Complex sheet metal part bending device based on machine learning and bending parameter optimization process

Abstract

The invention belongs to the technical field of complex sheet metal part bending, and discloses a machine learning-based complex sheet metal part bending device and a bending parameter optimization process, the machine learning-based complex sheet metal part bending device comprises a bending frame body, a reinforcing table and a bending table, the reinforcing table is arranged in the bending frame body, and the bending table is fixedly mounted in the bending frame body; two reinforcing rods are fixedly mounted in each of the bending frame body and the reinforcing table, through cooperation of a fitting plate, a two-way threaded rod, a reinforcing strip, a contact block, a leveling plate, a protective cover, a twisting strip and an arc-shaped plate, the complex sheet metal part is placed on the surface of the fitting plate for subsequent bending, the twisting strip is rotated clockwise, and therefore the complex sheet metal part is bent. According to the sheet metal part bending device, the two-way threaded rod connected with the two-way threaded rod is driven to rotate, workers are prevented from being accidentally injured in the sheet metal part bending process, the safety of the bending device is improved, the contact block moves at one end of the reinforcing strip, and the contact block is supported through the two-way threaded rod and the reinforcing strip, so that the moving stability of the contact block is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of complex sheet metal bending, and specifically provides a complex sheet metal bending device and a bending parameter optimization process based on machine learning. Background Art

[0002] The complex sheet metal bending device is based on the application of machine learning, which mainly uses intelligent algorithms to optimize the bending process, improve bending accuracy and efficiency, and reduce production costs. The premise of machine learning is high-quality data. Therefore, the bending device needs to be equipped with various sensors and monitoring equipment to collect data in the bending process in real time. The bending device based on machine learning optimizes the bending process by learning from historical data. Supervised learning is the most commonly used technology in machine learning, especially when it is necessary to predict the bending results. The bending device is a key equipment for sheet metal processing. The bending process is monitored in real time through sensors and machine learning algorithms, and the motion trajectory and pressure of the bending machine are dynamically adjusted to ensure the bending accuracy of each sheet metal part.

[0003] However, the existing complex sheet metal bending devices and bending parameter optimization processes based on machine learning are not perfect and still have certain defects:

[0004] In the bending device, especially when bending and loading complex sheet metal parts, there is no abutment structure on both sides of the complex sheet metal parts to keep them balanced, resulting in unstable positioning of the sheet metal parts during the bending process, which may cause the position of the workpiece to shift and affect the bending accuracy. Due to unstable positioning, manual readjustment and correction may be required, which wastes time. The lack of a protective structure will expose the operator to a dangerous bending area. Especially when the bending machine is running, the workpiece may bounce and fall due to improper operation, thereby accidentally injuring the operator. Summary of the Invention

[0005] The purpose of the present invention is to provide a complex sheet metal bending device and a bending parameter optimization process based on machine learning to solve the problems raised in the above background technology.

[0006] In order to achieve the above-mentioned objectives, the present invention provides the following technical solutions: a complex sheet metal bending device and a bending parameter optimization process based on machine learning, comprising a bending frame, a reinforcement platform and a bending platform, wherein the reinforcement platform is arranged inside the bending frame, the bending platform is fixedly installed inside the bending frame, and two reinforcement rods are fixedly installed inside the bending frame and the reinforcement platform, and a hydraulic cylinder is provided on the top of the reinforcement platform, and the output shaft of the hydraulic cylinder is connected to the bending knife in a transmission manner, and a bending seat is provided on the top of the bending seat, and a bonding plate is provided on one side of the bending seat, and two triangular plates are provided on the bottom of the bonding plate, and a vertical plate is fixedly installed on the bottom of the bonding plate, a bidirectional threaded rod is provided inside the vertical plate, and a reinforcement strip is provided inside the vertical plate, and contact blocks are sleeved on both ends of the bidirectional threaded rod and the reinforcement strip, and two leveling plates are provided on the top of the bonding plate, and a protective cover is provided on one side of the two leveling plates, and a pressure plate is fixedly installed on one side of the bending frame.

[0007] Preferably, a spring is fixedly installed between the inner walls of the two adjacent reinforcement rods, a torsion bar is rotatably installed on one side of the vertical plate, the two contact blocks are respectively threadedly sleeved on the two ends of the bidirectional threaded rod, and an arc plate is provided on the other side of the two leveling plates.

[0008] Preferably, one end of the torsion bar is fixedly connected to one end of the bidirectional threaded rod, a contact rod is fixedly installed on one side of the two contact blocks, one end of the two contact rods is fixedly connected to one side of the two arc-shaped plates respectively, two limit bars are provided on one side of the two leveling plates, and two embedded holes are opened on one side of the two arc-shaped plates.

[0009] Preferably, one end of the four limit strips is respectively connected to the internal threads of the four embedded holes, the positions of the two adjacent limit strips are corresponding, the tops of the two triangular plates are respectively fixedly connected to the bottoms of the bonding plates, and one side of the two triangular plates is respectively fixedly connected to one side of the bending seat, and a bending groove is provided at the top of the bending seat.

[0010] Preferably, positioning plates are fixedly installed on both sides of the bending seat, and multiple positioning strips are provided on the top of the two positioning plates. Multiple auxiliary holes are opened on the surface of the bending table, and one end of the multiple positioning strips is respectively connected to the internal threads of the multiple auxiliary holes.

[0011] Preferably, two arc-shaped blocks are fixedly mounted on the top of the two protective covers, an arc-shaped cover is provided on the top of the two protective covers, and an external force resistance plate is provided inside the two arc-shaped covers.

[0012] Preferably, a concave hole is opened on one side of the four arc blocks, and a limiting bolt is set on one side of the four arc blocks. The four limiting bolts are respectively connected to the four concave holes and the internal threads of the two anti-external force plates. A control panel is fixedly installed on one side of the bending frame, and a force sensor is set on one side of the bending knife.

[0013] The bending parameter optimization process of a complex sheet metal bending device based on machine learning includes the following steps:

[0014] Step 1: Determine the optimization goals: bending angle accuracy, surface quality, defect control, reducing production cycle and improving equipment utilization;

[0015] Step 2: Data acquisition: Obtain data about the bending process from the production equipment, including force, displacement, angle, temperature, material thickness and mold type. Data preprocessing: Clean the data, remove noise, and normalize the scale of different variables to prepare for subsequent analysis.

[0016] Step 3: Mathematical modeling and simulation: Use the finite element analysis method to establish a mathematical model of the bending process and simulate the influence of different bending parameters on the bending quality;

[0017] Step 4: Conduct experiments and adjust parameters: Based on the simulation results, conduct small-batch trial production to test the bending effect and adjust parameters: adjust the bending parameters according to the experimental results to ensure that they meet the target requirements;

[0018] Step 5: Use machine learning to optimize bending parameters: Use historical production data to train the machine learning model to predict bending parameter combinations. Based on real-time data, the machine learning model automatically adjusts the bending parameters.

[0019] Step 6: Quality inspection and feedback: Sensors are used to monitor various parameters of the bending process in real time. During the production process, new production data is continuously collected and used to further train the machine learning model, thereby achieving self-optimization of the system.

[0020] The beneficial effects of the present invention are as follows:

[0021] 1. The present invention cooperates with the provided bonding plate, bidirectional threaded rod, reinforcement strip, contact block, smoothing plate, protective cover, torsion strip and curved plate. The complex sheet metal part is placed on the surface of the bonding plate for subsequent bending. The torsion strip is rotated clockwise to drive the bidirectional threaded rod connected to it to rotate, and the two contact blocks threadedly sleeved thereon move relatively. Under the connection of the contact rod and the curved plate, the two smoothing plates are driven to move relatively. The two smoothing plates are respectively in contact with the two sides of the complex sheet metal part, so that the complex sheet metal part is in the same straight line, thereby improving the quality of the sheet metal bending. The two protective covers are located on both sides of the top of the bonding plate, which can shield the surrounding area, avoid accidental injury to workers during the sheet metal bending process, and improve the safety of the bending device. The contact block moves at one end of the reinforcement strip, and the support of the contact block by the bidirectional threaded rod and the reinforcement strip improves the stability of the contact block movement.

[0022] 2. The present invention cooperates with the positioning plate, positioning strip, auxiliary hole, arc cover, anti-external force plate, vertical plate and pressure-resistant plate. The arc cover and anti-external force plate cover the top of the protective cover, thereby increasing the shielding range of the surrounding environment. After the bottom surface of the positioning plate is tightly fitted with the surface of the bending table, one end of the multiple positioning strips is embedded into the inside of the auxiliary hole. With the threaded connection between the two, the installation of the bending seat and the bending table can be realized. This structure facilitates the replacement of damaged bending seats, and solves the problem of uneven surface of the bending groove affecting the bending quality. The bottom of the vertical plate abuts against the top of the pressure-resistant plate, thereby ensuring the firmness of the vertical plate. The two triangular plates are located between the bonding plate and the bending seat, and the triangle is stable, which enhances the stability of the bottom of the bonding plate.

[0023] 3. The present invention cooperates with the arc block, concave hole, limit bolt, limit strip, embedded hole, bending knife and bending seat, and rotates the limit bolt in the opposite direction to pull one end out from the inside of the concave hole and the anti-external force plate, and moves the anti-external force plate to one side to separate it from the protective cover, so that the arc cover and the anti-external force plate can be maintained and replaced separately. After one side of the arc plate is fitted with one side of the smoothing plate, one end of the limit strip is embedded into the inside of the embedded hole, and the threaded connection between the two can realize the limitation and fixation of the arc plate and the smoothing plate, and the protective cover can be installed on one side of the arc plate. The force sensor is mainly used to monitor the pressure applied to the sheet metal to ensure that the bending force is moderate. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic diagram of the three-dimensional structure of the present invention;

[0025] Figure 2 For the present invention Figure 1 Enlarged view of point A in the middle;

[0026] Figure 3 It is a front view structural schematic diagram of the present invention;

[0027] Figure 4 It is a side structural schematic diagram of the present invention;

[0028] Figure 5 It is a partial rear view structural diagram of the present invention;

[0029] Figure 6 It is a schematic diagram of the back structure of the present invention.

[0030] In the figure: 1. Bending frame; 2. Reinforcement platform; 3. Bending platform; 4. Reinforcement rod; 5. Spring; 6. Hydraulic cylinder; 7. Bending knife; 8. Bending seat; 9. Bending groove; 10. Positioning plate; 11. Positioning bar; 12. Triangular plate; 13. Laminating plate; 14. Vertical plate; 15. Pressure plate; 16. Bidirectional threaded rod; 17. Reinforcement bar; 18. Twisting bar; 19. Contact block; 20. Contact rod; 21. Arc plate; 22. Screed plate; 23. Protective cover; 24. Limiting bar; 25. Arc cover; 26. Anti-external force plate; 27. Arc block; 28. Concave hole; 29. ​​Limiting bolt; 30. Control panel. DETAILED DESCRIPTION

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0032] like Figures 1 to 6 As shown, the embodiment of the present invention provides a technical solution for a complex sheet metal bending device and a bending parameter optimization process based on machine learning:

[0033] Example 1:

[0034] like Figure 1-2As shown, a complex sheet metal bending device and a bending parameter optimization process based on machine learning include a bending frame 1, a reinforcement platform 2 and a bending platform 3. The reinforcement platform 2 is arranged inside the bending frame 1, and the bending platform 3 is fixedly installed inside the bending frame 1. Two reinforcement rods 4 are fixedly installed inside the bending frame 1 and the reinforcement platform 2. A hydraulic cylinder 6 is provided on the top of the reinforcement platform 2. The output shaft of the hydraulic cylinder 6 is connected to a bending knife 7. A bending seat 8 is provided on the top of the bending platform 3. A bonding plate 13 is provided on one side of the bending seat 8. Two triangular plates 12 are provided at the bottom of the bonding plate 13. A vertical plate 14 is fixedly installed at the bottom of the bonding plate 13. A bidirectional threaded rod 16 is provided inside the vertical plate 14. A reinforcement strip 17 is provided inside the vertical plate 14. The two ends of the threaded rod 16 and the reinforcement strip 17 are both sleeved with contact blocks 19, and two leveling plates 22 are provided on the top of the bonding plate 13. A protective cover 23 is provided on one side of the two leveling plates 22. A pressure-resistant plate 15 is fixedly installed on one side of the bending frame 1. The two leveling plates 22 are respectively abutted against the two sides of the complex sheet metal parts, so that the complex sheet metal parts are in the same straight line, thereby improving the quality of the sheet metal bending. The two protective covers 23 are located on both sides of the top of the bonding plate 13, which can shield the surroundings and avoid accidental injury to the staff during the sheet metal bending process, thereby improving the safety of the bending device. The contact block 19 moves at one end of the reinforcement strip 17, and the support of the contact block 19 by the bidirectional threaded rod 16 and the reinforcement strip 17 improves the stability of the movement of the contact block 19.

[0035] Example 2:

[0036] Based on the first embodiment, Figure 2-4 As shown, a spring 5 is fixedly installed between the inner walls of the two adjacent reinforcing rods 4, and a torsion bar 18 is rotatably installed on one side of the vertical plate 14. The two contact blocks 19 are respectively threadedly sleeved on the two end portions of the bidirectional threaded rod 16, and the other sides of the two leveling plates 22 are provided with an arc plate 21, one end of the four limit bars 24 is respectively connected to the internal threads of the four embedded holes, and the positions of the two adjacent limit bars 24 are corresponding. The tops of the two triangular plates 12 are respectively fixedly connected to the bottoms of the bonding plates 13, and one side of the two triangular plates 12 is respectively fixedly connected to one side of the bending seat 8. A bending groove 9 is provided at the top of the bending seat 8. This structure facilitates the replacement of the damaged bending seat 8, solves the problem that the uneven surface of the bending groove 9 affects the bending quality, and the bottom of the vertical plate 14 abuts against the top of the pressure plate 15 to ensure the firmness of the vertical plate 14. The two triangular plates 12 are located between the bonding plate 13 and the bending seat 8, and the triangle has stability, which enhances the stability of the bottom of the bonding plate 13.

[0037] Example 3:

[0038] Based on the first and second embodiments, Figure 1 、 Figure 5 and Figure 6 As shown, positioning plates 10 are fixedly installed on both sides of the bending seat 8, and multiple positioning strips 11 are provided on the top of the two positioning plates 10. Multiple auxiliary holes are opened on the surface of the bending table 3, and one end of the multiple positioning strips 11 is respectively connected to the internal threads of the multiple auxiliary holes. Two arc blocks 27 are fixedly installed on the top of the two protective covers 23, and arc covers 25 are provided on the top of the two protective covers 23. The internal parts of the two arc covers 25 are provided with anti-external force plates 26. Through the arc blocks 27, concave holes 28, limit bolts 29, and limit With the cooperation of the strip 24, the embedding hole, the bending knife 7 and the bending seat 8, the limiting bolt 29 is rotated in the opposite direction so that one end of it is pulled out from the inside of the recessed hole 28 and the anti-external force plate 26, and the anti-external force plate 26 is moved to one side to separate it from the protective cover 23, so that the arc cover 25 and the anti-external force plate 26 can be maintained and replaced separately. After one side of the arc plate 21 is fitted with one side of the smoothing plate 22, one end of the limiting strip 24 is embedded into the inside of the embedding hole, and the threaded connection between the two can realize the limiting and fixation of the arc plate 21 and the smoothing plate 22.

[0039] The bending parameter optimization process of a complex sheet metal bending device based on machine learning includes the following steps:

[0040] Step 1: Determine the optimization goals: bending angle accuracy, surface quality, defect control, reducing production cycle and improving equipment utilization;

[0041] Step 2: Data acquisition: Obtain data about the bending process from the production equipment, including force, displacement, angle, temperature, material thickness and mold type. Data preprocessing: Clean the data, remove noise, and normalize the scale of different variables to prepare for subsequent analysis.

[0042] Step 3: Mathematical modeling and simulation: Use the finite element analysis method to establish a mathematical model of the bending process and simulate the influence of different bending parameters on the bending quality;

[0043] Step 4: Conduct experiments and adjust parameters: Based on the simulation results, conduct small-batch trial production to test the bending effect and adjust parameters: adjust the bending parameters according to the experimental results to ensure that they meet the target requirements;

[0044] Step 5: Use machine learning to optimize bending parameters: Use historical production data to train the machine learning model to predict bending parameter combinations. Based on real-time data, the machine learning model automatically adjusts the bending parameters.

[0045] Step 6: Quality inspection and feedback: Sensors are used to monitor various parameters of the bending process in real time. During the production process, new production data is continuously collected and used to further train the machine learning model, thereby achieving self-optimization of the system.

[0046] The working principle and usage process of the present invention: The complex sheet metal bending device is based on the application of machine learning, which mainly optimizes the bending process through intelligent algorithms to improve bending accuracy and efficiency and reduce production costs. The premise of machine learning is high-quality data. Therefore, the bending device needs to be equipped with various sensors and monitoring equipment to collect data in the bending process in real time. The bending device based on machine learning optimizes the bending process by learning from historical data. Supervised learning is the most commonly used technology in machine learning, especially when it is necessary to predict the bending results. The bending device is the key equipment for sheet metal processing. The bending process is monitored in real time through sensors and machine learning algorithms, and the motion trajectory and pressure of the bending machine are dynamically adjusted to ensure the bending accuracy of each sheet metal part. In the bending device, especially when bending and loading complex sheet metal parts, there is no abutment structure on both sides of the complex sheet metal parts to keep them balanced, resulting in unstable positioning of the sheet metal parts during the bending process, which may cause the position of the workpiece to shift and affect the bending accuracy. Due to unstable positioning, manual readjustment and correction may be required, wasting time. The lack of a protective structure will expose the operator to a dangerous bending area. Especially when the bending machine is running, the workpiece may bounce and fall due to improper operation, thereby accidentally injuring the operator. Determine the optimization goals: bending angle accuracy, surface quality, defect control, reducing production cycle and improving equipment utilization. Data acquisition: Obtain data about the bending process from the production equipment, including force, displacement, angle, temperature, material thickness Degree and mold type, data preprocessing: clean data, remove noise, normalize the scale of different variables, prepare for subsequent analysis, mathematical modeling and simulation: use finite element analysis method to establish a mathematical model of the bending process, simulate the influence of different bending parameters on bending quality, use machine learning to optimize bending parameters: use historical production data to train machine learning models, predict bending parameter combinations, based on real-time data, machine learning models automatically adjust bending parameters, complex sheet metal parts are placed on the surface of the bonding plate 13 for subsequent bending, and the torsion bar 18 is rotated clockwise, driving the bidirectional threaded rod 16 connected to it to rotate, and the two contact blocks 19 threaded with it move relative to each other. Under the connection between the contact rod 20 and the curved plate 21, the two leveling plates 22 are driven to rotate. For movement, the two leveling plates 22 are respectively in contact with the two sides of the complex sheet metal parts, so that the complex sheet metal parts are in the same straight line, which improves the quality of the sheet metal bending. The two protective covers 23 are located on both sides of the top of the bonding plate 13, which can block the surrounding area, avoid accidental injury to workers during the sheet metal bending process, and improve the safety of the bending device. The contact block 19 moves at one end of the reinforcement bar 17, and the contact block 19 is supported by the two-way threaded rod 16 and the reinforcement bar 17, which improves the stability of the movement of the contact block 19. The arc cover 25 and the anti-external force plate 26 cover the top of the protective cover 23, which increases the shielding range of the surrounding environment. After the bottom surface of the positioning plate 10 is tightly fitted with the surface of the bending table 3, one end of the multiple positioning bars 11 is embedded in the inside of the auxiliary hole.With the threaded connection between the two, the installation of the bending seat 8 and the bending table 3 can be realized. This structure makes it easy to replace the damaged bending seat 8, and solves the problem that the uneven surface of the bending groove 9 affects the bending quality. The bottom of the vertical plate 14 abuts against the top of the pressure plate 15, ensuring the firmness of the vertical plate 14. The two triangular plates 12 are located between the bonding plate 13 and the bending seat 8, and the triangle has stability, which enhances the stability of the bottom of the bonding plate 13. The limiting bolt 29 is rotated in the opposite direction so that one end of it is removed from the concave hole 28 and the anti-external force plate 26. By pulling it out from the inside and moving the anti-external force plate 26 to one side, it can be separated from the protective cover 23, making it easier to maintain and replace the curved cover 25 and the anti-external force plate 26 separately. After one side of the curved plate 21 is fitted with one side of the screed plate 22, one end of the limit bar 24 is embedded into the inside of the embedding hole. The threaded connection between the two can achieve the limit and fixation of the curved plate 21 and the screed plate 22. The protective cover 23 can then be installed on one side of the curved plate 21. The force sensor is mainly used to monitor the pressure applied to the sheet metal to ensure that the bending force is moderate.

[0047] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A complex sheet metal bending device based on machine learning, comprising a bending frame (1), a reinforcement platform (2) and a bending platform (3), wherein the reinforcement platform (2) is arranged inside the bending frame (1), and the bending platform (3) is fixedly installed inside the bending frame (1), characterized in that: Two reinforcing rods (4) are fixedly installed inside the bending frame (1) and the reinforcing platform (2); a hydraulic cylinder (6) is provided on the top of the reinforcing platform (2); the output shaft of the hydraulic cylinder (6) is connected to the bending knife (7); a bending seat (8) is provided on the top of the bending platform (3); a bonding plate (13) is provided on one side of the bending seat (8); two triangular plates (12) are provided at the bottom of the bonding plate (13); a vertical plate (12) is fixedly installed at the bottom of the bonding plate (13); A straight plate (14), a bidirectional threaded rod (16) is provided inside the vertical plate (14), a reinforcement strip (17) is provided inside the vertical plate (14), both ends of the bidirectional threaded rod (16) and the reinforcement strip (17) are sleeved with contact blocks (19), two leveling plates (22) are provided on the top of the bonding plate (13), and a protective cover (23) is provided on one side of the two leveling plates (22), and a pressure-resistant plate (15) is fixedly installed on one side of the bending frame (1).

2. The complex sheet metal bending device based on machine learning according to claim 1, characterized in that: A spring (5) is fixedly installed between the inner walls of two adjacent reinforcement rods (4), a torsion bar (18) is rotatably installed on one side of the vertical plate (14), the two contact blocks (19) are respectively threadedly sleeved on the two ends of the bidirectional threaded rod (16), and an arc plate (21) is provided on the other side of the two leveling plates (22).

3. The complex sheet metal bending device based on machine learning according to claim 2, characterized in that: One end of the torsion bar (18) is fixedly connected to one end of the bidirectional threaded rod (16), one side of each of the two contact blocks (19) is fixedly mounted with a contact rod (20), one end of each of the two contact rods (20) is fixedly connected to one side of each of the two arc-shaped plates (21), one side of each of the two leveling plates (22) is provided with two limit bars (24), and one side of each of the two arc-shaped plates (21) is provided with two embedded holes.

4. The complex sheet metal bending device based on machine learning according to claim 3, characterized in that: One end of the four limiting strips (24) is respectively connected to the internal threads of the four embedded holes, and the positions of the two adjacent limiting strips (24) are corresponding. The tops of the two triangular plates (12) are respectively fixedly connected to the bottoms of the bonding plates (13), and one side of the two triangular plates (12) is respectively fixedly connected to one side of the bending seat (8), and a bending groove (9) is provided at the top of the bending seat (8).

5. The complex sheet metal bending device based on machine learning according to claim 1, characterized in that: Positioning plates (10) are fixedly installed on both sides of the bending seat (8), and a plurality of positioning strips (11) are provided on the top of the two positioning plates (10). A plurality of auxiliary holes are opened on the surface of the bending table (3), and one end of the plurality of positioning strips (11) is respectively connected to the internal threads of the plurality of auxiliary holes.

6. The complex sheet metal bending device based on machine learning according to claim 1, characterized in that: Two arc blocks (27) are fixedly mounted on the tops of the two protective covers (23), arc covers (25) are provided on the tops of the two protective covers (23), and external force resistance plates (26) are provided inside the two arc covers (25).

7. The complex sheet metal bending device based on machine learning according to claim 6, characterized in that: One side of the four arc blocks (27) is provided with a concave hole (28), and one side of the four arc blocks (27) is provided with a limiting bolt (29). The four limiting bolts (29) are respectively connected to the four concave holes (28) and the internal threads of the two anti-external force plates (26). A control panel (30) is fixedly installed on one side of the bending frame (1), and a force sensor is provided on one side of the bending knife (7).

8. A bending parameter optimization process for a complex sheet metal bending device based on machine learning, the method being applicable to the complex sheet metal bending device based on machine learning described in claims 1-7, characterized in that : It includes the following steps: Step 1: Determine the optimization goals: bending angle accuracy, surface quality, defect control, reducing production cycle and improving equipment utilization; Step 2: Data acquisition: Obtain data about the bending process from the production equipment, including force, displacement, angle, temperature, material thickness and mold type. Data preprocessing: Clean the data, remove noise, and normalize the scale of different variables to prepare for subsequent analysis. Step 3: Mathematical modeling and simulation: Use the finite element analysis method to establish a mathematical model of the bending process and simulate the influence of different bending parameters on the bending quality; Step 4: Conduct experiments and adjust parameters: Based on the simulation results, conduct small-batch trial production to test the bending effect and adjust parameters: adjust the bending parameters according to the experimental results to ensure that they meet the target requirements; Step 5: Use machine learning to optimize bending parameters: Use historical production data to train the machine learning model to predict bending parameter combinations. Based on real-time data, the machine learning model automatically adjusts the bending parameters. Step 6: Quality inspection and feedback: Sensors are used to monitor various parameters of the bending process in real time. During the production process, new production data is continuously collected and used to further train the machine learning model, thereby achieving self-optimization of the system.

Images