Stamping die crane sling control method

Abstract

The invention discloses a control method for a stamping die crane lifting appliance. The control method comprises the following steps of S1, die identity recognition and model matching; s2, dynamic lifting point calculation; s3, load balance control; s4, generating an anti-swing track; s5, carrying out digital twinborn verification; and S6, performing closed-loop feedback optimization. According to the method, model registration is carried out by scanning QR codes or laser marking numbers and scanning mold surface features in real time by using a depth camera, so that accurate recognition and spatial positioning of the mold are realized; an ICP algorithm is adopted to align scanning feature data with a standard model, and a regional lifting point is selected based on a finite element analysis result, so that the risk of mold damage caused by overlarge stress is avoided; the lifting point load is monitored in real time through the weighing sensor, and the lifting speed is dynamically adjusted by utilizing a speed correction formula, so that the unbalance loading problem is solved; an acceleration trajectory is generated by adopting a seven-section S curve, so that the swing amplitude of goods is reduced; and the safety, the accuracy and the efficiency of mold hoisting are obviously improved.

Description

Technical Field

[0001] This invention relates to the field of crane lifting control technology, specifically a method for controlling the lifting device of a stamping die crane. Background Technology

[0002] In the industrial manufacturing sector, stamping dies are widely used in various industries such as automobiles and home appliances because they can efficiently and accurately form metal parts. However, due to their large size and weight, the handling and installation process is complex and has a high risk factor, especially when crane lifting is involved, which places extremely high demands on operational precision and safety.

[0003] Stamping dies often have complex shapes (e.g., car door dies frequently have curved surfaces and inverted structures). Traditional lifting equipment relies on manual experience to select lifting points and manually adjusts the crane for positioning, which takes more than 15 minutes per operation. Uneven weight distribution in the dies (eccentric loading can reach ±30%) makes existing lifting equipment prone to tilting, leading to die collisions and a scrap rate as high as 0.5%, potentially causing safety accidents. Excessive manual intervention makes it difficult to align with the cycle time of automated stamping production lines (requiring die switching within 90 seconds). Industry demands: The trend towards lightweighting in automobiles increases the frequency of die replacements (new energy vehicle dies are switched 8-12 times daily); Die precision requirements are rising (positioning error must be ≤±1mm), which is difficult to meet with manual operation.

[0004] In recent years, with the advancements in computer vision, robotics, and intelligent control systems, more and more technologies have been introduced into the mold hoisting process, such as using QR codes or laser-engraved numbers for identification and depth camera scanning to acquire 3D model data. However, the application of these technologies still faces some challenges, such as how to accurately match the CAD model of the mold with its actual physical characteristics, how to ensure that the selection of dynamic lifting points can both guarantee the structural safety of the mold and maintain good balance, and how to effectively avoid the swaying of goods along the handling path. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the existing defects and provide a control method for the lifting device of a stamping die crane. This method improves the accuracy of die identification and registration, optimizes the selection of lifting points, enhances the load balance control capability, improves the level of handling path planning, and realizes digital twin verification and closed-loop feedback optimization. It significantly improves the safety, accuracy and efficiency of die lifting and can effectively solve the problems in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for controlling the lifting device of a stamping die crane, comprising the following steps:

[0007] S1. Mold Identification and Model Matching: Scan the QR code or laser-engraved number on the mold, retrieve the mold's CAD model and historical hoisting parameters from the database. If the mold has no preset data, execute a real-time depth camera scan of the mold's surface features to generate point cloud data. Combine this with Open3D library algorithms to extract key feature points, register them with the mold's CAD model, and calculate the mold's precise coordinates and orientation in space. Model registration error function:

[0008] ;

[0009] Among them, the weighting coefficient , Let R be the target optimization value, and R be a 3x3 rotation matrix. Let be a point in the source point set, and t be the translation vector. For the points in the target point set, The initial distance between the i-th point pairs;

[0010] S2. Dynamic Lifting Point Calculation: The scanned feature data is aligned with the standard model using the ICP algorithm. Based on the finite element analysis results, areas with stress concentration < 20 MPa are selected as lifting points. The dynamic lifting point optimization algorithm is as follows:

[0011] Multi-objective evaluation function: ;

[0012] Stress gradient constraint: ;

[0013] in, The maximum stress in the candidate lifting point area, For the effective contact area between the lifting point and the mold, Minimum effective contact area threshold, For centroid shift, For the centroid coordinates of the mold, The coordinates of the center of gravity formed by the combination of lifting points;

[0014] S3. Load Balance Control: Real-time monitoring of load at each lifting point during the hoisting phase. - When off-center load is detected, speed adjustment is performed, and the speed correction amount is as follows:

[0015] ;

[0016] in, For average load, =0.05、 =0.01, and the speed command is sent to the corresponding electric hoist via the CAN bus;

[0017] S4. Anti-sway trajectory generation: Based on the length L of the transport path, a seven-segment S-curve is generated, and a reverse correction torque is applied in advance at the turning section.

[0018] acceleration ;

[0019] in, , The value is 0.3 m / s².

[0020] Improved acceleration curve:

[0021] ,

[0022] Time allocation: : : =3:4:3;

[0023] S5: Digital Twin Verification: The following disturbance factors are injected into the virtual environment for pre-simulation: workshop lateral wind load, track docking error and emergency braking conditions. After verification, the hoisting operation is performed.

[0024] S6: Closed-loop feedback optimization: After hoisting is completed, record the actual data, positioning error, energy consumption data and adjustment parameters, and then update the mold database.

[0025] Preferably, in step S1, the key feature points of the mold include the center of the lifting lug, the groove, and the reinforcing rib.

[0026] Preferably, in step S3, the load at each lifting point is collected in real time by four weighing sensors. , , and .

[0027] Preferably, in step S6, the mold database parameter update rules are as follows:

[0028] .

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] 1. Improve mold recognition and registration accuracy: By scanning QR codes or laser-engraved numbers and using a depth camera to scan the surface features of the mold in real time, and combining the Open3D library algorithm to extract key feature points for model registration, efficient and accurate recognition and spatial positioning of molds without preset data are achieved;

[0031] 2. Optimize dynamic lifting point selection: The ICP algorithm is used to align the scanned feature data with the standard model. Based on the finite element analysis results, areas with stress concentration of less than 20MPa are selected as lifting points, thereby avoiding the risk of mold damage caused by excessive stress and improving the stability and safety of the lifting device.

[0032] 3. Enhanced load balance control capability: By using four weighing sensors to monitor the load at each lifting point in real time and using a speed correction formula to dynamically adjust the lifting speed, the problem of uneven load is effectively solved, ensuring the stability during the lifting process and reducing the possibility of mold damage.

[0033] 4. Improve the level of handling path planning and anti-sway control: The seven-segment S-curve is used to generate acceleration trajectory, and the reverse correction torque is applied in advance in the turning section, which makes the crane more stable during handling, reduces the swing amplitude of the goods, and improves operating efficiency and safety.

[0034] In summary, this significantly improves the safety, accuracy, and efficiency of mold hoisting. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0036] The present invention can be explained in detail through the following embodiments. The purpose of disclosing the present invention is to protect all technical improvements within the scope of the present invention. In the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "front", "rear", "left", "right" indicating the orientation or positional relationship, they are only corresponding to the drawings of this application for the convenience of describing the present invention, and are not intended to indicate or imply that the device or element referred to must have a specific orientation.

[0037] Please see Figure 1 This invention provides a technical solution: a method for controlling the lifting device of a stamping die crane, comprising the following steps:

[0038] S1. Mold Identification and Model Matching: Scan the QR code or laser-engraved number on the mold, retrieve the mold's CAD model and historical hoisting parameters from the database. If the mold has no preset data, execute a real-time depth camera scan of the mold's surface features to generate point cloud data. Combine this with Open3D library algorithms to extract key feature points, register them with the mold's CAD model, and calculate the mold's precise coordinates and orientation in space. Model registration error function:

[0039] ;

[0040] Among them, the weighting coefficient , Let R be the target optimization value, and R be a 3x3 rotation matrix. Let be a point in the source point set, and t be the translation vector. For the points in the target point set, The initial distance between the i-th point pairs;

[0041] The error function takes into account the effects of the rotation matrix R and the translation vector t, ensuring the accuracy of model matching;

[0042] Specifically, the key features of the mold include the center of the lifting lug, the groove, and the reinforcing rib;

[0043] S2. Dynamic Lifting Point Calculation: The scanned feature data is aligned with the standard model using the ICP algorithm. Based on the finite element analysis results, the stress distribution in different regions is simulated using ANSYS or Abaqus software to determine safe lifting points. Regions with stress concentration < 20 MPa are selected as lifting points. The dynamic lifting point optimization algorithm is as follows:

[0044] Multi-objective evaluation function: ;

[0045] Stress gradient constraint: ;

[0046] in, The maximum stress in the candidate lifting point area, For the effective contact area between the lifting point and the mold, Minimum effective contact area threshold, For centroid shift, For the centroid coordinates of the mold, The coordinates of the center of gravity formed by the combination of lifting points;

[0047] Based on the feature data aligned by the ICP algorithm, and combined with the finite element analysis results, areas with stress concentration less than 20MPa are selected as lifting points. During the optimization process, a multi-objective evaluation function is used to balance factors such as stress gradient constraints, ensuring that the selected lifting points can withstand the necessary load without causing excessive local deformation of the mold. This is crucial for avoiding mold damage caused by improper selection of lifting points.

[0048] S3. Load Balance Control: Real-time monitoring of load at each lifting point during the hoisting phase. - The load at each lifting point is collected in real time by weighing sensors at four locations. , , and When off-center load is detected, speed adjustment is performed, and the speed adjustment amount is as follows:

[0049] ;

[0050] in, For average load, =0.05、 =0.01, the speed command is sent to the corresponding electric hoist via the CAN bus; through load balance control, the risk of imbalance is reduced, which helps to maintain the stability of the entire handling process, especially for large molds with uneven weight distribution;

[0051] S4. Anti-sway trajectory generation: Based on the length L of the transport path, a seven-segment S-curve is generated, and a reverse correction torque is applied in advance at the turning section.

[0052] acceleration ;

[0053] in, , The value is 0.3 m / s².

[0054] Improved acceleration curve:

[0055] ,

[0056] Time allocation: : : =3:4:3;

[0057] To further improve operational stability, a seven-segment S-curve acceleration planning method was adopted on the transport path, and a reverse correction torque was applied at the turning point to reduce swaying. This not only improved the movement accuracy, but also effectively reduced the impact caused by sudden acceleration or deceleration.

[0058] S5: Digital Twin Verification: The following disturbance factors are injected into the virtual environment for pre-simulation: workshop lateral wind load, track docking error and emergency braking conditions. After verification, the hoisting operation is performed. Only after full verification will the real hoisting task be performed, thereby greatly reducing the risk of on-site operation.

[0059] S6: Closed-loop feedback optimization: After hoisting is completed, record the actual data, positioning error, energy consumption data, and adjustment parameters, and then update the mold database. The mold database parameter update rules are as follows:

[0060] .

[0061] Example of implementation: Upgrading the mold hoisting system in the stamping workshop of a large automobile manufacturing plant

[0062] 1. Project Background: A leading domestic automobile manufacturer (hereinafter referred to as "Plant A") has an annual production capacity of 800,000 vehicles. Its stamping workshop is equipped with 10 large stamping production lines, each using multiple sets of large dies (size: 4m × 2.5m × 1.2m, weight: 8–12 tons). The original hoisting system adopted a manual visual positioning + fixed lifting points + manual balance adjustment method, which had the following problems: low hoisting efficiency (average single hoisting time of 18 minutes); high die damage rate (approximately 5 times per year due to improper hoisting causing die deformation or cracking of lifting lugs); and high safety risks (once a slight off-center load caused the die to shake and collide with the equipment).

[0063] To improve automation and safety, Plant A decided to introduce a lifting device system based on digital twins and intelligent control, and adopt the control method described in this article;

[0064] 2. System Deployment and Parameter Configuration:

[0065]

[0066] 3. Implementation process and application of key technologies:

[0067] S1: Mold Identification and Model Matching

[0068] Scenario: New mold launched, no preset data;

[0069] Process: When the lifting device moves to a position 1.5m above the mold, the depth camera automatically triggers a scan; after collecting point cloud data, coarse registration is performed using Open3D's Fast Global Registration (FGR) algorithm, and then fine registration is performed using Point-to-Plane ICP; Feature point matching: The system automatically identifies the centers of 4 lifting lugs, the edges of 2 main grooves, and the inflection points of 3 reinforcing ribs;

[0070] Results: The registration error was controlled within 0.08 mm RMS, and the attitude angle error was < 0.3°, meeting the hoisting and positioning requirements;

[0071] S2: Dynamic Lifting Point Calculation (Finite Element Integration)

[0072] System integration: The hoisting control software interfaces with ANSYS Mechanical APDL and has a pre-stored database of FEA results for 500+ molds;

[0073] Calculation process: After model registration, the system retrieves the stress cloud map of the corresponding mold; a search is performed within the lifting lug area (φ120mm circular area). A sub-region with a pressure of 20 MPa; the output is calculated using a multi-objective optimization function; the system automatically recommends four optimal lifting point coordinates (error ±2 mm) and sends them to the electric hoist positioning system;

[0074] Case Study: A side panel mold (weighing 11.2t) traditionally had its lifting points located at the edge of the lifting lugs, resulting in a stress of 25MPa. The new system recommends offsetting the lifting points inward by 30mm, reducing the stress to 16.8MPa and significantly improving safety.

[0075] S3: Load balancing control (real-time closed loop)

[0076] Sensor configuration: Each electric hoist integrates an S-type tension sensor with a sampling frequency of 100Hz;

[0077] Control logic: Initial lifting speed: 0.2 m / s, when 5% Time-triggered correction, speed correction amount: Speed ​​commands are sent via CANopen, and the electric hoist's response time is <150ms.

[0078] Actual measurement data: During a certain lifting operation, F1=2850kg, F2=2920kg, F3=3010kg, F4=3080kg, the system reduced the deviation from 9.8% to 1.2% within 2.3 seconds, achieving a smooth lift;

[0079] S4: Anti-sway trajectory generation (path optimization)

[0080] Transportation route: From the mold storage area to the stamping machine, the total length L=28m, including one 90° turn;

[0081] Trajectory planning: Straight line segment: Using an improved S-curve, T= = ≈9.66s, acceleration time =2.898s, =3.864s, =2.898s; Turning section (5m in advance): Apply reverse torque, and control the lateral correction force within ±150N;

[0082] Results: The end swing amplitude was reduced from ±120mm to ±15mm compared to the traditional trapezoidal curve, and the positioning accuracy reached ±3mm;

[0083] S5: Digital Twin Verification (Virtual Pre-Show)

[0084] Simulation scenario: Crosswind: Level 3 wind (5.5 m / s), applied lateral wind load ≈ 8 N / m; Track error: ±1.5 mm step disturbance; Emergency braking: deceleration 1.0 m / s²;

[0085] Verification results: Mold sway < 50mm under all working conditions; maximum stress did not exceed the material yield strength; the system determined "verification passed", allowing actual hoisting to proceed;

[0086] S6: Closed-loop feedback optimization (data-driven iteration)

[0087] Data records: Average time per hoisting operation: 10.3 minutes; Average positioning error: ±2.1 mm; Energy consumption: Average 4.8 kWh / operation;

[0088] Database update: Added "optimal lifting point offset" for this mold to +30mm (based on stress optimization); Updated PID parameters: Kp adjusted from 0.05 to 0.055 (for heavy mold response lag);

[0089] 4. Comparison of Implementation Results:

[0090]

[0091] This case study demonstrates the high precision, high safety, and high efficiency of the stamping die crane lifting control method in a real industrial environment.

[0092] The parts of this invention not described in detail are prior art. It will be apparent to those skilled in the art that this invention is not limited to the details of the above exemplary embodiments, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects, and are intended to encompass all changes that fall within the meaning and scope of equivalents in the content of this invention.

Claims

1. A method for controlling the lifting device of a stamping die crane, characterized in that: Includes the following steps: S1. Mold Identification and Model Matching: Scan the QR code or laser-engraved number on the mold, retrieve the mold's CAD model and historical hoisting parameters from the database. If the mold has no preset data, execute a real-time depth camera scan of the mold's surface features to generate point cloud data. Combine this with Open3D library algorithms to extract key feature points, register them with the mold's CAD model, and calculate the mold's precise coordinates and orientation in space. Model registration error function: ； Among them, the weighting coefficient , Let R be the target optimization value, and R be a 3x3 rotation matrix. Let be a point in the source point set, and t be the translation vector. For the points in the target point set, The initial distance between the i-th point pairs; S2. Dynamic Lifting Point Calculation: The scanned feature data is aligned with the standard model using the ICP algorithm. Based on the finite element analysis results, areas with stress concentration < 20 MPa are selected as lifting points. The dynamic lifting point optimization algorithm is as follows: Multi-objective evaluation function: ; Stress gradient constraint: ; in, The maximum stress in the candidate lifting point area, For the effective contact area between the lifting point and the mold, Minimum effective contact area threshold, For centroid shift, For the centroid coordinates of the mold, The coordinates of the center of gravity formed by the combination of lifting points; S3. Load Balance Control: Real-time monitoring of load at each lifting point during the hoisting phase. - When off-center load is detected, speed adjustment is performed, and the speed correction amount is as follows: ； in, For average load, =0.05、 =0.01, and the speed command is sent to the corresponding electric hoist via the CAN bus; S4. Anti-sway trajectory generation: Based on the length L of the transport path, a seven-segment S-curve is generated, and a reverse correction torque is applied in advance at the turning section. acceleration ; in, , The value is 0.3 m / s². Improved acceleration curve: ， Time allocation: : : =3:4:3; S5: Digital Twin Verification: The following disturbance factors are injected into the virtual environment for pre-simulation: workshop lateral wind load, track docking error and emergency braking conditions. After verification, the hoisting operation is performed. S6: Closed-loop feedback optimization: After hoisting is completed, record the actual data, positioning error, energy consumption data and adjustment parameters, and then update the mold database.

2. The method for controlling the lifting device of a stamping die crane according to claim 1, characterized in that: In step S1, the key feature points of the mold include the center of the lifting lug, the groove, and the reinforcing rib.

3. The method for controlling the lifting device of a stamping die crane according to claim 1, characterized in that: In step S3, the load at each lifting point is collected in real time by weighing sensors at four locations. , , and .

4. The method for controlling the lifting device of a stamping die crane according to claim 1, characterized in that: In step S6, the mold database parameter update rules are as follows: 。

Images