological damper adjusts the damping coefficient according to the applied pressure Qi to prevent workpiece clamping.

[0112] In this embodiment, the magnetorheological fluid 211B is in a liquid state when not energized, and the flexible strip 211 can bend freely to conform to the curved surface; after energization, the magnetorheological fluid 211B instantly becomes semi-solid, locking the shape of the suction cup array and providing rigid support. Then, distributed active compliant unit independent closed-loop control is performed to solve the limitations of traditional robotic arms in overall posture adjustment.

[0113] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-station adaptive material handling robot for injection molding machines, comprising a six-axis drive system (10) and a material handling system (20), wherein the six-axis drive system (10) drives the material handling system (20) via multiple degrees of freedom. Its features are, The material handling system (20) includes a flexible strip (211), with multiple suction cups (212) at the bottom of the flexible strip (211). The flexible strip (211) is composed of a flexible outer wall (211A) and a magnetorheological fluid (211B). Among them, multiple suction cups (212) integrate ball joint structures, magnetorheological dampers, linear servo motors, and pressure sensors at the bottom of the flexible strip (211), forming an active compliance unit. The specific steps are as follows: Step 1: Target detection, identifying the gripping surface of the curved injection molded product; Step 2: Calculate the initial gripping posture based on the detected curved injection molded product; Step 3: Define the suction cup that is completely in contact with the gripping surface of the curved injection molded product as the core suction cup, and define the suction cup that is not completely in contact with the gripping surface of the curved injection molded product as the non-core suction cup. Step 4: Adjust the posture of non-core suction cups to optimize gripping stability; Step 5: Evaluate the gripping strength of all suction cups to determine the probability of a successful grip. Here, a pressure threshold Q is set. th For each suction cup, measure its pressure sensor reading Q. i If Q i Greater than Q th If Q is the core suction cup, then it is determined to be the core suction cup; if Q i Less than Q th If so, it is determined to be a non-core suction cup. The suction cup is equipped with an active compliant unit that independently adjusts the suction cup tilt angle and axial position. The suction cup integrates a ball joint structure, magnetorheological damper, linear servo motor and pressure sensor at the bottom of the flexible strip to form an active compliant unit. The ball joint structure allows the suction cup to tilt freely within a range of ±25° and dynamically adjusts the axial position of the suction cup based on the feedback from the pressure sensor.

2. The multi-station adaptive material handling robot for injection molding machines according to claim 1, characterized in that, Scan the gripping surface of the curved injection molded part to obtain the gripping surface S. Establish an objective function that includes the distance errors from all suction cups to the gripping surface S, as well as the angular errors between the suction cup axis direction and the normal vector of the gripping surface S. The objective function is: ； Where the number of suction cups is N, and T represents the transformation matrix of multiple degrees of freedom of all suction cups; Distance error term: p i This represents the local position coordinates of the i-th suction cup. This means that the local position coordinates of the suction cup are transformed into the global position coordinates of the suction cup using the transformation matrix T; This represents the distance function from the point to the grasping surface S. Represents the weighting coefficient of the distance term; Angle error term: m is the projection point q i The surface normal vector at point q i for The projection point on the grasping surface S is determined by the distance function. The nearest surface point obtained through calculation; n i This represents the local axial direction of the i-th suction cup, which is multiplied by T to obtain the axial direction. The measure of angular error is the dot product of two vectors, equal to... , Let be the angle between the two vectors.

3. The multi-station adaptive material handling robot for injection molding machines according to claim 2, characterized in that, The residual is constructed using Ceres / PyTorch, and the transformation matrix T is automatically differentiated and optimized. The optimal transformation matrix T is then output, and the initial grasping posture is output based on the optimal transformation matrix T.

4. The multi-station adaptive material handling robot for injection molding machines according to claim 1, characterized in that, The linear servo motor adjusts the displacement along the axial direction. Until Q i Approaching Q th Set pressure threshold, The calculation formula is as follows: ； Indicates pressure deviation, where K p K i K d This indicates the adjustment parameters, where the tilt angle of the suction cup is adjusted through a ball joint structure during the adjustment of the linear servo motor.

5. The multi-station adaptive material handling robot for injection molding machines according to claim 1, characterized in that, The material handling system (20) includes a mounting shaft (201) that engages with the wrist arm (108). An opening and closing motor (202) is provided on the side of the mounting shaft (201), and a first transmission gear (203) is provided on the output shaft of the opening and closing motor (202). The first transmission gear (203) meshes with a second transmission gear (204). The second transmission gear (204) is rotatably connected to the bottom of the mounting shaft (201), and the mounting shaft (201) passes through the second transmission gear (204) and is fixedly connected to the cage (201A).

6. The multi-station adaptive material handling robot for injection molding machines according to claim 5, characterized in that, A worm gear (205) is fixedly connected to the middle of the cage (201A). The worm gear (205) is fixedly connected to the second transmission gear (204) and rotates with the second transmission gear (204). A turbine (206) is provided on both sides of the cage (201A), and the turbine (206) meshes with the worm gear (205).

7. The multi-station adaptive material handling robot for injection molding machines according to claim 6, characterized in that, The turbine (206) is connected to one end of the first connecting rod (207), and the other end of the first connecting rod (207) is hinged to one end of the second connecting rod (208). A gripper (210) is installed on the end of the second connecting rod (208) away from the first connecting rod (207). The bottom of the retainer (201A) is connected to one end of the third link (209), and the other end of the third link (209) is hinged to the middle of the second link (208).

8. The multi-station adaptive material handling robot for injection molding machines according to claim 7, characterized in that, The flexible strip (211) is located at the bottom of the third link (209).

9. The multi-station adaptive material handling robot for injection molding machines according to claim 1, characterized in that, The six-axis drive system (10) includes a base (101), a circumferential rotation axis (102), a first drive motor (103), a lower swing arm (104), a second drive motor (105), an upper swing arm (106), a third drive motor (107), and a wrist arm (108).

Images