An automatic control method for side spraying after rolling

Abstract

The present application discloses an automatic control method for side spraying after rolling, belonging to the technical field of hot rolling cooling. The method automatically extracts the width and thickness information of the steel plate by using computer vision technology and deep learning algorithms, and determines the target coordinates of the side spraying device according to the edge shielding width, the side spraying water sealing requirements, the movement of the side spraying device, and the water outlet law parameters, and then automatically adjusts the side spraying position in real time to meet the cooling requirements of steel plates of different specifications. The side spraying control method involved in the present application has a high degree of automation and flexibility. It can not only accurately adjust the water sealing area in real time according to the steel plate information to meet the diversified production of steel products, but also can adjust the layout and position of the side spraying device in real time in combination with the cooling process requirements to control the intensity and effect of side spraying water sealing, and has good engineering application prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of hot rolling cooling, and more particularly, to an automatic control method for side spraying after rolling. Background Art

[0002] In the modern steel industry, the cooling process of steel plates plays a crucial role in ensuring the quality of the final products. Traditional cooling processes mostly adopt fixed side spraying and water sealing systems. By spraying water jets on both sides of the moving direction of the steel plate, uniform cooling of the steel plate surface is achieved. This method not only avoids uneven accumulation of coolant on the steel plate but also effectively breaks the water vapor film generated during the cooling process, thereby improving the cooling efficiency and simplifying the analysis process of the cooling results.

[0003] However, with the development of steel products towards diversification and specialization, fixed side spraying devices are difficult to meet the cooling requirements of steel plates of different specifications. The limitations of the existing technology are that it cannot adapt to changes in the width of the steel plate, resulting in areas where water cannot be sealed in the middle during the production of wide steel plates, while in the case of narrow steel plates, there is an over-coverage of the edge shielding area, affecting the overall cooling quality of the steel plate.

[0004] To solve the above problems, a large number of studies have been carried out by those skilled in the art, and various improvement schemes have been proposed. For example, the patent with the application number 201810298000.2 discloses a side spraying method for post-rolling cooling of hot-rolled strip steel, which adapts to steel plates of different widths by adjusting the arrangement of side spraying devices on both sides of the roller table and the angles and heights of the nozzles. Although this technology improves the production efficiency to a certain extent, it still requires manual pre-setting of the steel plate size and the height of the nozzles. Not only is the degree of automation low, but also manual errors will affect the quality of the finished products. Summary of the Invention

[0005] In view of this, the present application provides an automatic control method for side spraying after rolling, mainly aiming to solve the problems that the current fixed side spraying device cannot meet the cooling requirements of steel plates of different sizes, the efficiency of manually presetting the size and adjusting the nozzle height of the adjustable side spraying device is low and the cost is high, and the overall production flexibility and automation degree are insufficient.

[0006] To achieve the above object, the present application provides the following technical solutions, including:

[0007] When the steel passes through after rolling, a video image of the steel plate is obtained as input information, and the width and thickness of the steel plate to be cooled are predicted through a network model established based on a deep learning algorithm;

[0008] Taking the width and thickness directions of the steel plate as a reference to establish a coordinate system, according to the width, thickness, edge shielding width of the steel plate, the requirements of side spray water sealing, the movement of the side spray device, and the water outlet law parameters, determine the position coordinate formula of the side spray device and save it in the computer;

[0009] When the computer receives the predicted data of the width and thickness of the steel plate fed back by the deep learning algorithm, it outputs the position coordinate result of the side spray device through formula operation and real-time feedbacks it to the cooling control system to realize the automatic movement of the side spray device.

[0010] In some embodiments, the specific process of establishing the network model based on the deep learning algorithm is as follows:

[0011] Collect the video image information of the steel plate during post-rolling steel passing and the actual steel plate label. After obtaining the corresponding data of the two, preprocess it to obtain the training data set;

[0012] Based on the YOLOv5 algorithm, train the input information, identify and extract the steel plate position as the target area. After convolution operation, use at least two layers of neural networks to obtain the prediction box measured by the steel plate position. After extracting the corresponding coordinates and dimensions, compare them with the actual data in the training data set, and fine-tune the data set to establish the optimal neural network model.

[0013] In some embodiments, the movement trajectory of the side spray device is set to reciprocate along the straight line where the right edge of the water beam is located; in the coordinate system, the lower surface in the width direction of the steel plate is the x-axis, the midpoint of the width is the origin, the left side of the steel plate is taken as the research object and is the positive direction, and the thickness direction of the steel plate is the y-axis; the parameters of the coordinate formula include: the width of the steel plate 2(a + b), the thickness of the steel plate c, the edge shielding width a, the water sealing width b, the side spray water beam emission angle θ, and the angle β formed by the side spray water beam and the upper surface of the steel plate; according to the mathematical relationship of each parameter in the coordinate system, the coordinate formula of the side spray device is obtained as:

[0014]

[0015] In some embodiments, N nozzle components that can be independently and automatically moved in the coordinate system through the cooling control system are arranged side by side along the length direction of the steel plate, N≥2; the nozzle components support 360° rotation to control the water outlet law.

[0016] In some embodiments, the nozzle components are numbered from 1 to N in sequence according to the moving direction of the steel plate during steel passing. The movement trajectories of each nozzle component are set to be the same in the coordinate system, but the ranges are different. The movement trajectory ranges between the odd-numbered nozzle components do not overlap, and the movement trajectory ranges between the even-numbered nozzle components and the adjacent odd-numbered nozzle components overlap by 1 / 2 each.

[0017] According to the above technical solution, an automatic control method for side spraying after rolling provided by the present application automatically detects the size of the steel plate through computer vision technology and deep learning algorithms, realizing a high degree of automatic adjustment of the side spraying device. This not only reduces manual intervention, lowers operating costs and error rates, but also improves product quality and production safety. In addition, the present application has strong adaptability, can meet the diverse market demands of steel products, and can adjust the layout and position of the side spraying device in real time in combination with the requirements of the cooling process to control the intensity and effect of side spraying water sealing, which is consistent with the development trend of steel industry automation and intelligence.

[0018] The above description is only an overview of the technical solution of the present application. In order to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the specific embodiments of the present application are specifically described below. Brief Description of the Drawings

[0019] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present application, where:

[0020] Figure 1 shows a schematic flow diagram of the automatic movement of the nozzle in some embodiments of the present application;

[0021] Figure 2 shows a nozzle position layout diagram of the side spraying device in some embodiments of the present application;

[0022] Figure 3 shows a schematic diagram of the automatic movement of the nozzle in some embodiments of the present application; Detailed Description of the Embodiments

[0023] The following uses specific specific examples to illustrate the embodiments of the present invention. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. Among them, the drawings are only for illustrative purposes, showing only schematic diagrams and not physical diagrams, and cannot be understood as a limitation of the present application; for those skilled in the art, it is understandable that some well-known structures or steps and their descriptions in the drawings may be omitted in order to better illustrate the embodiments of the present application.

[0024] Refer to Figures 1-3 , taking the hot rolling production line of a certain factory as an example, an automatic control method for side spraying after rolling involved in the embodiments of the present application will be described.

[0025] An automatic control method for side spraying after rolling provided by an embodiment of the present application, as Figure 1 shown, includes:

[0026] When steel passes through after rolling, the steel plate video image fed back by a high-temperature industrial camera or other imaging devices capable of collecting steel plate position and size information is used as input information. The width and thickness of the steel plate to be cooled are predicted through a network model established based on a deep learning algorithm, where:

[0027] The reason for predicting the width is to determine the width range that needs to be purged and covered by combining the edge shielding width and the side spraying water sealing requirements; predicting the thickness is because when the object of side spraying purging has a certain thickness, the existence of its thickness will affect the final height position of the side spraying device, so it cannot be ignored either.

[0028] And considering that the deep learning algorithm YOLOv5 can not only make good use of image features to extract important information, but also achieve high-speed target detection, can process images in a very short time, give prediction results, meet the high-speed production requirements and maintain high accuracy at the same time. In the scenario of predicting the width and thickness of the steel plate after rolling in the embodiment, it is relatively easy to implement and customize the application requirements for steel plate size measurement. Therefore, YOLOv5 is selected as the deep learning algorithm for establishing the network model. The specific process is as follows: collect the video image information of the steel plate when steel passes through after rolling and the actual steel plate width and thickness labels. After obtaining the corresponding data of the two, perform preprocessing, including adjusting the picture size, data augmentation, etc., to obtain the training data set; then use the pre-trained model based on the YOLOv5 algorithm to train the input information, identify and extract the target area with the steel plate position, and use at least two layers of neural networks to obtain a prediction box measured by the steel plate position after convolution operation. After extracting the corresponding coordinates and dimensions, compare them with the actual data in the training data set, and fine-tune the data set to establish the optimal neural network model.

[0029] Use the above-trained network model to predict the width and thickness of the steel plate to be cooled according to the newly input video image information of steel passing through. Generally, the width of the steel plate on the production line is about between 2000 mm and 5400 mm. In this embodiment, the predicted width of the steel plate is 2800 mm and the thickness is 300 mm.

[0030] A coordinate system is established with the width and thickness directions of the steel plate as references. The lower surface in the width direction of the steel plate is set as the x-axis. Since the side spray device is symmetric about the midpoint in the width direction, the midpoint in the width direction is taken as the origin. The left side of the steel plate is taken as the research object and set as the positive direction, and the thickness direction of the steel plate is the y-axis. Considering the refined water sealing requirements, N nozzle components are arranged in parallel along the length direction of the steel plate, N≥2; within the coordinate system, the side spray device can move automatically as a whole, and each nozzle component can also achieve independent automatic movement through the cooling control system, and each nozzle component supports 360° rotation to control the water outlet law; the nozzle components are numbered from 1 to N in sequence according to the moving direction of the steel plate during steel passing. The movement trajectory of the side spray device is designed to reciprocate along the straight line where the right edge of the water beam is located. The movement trajectories of each nozzle component are the same, but the ranges are different. It is set that the movement trajectory ranges between odd-numbered nozzle components do not overlap, and the movement trajectory ranges between even-numbered nozzle components and adjacent odd-numbered nozzle components overlap by 1 / 2 each, as shown in Figure 2 shown; the refined arrangement of the nozzle components, on the one hand, expands the range of the steel plate width that can be water-sealed, and on the other hand, can also combine the water outlet situation to control the intensity and effect of side spray water sealing in real time by adjusting the arrangement and position of the nozzle components.

[0031] Within the coordinate system, the automatic movement laws of the side spray device and each nozzle component are similar, as shown in Figure 3 shown. Taking a certain nozzle component of the side spray device as an example, h and h’ are two heights formed by the nozzle moving in its y direction respectively. The parameters within the coordinate system include: the steel plate width 2(a + b), the steel plate thickness c, the edge shielding width a, the water sealing width b, the side spray water beam emission angle θ, and the angle β formed by the side spray water beam and the upper surface of the steel plate; according to the mathematical relationships of the parameters within the coordinate system, the coordinate formula of the nozzle component is obtained as:

[0032]

[0033] The above horizontal and vertical coordinate formulas are saved as equalities in the python code. When the computer receives the predicted data of the steel plate width and thickness fed back by the YOLOv5 algorithm, that is, 2(a + b) = 2800mm and c = 300mm, combined with the known edge shielding width a = 150mm, the set spray water beam emission angle θ = 25°, and the angle β formed by the side spray water beam and the upper surface of the steel plate β = 30°, through python calculation, the target coordinate output result (1473.25, 428.89) of the final movement of the nozzle component is obtained.

[0034] The cooling control system receives the target coordinate output result of the feedback nozzle component and controls in real time to make the nozzle component automatically move to the corresponding position, realizing the automatic control of side spray during post-rolling cooling.

[0035] As used in the specification, claims and the above drawings of this application, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that comprises a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0036] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the present technical solution, and they should all be covered within the scope of the claims of the present invention.

Claims

1. An automatic control method for side spraying after rolling, characterized in that, It includes the following steps: a) When the steel plate passes through after rolling, obtain the video image of the steel plate as input information, and predict the width and thickness of the steel plate to be cooled through a network model established based on the deep learning algorithm; b) Establish a coordinate system with the width and thickness directions of the steel plate as references. According to the width, thickness, edge shielding width of the steel plate, as well as the side spray water sealing requirements, the movement of the side spray device, and the water outlet law parameters, determine the position coordinate formula of the side spray device and save it in the computer; c) When the computer receives the predicted data of the width and thickness of the steel plate fed back by the deep learning algorithm, output the position coordinate result of the side spray device through formula operation and feedback it to the cooling control system in real time to realize the automatic movement of the side spray device.

2. The automatic control method for side spraying after rolling according to claim 1, wherein The specific process of establishing the network model based on the deep learning algorithm is as follows: a1) Collect the video image information of the steel plate when it passes through after rolling and the actual steel plate label. After obtaining the corresponding data of the two, obtain the training data set through preprocessing; a2) Based on the YOLOv5 algorithm, train the input information, identify and extract the steel plate position as the target area. After convolution operation, use at least two layers of neural network to obtain the prediction box measured by the steel plate position. After extracting the corresponding coordinates and dimensions, compare them with the actual data in the training data set, and fine-tune the training data set to establish the optimal neural network model.

3. The automatic control method for side spray after rolling according to claim 1, wherein The movement trajectory of the side spray device is set to reciprocate along the straight line where the right edge of the water beam is located; in the coordinate system, the lower surface in the width direction of the steel plate is the x-axis, the midpoint of the width is the origin, and the left steel plate is taken as the research object and taken as the positive direction, and the thickness direction of the steel plate is the y-axis; the parameters of the position coordinate formula include: the width of the steel plate 2(a + b), the thickness of the steel plate c, the edge shielding width a, the water sealing width b, the emission angle θ of the side spray water beam, and the angle β formed by the side spray water beam and the upper surface of the steel plate; according to the mathematical relationship of each parameter in the coordinate system, the position coordinate formula of the side spray device is obtained as:

4. A post-rolling cooling side-spray automatic control method according to any one of claims 1 to 3, characterized in that, The side spray device is provided with N nozzle components arranged side by side along the length direction of the steel plate, which can realize independent automatic movement in the coordinate system through the cooling control system, N≥2; the nozzle components support 360° rotation to control the water outlet law.

5. The automatic control method for side spraying after rolling according to claim 4, wherein The nozzle components are numbered from 1 to N in sequence according to the moving direction of the steel plate when it passes through. The movement trajectories of each nozzle component are set to be the same in the coordinate system, but the ranges are different. The movement trajectory ranges between the odd-numbered nozzle components do not overlap, and the movement trajectory ranges between the even-numbered nozzle components and the adjacent odd-numbered nozzle components overlap by 1 / 2 each.

Images