Unmanned head cutting method and device for coil dividing unit

Abstract

The invention relates to an unmanned head cutting method and device for a coil dividing unit. The method comprises the following steps that (1) a high-definition camera is arranged to collect and recognize the head position of strip steel on a ground roller; (2) the head of the strip steel enters the camera shooting range and starts to track the position of the strip head, and positioning is completed when the head of the strip steel reaches the positioning 4 o'clock or 5 o'clock; (3) cameras are arranged on the crop shear frame and the preparation station platform, and images of the strip steel entering the right pressing roller are collected; (4) calculating the length of an undesirable belt head in real time, when the tracking length reaches the set length or the calculated length of the undesirable belt head, automatically cutting the head by the crop shear, and repeating the step (4) when the cut is unqualified; (5) if the strip steel cannot enter the crop shear smoothly, alarm information is sent out; steel coil strip head position detection and bad strip head length detection are introduced, and meanwhile strip head detection angle closed-loop control and bad strip head length closed-loop control are added in a basic automatic system to achieve automatic strip head cutting.

Description

Technical Field

[0001] This invention relates to the field of hot-rolled strip technology, and in particular to a method and apparatus for unmanned head cutting of a slitting line unit. Background Technology

[0002] The hot rolling slitting mill is a key piece of equipment in the hot rolling process, used to slit hot-rolled steel coils. Its existing basic automation system, particularly the head-cutting control at the preparation station, relies primarily on manual control. The control process is as follows:

[0003] 1. The inlet trolley transports the steel coil from the walking beam to the ground roller at the preparation station;

[0004] 2. The operator manually controls the ground roller to rotate the steel coil head to the 4 o'clock position;

[0005] 3. The right pressure roller presses down, introducing the belt head;

[0006] 4. The operator observes the length of the defective head-down section to determine the cutting length;

[0007] 5. Poor cutting quality of the cutting head due to improper operation of the cutting shears;

[0008] 6. The operator manually controls the ground roller to rotate the steel coil head to the 5 o'clock position.

[0009] However, relying on manual operation to control the position and observe the condition of the lead head requires experienced operators. If the operator is negligent or makes a mistake, it will directly cause unqualified sheared lead heads to enter the slitting line, requiring them to be returned and re-cut. It may even cause the re-cutting and curling of the slitting, which is time-consuming, labor-intensive, and affects production efficiency.

[0010] Chinese Patent Publication No. CN110293137B discloses an automatic strip threading control method for a hot-rolled leveling and slitting mill, comprising the following steps: 1) Determining whether strip threading preparation is complete by acquiring the position of the coiler jaws and the position of the threading guide plate; if threading preparation is complete, proceed to the next step; 2) Starting the hot-rolled leveling and slitting mill to move the strip forward and acquiring the strip head position in real time; 3) When the strip head is positioned at the coiler jaw position, triggering the expansion of the coiler mandrel; 4) After the mandrel expands to the correct position, the coiler automatically establishes micro-tension. Simultaneously, the strip coiling number positioning control is triggered, and the coiling machine tension is dynamically and linearly increased according to the calculated coiling number; 5) When the coiling machine has coiled to the set number of coils, the coiling machine switches to normal production operating tension; The automatic strip threading method of the hot rolling flattening and slitting inspection line of this invention is easy to apply based on the rapid development of current electrical automation control technology. This method realizes the function of automatically threading hot-rolled strip from the exit shear to the coiling machine, effectively improving production efficiency; however, the specific automation process of the hot-rolled slitting strip head shearing part of this patent is not disclosed. Summary of the Invention

[0011] This invention provides a method and apparatus for unmanned cutting of steel coil heads in a slitting line unit. It introduces detection of the position of the steel coil head and detection of the length of the defective head. At the same time, it adds closed-loop control of the head detection angle and closed-loop control of the length of the defective head to the basic automation system to realize automatic cutting at the preparation station.

[0012] To achieve the above objectives, the present invention employs the following technical solution:

[0013] A method for unmanned head cutting of a wire reel unit includes the following steps:

[0014] (1) Set up a high-definition camera to collect and identify the image of the steel head of the steel coil on the ground roller, establish a strip image library, use image recognition technology to screen, clean and correct the image library, standardize the images, and then train the corresponding model through the AI ​​intelligent recognition model Inception-ResNet to classify and detect the images, and transmit the strip data to the basic automation system to complete the strip position detection.

[0015] (2) The basic automation system uses the strip head position information to track the positioning of the strip head. The positioning is completed when the strip head reaches the 4 o'clock or 5 o'clock position. The 4 o'clock position is used to prepare the station to uncoil and shear the defective strip head. The 5 o'clock position is used to position after shearing and prepare the conditions for the next step of sequential control.

[0016] (3) Collect images of the strip head after it enters the right pressure roller. By collecting images of defective strip heads, a dataset of defective strip heads is formed. The defective strip head model is trained to obtain the defective strip head Inception-ResNet model. The images are classified and detected by the model of the AI ​​detection system to obtain the length of the defective strip head. At the same time, the defective strip head and its length are transmitted to the grassroots automation system through industrial Ethernet to realize the automatic detection and cutting of defective strip head and its length.

[0017] (4) The encoder on the ground roller is used to calculate the length of the defective belt head in real time. When the tracking length reaches the detected length of the defective belt head, the head cutter automatically cuts the head. If the cut is not qualified, step (4) is repeated.

[0018] (5) If the AI ​​detection system determines that the strip steel has failed to enter the cutting head shear smoothly, an alarm message will be issued;

[0019] (6) After the defective lead shearing is completed, the basic automated control system uses AI intelligent recognition model to detect and automatically locate the 5-point position as a preparation condition for the next step of sequential control.

[0020] Furthermore, the high-definition camera and AI detection system are used to identify the entry and exit directions of the steel coil rolling, namely the position of the steel coil head within the range of 3 o'clock ±30° and 9 o'clock ±30°.

[0021] Furthermore, the image of the strip entering the right pressure roller is acquired using cameras mounted on the head shear frame and the preparation station platform.

[0022] A device for unmanned head cutting on a slitting line includes a ground roller, a right pressure roller, a bending roller, and a head cutting shear. It also includes a high-definition camera, a camera, an AI detection system, a right pressure roller pressure sensor, a bending roller pressure sensor, and a ground roller encoder. After the ground roller is mounted on the preparation station along the running direction of the slitting line, the right pressure roller, bending roller, and head cutting shear are sequentially activated. The right pressure roller pressure sensor is located at the valve platform of the preparation station to detect the pressure on the rod-side of the right pressure roller hydraulic cylinder; the bending roller pressure sensor detects the pressure on the rod-side of the bending roller hydraulic cylinder; and the ground roller encoder is located on the operating side of the ground roller at the inlet side. The high-definition camera is respectively located on the inlet and outlet sides of the ground roller at the preparation station, and the camera is respectively mounted on the head cutting shear frame and the preparation station platform.

[0023] Furthermore, the ground roller encoder is an incremental encoder, which calculates and tracks the position of the belt head using speed feedback for belt head position calibration.

[0024] Furthermore, when the pressure sensor of the right pressure roller detects that the pressure has reached the set target pressure, the pressing action of the right pressure roller stops.

[0025] Furthermore, when the bending roller pressure sensor detects that the pressure has reached the set target pressure, the bending roller lifting action stops.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] 1) Introduce steel coil head position detection and defective head length detection, and add head detection angle closed-loop control and defective head length closed-loop control to the basic automation control system to realize automatic head cutting at the preparation station;

[0028] 2) Achieve unmanned cutting function in the preparation station area of ​​the slitting line unit, realize full automation in the preparation station area, reduce the workload of slitting line operators, and optimize the manpower of slitting line unit operators;

[0029] 3) The device has a simple structure and can be modified based on the original system with minimal changes, making it easy to implement and promote. Attached Figure Description

[0030] Figure 1 This is a schematic diagram showing the positions of the high-definition cameras installed on the inlet and outlet sides of the preparation station ground roller described in this invention.

[0031] Figure 2 This is a schematic diagram of the camera position on the head-cutting shear frame described in this invention.

[0032] Figure 3This is a schematic diagram of the camera positions set on the preparation station platform described in this invention.

[0033] Figure 4 This is a flowchart illustrating the present invention.

[0034] In the diagram: 1. High-definition camera a 2. High-definition camera b 3. Camera a 4. Camera b 5. Ground roller 6. Steel coil 7. Head cutter frame 8. Right pressure roller 9. Bending roller Detailed Implementation

[0035] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings:

[0036] A device for unmanned head cutting on a slitting line unit includes a ground roller 5, a right pressure roller 8, a bending roller 9, and a head cutting shear. It also includes a high-definition camera, a camera, a right pressure roller pressure sensor, a bending roller pressure sensor, and a ground roller 5 encoder. After the slitting station is positioned on the ground roller 5 along the slitting direction of the steel coil 6, the right pressure roller 8, bending roller 9, and head cutting shear are sequentially set up, with the bending roller 9 positioned above the right pressure roller 8. After the slitting station is positioned on the ground roller along the slitting direction of the steel coil 6, the right pressure roller 8, bending roller 9, and head cutting shear are sequentially activated. The right pressure roller pressure sensor is located at the valve seat of the right pressure roller 8 preparation station and detects the pressure on the rod-side of the hydraulic cylinder of the right pressure roller 8. When the set target pressure is reached, the right pressure roller 8 stops pressing down. The bending roller pressure sensor is located at the valve platform of the preparation station of the bending roller 9. The bending roller pressure sensor detects the pressure on the rod-side of the hydraulic cylinder of the bending roller 9. When the pressure reaches the set target pressure, the bending roller 9 stops lifting up. The ground roller 5 encoder is located on the operating side of the ground roller 5 at the entrance. The ground roller 5 encoder is an incremental encoder that calculates and tracks the position of the lead head for speed feedback and is used for lead head position calibration. The high-definition cameras are respectively located on the entrance and exit sides of the ground roller 5 at the preparation station. The cameras are respectively located on the head cutter frame 7 and the preparation station platform.

[0037] See Figure 4 The present invention discloses a method for unmanned head cutting of a wire reel unit, comprising the following steps:

[0038] (1) See Figure 1A high-definition camera a1 is installed on the inlet side of the preparation station ground roller 5 to identify the head position of the strip steel 6 within a ±30° range of the 3 o'clock position on the cylindrical bottom surface of the steel coil 6. A high-definition camera b2 is installed on the outlet side of the preparation station ground roller 5 to identify the head position of the strip steel 6 within a ±30° range of the 9 o'clock position on the cylindrical bottom surface of the strip steel 6. The images of the head positions of the strip steel 6 on the ground roller 5 are collected and identified to establish a strip head image library. Image recognition technology is used to filter, clean, and correct the images in the library, standardize the images, and then train the corresponding model through the AI ​​intelligent recognition model Inception-ResNet to perform image classification and detection. The strip head data is then transmitted to the basic automation system to complete the strip head position detection.

[0039] (2) The basic automation system uses the lead position information to realize the positioning of the steel coil lead. When the inlet trolley transports the steel coil 6 to the ground roller 5 of the preparation station, the ground roller 5 starts to rotate clockwise. The initial lead may be on the left or right. The steel coil 6 lead enters the camera range of the high-definition camera a1 or high-definition camera b2 and starts to track the lead position. When the head of the steel coil 6 reaches the 4 o'clock position of the positioning, the positioning is completed. The steel coil 6 lead stops at the 4 o'clock position. The 5 o'clock position is used for positioning after shearing and preparation conditions for the next step of sequential control.

[0040] (3) See Figure 2 Camera a3 is set on the head-cutting frame 7, see Figure 3 Camera b4 is set up on the preparation station platform to capture images of the strip 6 after it enters the right pressure roller 8, and images of the strip head after it enters the right pressure roller 8. By acquiring images of defective strip heads, a dataset of defective strip heads is formed. The defective strip head model is trained to obtain the defective strip head Inception-ResNet model. The right pressure roller 8 is raised. When the right pressure roller pressure sensor reaches the set target pressure, the raising action of the right pressure roller 8 stops. The model of the AI ​​detection system is used to classify and detect the images to obtain the length of the defective strip head. At the same time, it is transmitted to the grassroots automation system through industrial Ethernet to realize the automatic detection and cutting of defective strip heads and their lengths.

[0041] (4) When the strip 6 is uncoiled and enters the field of view of the camera a3 on the head cutter frame 7, the camera starts to take pictures continuously and calculates the length of the defective strip head in real time through image recognition technology. When the head of the strip 6 enters the head cutter, the encoder on the ground roller operation side of the ground roller 5 is used to track the length of the strip head. When the tracked length reaches the set length or the calculated length of the defective strip head, the ground roller 5 stops rotating and the head cutter performs automatic head cutting. If the strip head plate shape is not qualified, step (4) is repeated to advance the strip head and cut it again, so as to realize the automatic head cutting function.

[0042] (5) Use the camera on the preparation station platform to monitor the cutting blade of the head cutter. If the AI ​​detection system determines that the strip steel 6 has failed to enter the head cutter smoothly, an alarm message will be issued.

[0043] (6) After the defective strip head is cut off, the ground roller 5 drives the steel coil to rotate clockwise. The encoder of the ground roller 5 determines the position of the strip head. According to the basic automated control system, the AI ​​intelligent recognition model is used for automatic detection, and the high-definition camera a1 or high-definition camera b2 is used to assist in precise positioning. The steel coil strip head stops at the 5 o'clock position.

[0044] The above embodiments are implemented based on the technical solution of the present invention, providing detailed implementation methods and specific operation processes. However, the scope of protection of the present invention is not limited to the above embodiments. Unless otherwise specified, the methods used in the above embodiments are conventional methods.

Claims

1. A method for unmanned cutting of a winding machine unit, characterized in that: The steps include: (1) Set up a high-definition camera to collect and identify the head image of the steel coil on the ground roller, establish a head image library, use image recognition technology to screen, clean, and correct the image library, standardize the image, and then train the AI ​​intelligent recognition model Inception-ResNet to obtain the corresponding model, classify and detect the image, and transmit the head data to the basic automation system to complete the head position detection; (2) The basic automation system uses the strip head position information to track the strip head positioning. When the strip head reaches the 4 o'clock or 5 o'clock position, the positioning is completed. The 4 o'clock position is used for the preparation station to unwind and shear the bad strip head. The 5 o'clock position is used for positioning after the shearing is completed and the next step of sequential control preparation conditions; (3) The image of the strip head after the strip enters the right pressure roller is collected, and a data set of bad strip heads is formed by collecting images of bad strip heads. The model of bad strip heads is trained to obtain the Inception-ResNet model of bad strip heads. The image is classified and detected by the model of the AI ​​detection system to obtain the length of the bad strip heads. At the same time, it is transmitted to the grassroots automation system through industrial Ethernet to realize automatic detection and cutting of bad strip heads and their length; (4) The length of the bad tape head is calculated in real time using the encoder on the ground roller. When the tracking length reaches the detected bad tape head length, the head cutting shear performs automatic head cutting. If the cutting fails, step (4) is repeated. (5) If the AI ​​detection system determines that the strip fails to enter the crop shear smoothly, an alarm message is issued; (6) After the defective lead is cut, the basic automatic control system uses AI intelligent recognition model detection to automatically locate at the 5 o'clock position as a preparation condition for the next step of sequential control.

2. The unmanned cutting method for a winding machine assembly according to claim 1, characterized in that: The high-definition camera and AI detection system are used to identify the rolling entry and exit directions of the steel coil, that is, the steel coil head position within the range of ±30° at 3 o'clock and ±30° at 9 o'clock.

3. The unmanned cutting method for a winding machine assembly according to claim 1, characterized in that: The image of the strip head after the strip enters the right pressure roller is collected by using cameras arranged on the head cutting shear frame and on the preparation station platform.

4. The unmanned head cutting device of the winding machine unit according to claim 1 comprises a ground roller, a right pressure roller, a bending roller and a head cutting shear, characterized in that: It also includes a high-definition camera, a camera, an AI detection system, a right pressure roller pressure sensor, a bending roller pressure sensor and a ground roller encoder. After the ground roller is put on the preparation station along the running direction of the steel coil 5-division winding line preparation station, the right pressure roller, the bending roller and the head cutting shear are linked in sequence. The right pressure roller pressure sensor is arranged at the valve station of the preparation station to detect the pressure on the rod cavity side of the right pressure roller hydraulic cylinder; the bending roller pressure sensor detects the pressure on the rod cavity side of the bending roller hydraulic cylinder, and the ground roller encoder is arranged on the ground roller operating side on the inlet side; the high-definition cameras are respectively arranged on the inlet side and the outlet side of the ground roller of the preparation station, and the cameras are respectively arranged on the head cutting shear frame and the preparation station platform.

5. The unmanned cutting device of the winding machine unit according to claim 3 is characterized in that: The ground roller encoder adopts an incremental encoder, and the speed feedback calculates and tracks the belt head position for belt head position calibration.

6. The unmanned cutting device of the winding machine unit according to claim 3 is characterized in that: When the pressure detected by the right pressure roller pressure sensor reaches the set target pressure, the pressing action of the right pressure roller stops.

7. The unmanned cutting device of the winding machine unit according to claim 3 is characterized in that: The bending roller pressure sensor detects that the pressure reaches a set target pressure and the bending roller lifting action stops.

Images