A control method of a pickling line stretch leveler
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BAOSTEEL ZHANJIANG IRON & STEEL CO LTD
- Filing Date
- 2024-01-02
- Publication Date
- 2026-05-12
AI Technical Summary
Existing tension leveling machine control methods are insufficient in terms of accuracy and operational complexity, resulting in defects such as bends, roll marks, and indentations during the production process. Furthermore, they are not effective in controlling hot-formed steel, affecting production pace and yield.
By establishing a data acquisition and analysis system, and using a linear regression model of process parameters, parameters such as the elongation rate and bending roll insertion depth of the tension leveler are automatically set, achieving fully automatic control and reducing operator intervention.
It improved the control accuracy and production efficiency of the tension leveler, reduced the defect rate, simplified the operation process, and ensured product quality.
Smart Images
Figure CN117718361B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of pickling technology, and specifically relates to a method for controlling the process parameters of a tension leveling machine in a cold rolling pickling production line. Background Technology
[0002] The tension leveler, located before the pickling section, is used to level hot-rolled strip steel, breaking down the iron oxide scale and ensuring thorough pickling. It plays a crucial role in the pickling effect. For example, the tension leveler designed and manufactured by Xiangyang Boya Precision Machinery includes a main tension leveler and front and rear tension rollers. The main tension leveler applies bending stress to the strip steel; the front and rear tension rollers apply tensile stress (creating a controllable speed difference). Under the combined effect of these two stresses, the strip steel can achieve moderate elongation even when the tensile stress is far below the material's yield strength. This elongation improves the strip shape, eliminating defects such as waviness, warping, and sickle bends; it also eliminates the yield plateau, improving the strip's mechanical properties. Furthermore, due to the significant difference in properties between the oxide layer and the iron matrix on the strip surface, a large amount of the oxide layer peels off under bending and stretching, and the remaining oxide layer on the surface cracks, thus improving the efficiency of pickling and descaling in the pickling tank. In this way, the straightening machine and the acid tank work together to ensure the descaling effect.
[0003] Most domestic steel mills lack effective control over tension levelers, and several issues remain regarding their control. For example, patent application CN200710042975.0 discloses a tension leveler control method, which mainly controls the tension leveler by setting deviation conditions between the front and rear strips and their corresponding control modes. However, this invention has significant limitations in applicability, especially in controlling special process parameters, where it exhibits considerable errors. Another example is patent application CN201610857948.8, which discloses a tension leveler control method for pickling and rolling mills. This method sets different tension leveler parameters based on the yield strength level of the target steel grade, achieving graded control of the tension leveling parameters. However, this method has certain accuracy issues and is overly complex to operate.
[0004] Currently, the main problems in the control of tension leveling machines are as follows:
[0005] (1) The parameters of tension leveling elongation and tension leveling roller insertion depth are not set reasonably. Defects such as bends, roller marks, and indentations are prone to occur during the production process. There is a lot of human intervention during the production process, which seriously affects the production rhythm, easily leads to machine stoppage in the process section, generates stoppage spots, and affects the unit yield.
[0006] (2) The tension straightening process of hot-formed steel is not set properly. For hot-formed steel, the tension straightening elongation rate is not set properly, which makes it impossible to effectively control the shape of the steel plate. The control effect of hot-formed steel is poor. Micro-cracks are easily generated on the surface during tension straightening, and yellow spot defects exist. Summary of the Invention
[0007] The purpose of this invention is to provide a control method for a pickling line straightening machine that is highly accurate, fast, and easy to operate.
[0008] To achieve the above-mentioned technical objectives, the present invention adopts the following technical solution: a control method for a pickling production line straightening machine, characterized by comprising the following steps:
[0009] (1) Use a host computer server to connect to the PLC of the straightening machine and read the process parameters of the ID number of the current coiled strip. The process parameters include strip thickness, width, steel type, actual elongation, bending roll insertion depth 1, bending roll insertion depth 2, and straightening roll insertion depth.
[0010] (2) Record the relevant process parameters in the host computer at an interval of once per second and store them;
[0011] (3) After the strip passes through the tension leveler, the recorded process parameters are screened and processed. Based on the PLC signal given by the manual on-site observation, it is determined whether the production process of this coil of steel has recording value. If it has recording value, the average value of the process parameters of the tension leveler along the entire length of the coil of steel and the average value of all data in the length direction are calculated. The selected data are organized into a data sample for training a linear model.
[0012] (4) Add the sample obtained in step (2) to the data sample in step (3) and use it to train the linear regression model of process parameters. The input parameters of the model include strip thickness, width and steel grade, and the output parameters are target elongation, bending roll insertion depth 1, bending roll insertion depth 2 and straightening roll insertion depth.
[0013] (5) When the next coil of strip steel is produced, monitor and correct the production line tracking data. When the next coil of strip steel enters the uncoiler, calculate the corresponding process parameters on the host computer process server according to the linear model trained in step 4, and send them to the PLC for execution.
[0014] (6) Before the next coil of strip arrives at the inlet of the tension leveler, switch the execution process parameters of the tension leveler to the process parameters sent to the PLC in step (5) and execute them;
[0015] (7) Jump to step (2) and start the sample collection and training process of the new cycle.
[0016] Furthermore, in step (4), there are 60 different steel grades on site, each steel grade is represented by a number, from 1 to 60.
[0017] Furthermore, in step (6), when the position of the first weld of the next coil of strip is 5 meters away from the entrance of the straightening machine, the process parameters are switched.
[0018] The beneficial effects of this invention are as follows: By establishing data acquisition, analysis, and screening, and utilizing a linear model for process setting, this invention achieves reasonable parameter settings for parameters such as elongation and bending roll insertion depth during the tension straightening process of new steel coils. It has the following advantages:
[0019] 1. The continuously learning control system will continuously train the linear model based on the process parameters that allow the straightening plate to achieve a good shape on site.
[0020] 2. By setting reasonable parameters for elongation rate, bending roller insertion depth 1, bending roller insertion depth 2, and straightening roller insertion depth, the occurrence rate of plate shape defects can be reduced, thus ensuring product quality.
[0021] 3. Enables fully automatic parameter setting, reducing operator workload and avoiding the risk of operator error.
[0022] 4. The implementation and operation are simple and easy. Attached Figure Description
[0023] Figure 1 is a schematic diagram of the layout of the pickling and straightening machine.
[0024] Figure 2 is a flowchart of the method of the present invention.
[0025] Figure 3 This is a table of process data recorded in step 2 of the method of the present invention.
[0026] Among them, 1 is the steering roller, 2 is the upper bending roller box, 3 is the lower bending roller box, 4 is the lower straightening roller box, and 5 is the strip steel in production. Implementation
[0027] like Figure 2 As shown, the present invention provides a control method for a pickling line straightening machine, the steps of which are as follows.
[0028] Step 1: Connect a host computer server to the PLC controlling the straightening machine. The PLC is a Siemens S400 model. Establish a communication channel with the PLC to read the ID number, thickness, width, steel grade, actual elongation, bending roll insertion depth 1, bending roll insertion depth 2, and straightening roll insertion depth parameters of the current strip steel. This data is stored in the PLC's data storage block. The current production line's data is 220575500100, with a thickness of 2.52 mm, a width of 1250 mm, a steel grade of AN1541E5 (corresponding to the numerical code 10), an actual elongation of 1.25%, bending roll insertion depth 1 of 38.8 mm, bending roll insertion depth 2 of 38.8 mm, and straightening roll insertion depth of 34.7 mm.
[0029] Step 2: Record and store the relevant process data in the host computer at an interval of once per second; for example... Figure 3 As shown.
[0030] Step 3: After the strip passes through the straightening machine at the tail end, the recorded data is filtered. After observation, it is found that the shape of this coil of steel is good. At this time, a recording signal is given. Then, the average value of all data in the length direction is calculated. The thickness is 2.52 mm, the width is 1250 mm, the steel grade is AN1541E5 (corresponding to the number 10), the average actual elongation is 1.29%, the bending roll insertion depth 1 is 38.8 mm, the bending roll insertion depth 2 is 38.8 mm, and the straightening roll insertion depth is 34.7 mm. This data is organized into a data sample for training a linear model.
[0031] Step 4: Add the newly recorded data samples to the previously stored samples and use them to train a linear regression model for the process parameters. The input parameters of this model include thickness, width, and steel grade. There are 60 different steel grades on site, each represented by a number from 1 to 60. The output parameters are the target elongation, bending roll insertion depth 1, bending roll insertion depth 2, and straightening roll insertion depth.
[0032] The parameters of the trained model are as follows:
[0033]
[0034]
[0035]
[0036]
[0037] Here For the target elongation rate, , , These represent the insertion depths of the bending roll (1), the bending roll (2), and the straightening roll (2), respectively. Thick, width, and steel grade represent the thickness, width, and steel grade of the strip, respectively.
[0038] Step 5: When the next coil of strip steel (new coil of strip steel) is produced, monitor and correct the production line tracking data. When the next coil of strip steel enters the uncoiler, calculate its corresponding process parameters on the host computer process server according to the linear model trained in step 4, and send them to the PLC for execution.
[0039] Step 6: When the first weld seam of the new coiled strip is 5 meters away from the straightening machine, switch the process parameters, changing the execution process parameters of the straightening machine to the data calculated and sent to the PLC in step 5. Here, the first weld seam position of the new coiled strip refers to the weld seam position between the next coiled strip and the current strip.
[0040] Step 7, jump to step 2, and begin the sample collection and training process for the new cycle.
[0041] The above content is only used to illustrate the technical solution of the present invention. Simple modifications or equivalent substitutions made by those skilled in the art to the technical solution of the present invention do not depart from the essence and scope of the technical solution of the present invention.
Claims
1. A control method for a straightening machine in a pickling production line, characterized in that, Includes the following steps: (1) Use a host computer server to connect to the PLC of the straightening machine and read the process parameters of the ID number of the current coiled strip. The process parameters include strip thickness, width, steel type, actual elongation, bending roll insertion depth 1, bending roll insertion depth 2, and straightening roll insertion depth. (2) Record the relevant process parameters in the host computer at an interval of once per second and store them; (3) After the strip passes through the straightening machine at the tail end, the recorded process parameters are screened and processed. After observation, it is found that the plate shape of this coil is good. At this time, a recording signal is given. At this time, the average value of all data in the length direction is calculated, and the selected data is organized into a data sample for training a linear model. (4) Add the newly recorded data samples to the previously stored samples and use them to train the linear regression model of process parameters. The input parameters of the model include strip thickness, width and steel grade, and the output parameters are target elongation, bending roll insertion depth 1, bending roll insertion depth 2 and straightening roll insertion depth. (5) When the next coil of strip steel is produced, monitor and correct the production line tracking data. When the next coil of strip steel enters the uncoiler, calculate the corresponding process parameters on the host computer process server according to the linear model trained in step (4) and send them to the PLC for execution. (6) Before the next coil of strip arrives at the inlet of the tension leveler, switch the execution process parameters of the tension leveler to the process parameters sent to the PLC in step (5) and execute them; (7) Jump to step (2) and start the sample collection and training process of the new cycle.
2. The control method for the pickling line straightening machine according to claim 1, characterized in that: In step (4), there are 60 different steel grades on site, each steel grade is represented by a number from 1 to 60.
3. The control method for the pickling line straightening machine as described in claim 1, characterized in that: In step (6), when the first weld of the next strip is 5 meters away from the entrance of the straightening machine, the process parameters are switched.