Molten aluminum liquid level simulation control system and control method of casting and rolling machine

By collecting and processing images in real time on the casting and rolling mill, extracting surface quality characteristics, and combining machine learning models and PID control algorithms, the problem of difficult time detecting product defects in the casting and rolling mill is solved, and the surface quality and liquid level of the casting and rolling parts are achieved Automatic monitoring and control improves product quality and production efficiency.

CN118926497AInactive Publication Date: 2024-11-12CHANGZHOU INST OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410981196.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-12
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In the production process, existing casting and rolling mills are difficult to detect defects on the product surface in a timely manner, such as dents, spots, bubbles, etc., which affect the quality of casting and rolling products.

Method used

The cast-rolled image acquisition module, image preprocessing module, image feature extraction module and surface quality detection model are adopted, combined with the PID control algorithm, real-time online detection of the surface quality of cast-rolled parts and automatic liquid level adjustment.

Benefits of technology

It realizes timely detection of product defects during the casting and rolling mill production process, ensures the quality of casting and rolling products, and ensures high-quality surfaces and appropriate liquid levels through automatic liquid level adjustment, improving the automatic monitoring and control efficiency of aluminum liquid level.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118926497A_ABST
    Figure CN118926497A_ABST
Patent Text Reader

Abstract

The invention discloses a cast-rolling machine molten aluminum liquid level simulation control system and method, and the system comprises a cast-rolling piece image collection module, a cast-rolling piece image preprocessing module, a cast-rolling piece image feature extraction module, a cast-rolling piece surface quality detection model training module, and a liquid level simulation control system module. Technical means such as the cast-rolled piece image acquisition module are adopted, so that the technical problems in the prior art are effectively solved, and the surface quality of the cast-rolled piece is detected by acquiring image data of the cast-rolled piece and utilizing a trained model in the production process of the cast-rolled piece by the cast-rolling machine; according to the surface quality detection device, defects on a product can be conveniently detected in time, the quality of the cast-rolled product is ensured, and a surface quality detection result is taken as a feedback signal and is used for adjusting a liquid level control system, so that a proper liquid level is kept while a high-quality surface is ensured, and the effects of automatically monitoring and controlling the surface quality of a cast-rolled piece and the liquid level of molten aluminum are achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of casting and rolling mills, and more specifically, to a casting and rolling mill aluminum liquid level simulation control system and control method. Background Art

[0002] Casting and rolling mill is an important equipment for producing aluminum strips, aluminum alloy strips and other products. It is usually used for the continuous casting and rolling process of aluminum plates. This equipment continuously casts aluminum liquid into slabs, and then rolls the slabs into the required thickness and width through a series of rolling processes, and finally produces aluminum plate products with specific size and surface quality.

[0003] The document with prior art publication number CN204122710U provides an aluminum casting and rolling mill. The device uses a manipulator to break the aluminum plate containing impurities initially rolled out. It has automatic operation and high working efficiency, solves the problems of high labor intensity and poor working environment in manual breaking of aluminum plates, and is conducive to the automation of aluminum casting and rolling.

[0004] Although the above-mentioned prior art solution can achieve the relevant beneficial effects through the structure of the prior art, it still has the following defects: when the device is used, due to the defects (such as dents, spots, bubbles, etc.) on the surface of the cast-rolled product during production, the device is not convenient for timely detecting the defects on the product, which will affect the quality of the cast-rolled product and has poor practicality.

[0005] In view of this, we propose a casting and rolling mill aluminum liquid level simulation control system and control method. Summary of the invention

[0006] 1. Technical issues to be solved

[0007] The purpose of the present application is to provide a casting and rolling mill aluminum liquid level simulation control system and control method, which solves the technical problems raised in the above-mentioned background technology, and achieves the technical effect of facilitating and timely detecting defects on the product and ensuring the quality of the cast and rolled products.

[0008] 2. Technical solution

[0009] The embodiment of the present application provides a casting and rolling mill aluminum liquid level simulation control system and control method, including: a casting image acquisition module, a casting image preprocessing module, a casting image feature extraction module, a casting surface quality detection model training module, and a liquid level simulation control system module.

[0010] Casting image acquisition module: The casting image acquisition module installs a high-resolution camera on the casting mill to capture the surface image of the casting;

[0011] Casting image preprocessing module: the casting image preprocessing module preprocesses the image captured by the camera, including grayscale, edge detection, etc.;

[0012] Casting image feature extraction module: The casting image feature extraction module extracts features related to surface quality from the image, such as surface flatness, bubbles, cracks, etc.;

[0013] Casting surface quality detection model training module: the casting surface quality detection model training module trains a machine learning model for learning the relationship between image features and surface quality;

[0014] Liquid level simulation control system module: The liquid level simulation control system module adopts PID (proportional, integral, differential) control algorithm to automatically adjust the liquid level according to the feedback of the liquid level sensor.

[0015] As an optional solution to the technical solution of this application document, the cast image acquisition module installs a high-resolution camera on the casting machine to capture the surface image of the cast in real time and transmit it to the system for storage.

[0016] As an optional solution of the technical solution of the present application document, the cast image preprocessing module preprocesses the image captured by the camera, including graying, edge detection, etc. The graying: The image captured by the camera is usually in color, and graying is the process of converting a color image into a grayscale image, which reduces the dimension of the image data, simplifies subsequent processing, and improves calculation efficiency;

[0017] The edge detection is an important step in image processing, which is used to identify the boundaries and contours in the image. In the detection of castings, edge detection can help highlight the contours of the castings, thereby making it easier to identify and segment the target object.

[0018] As an optional solution of the technical solution of the present application document, the cast image feature extraction module extracts features related to surface quality from the image, such as surface flatness, bubbles, cracks, etc. The surface flatness: the texture feature information of the image is calculated by the gray level co-occurrence matrix method to evaluate the surface flatness;

[0019] The bubbles: bubbles are usually shown as bright areas in the image, and bubble features can be detected and extracted by setting morphological operations and other methods;

[0020] The crack: The crack is usually manifested as an elongated, continuous dark area in an image, and crack features can be extracted by an edge detection algorithm (such as Canny edge detection).

[0021] As an optional solution to the technical solution of the present application document, the casting surface quality detection model training module trains a convolutional neural network machine learning model to learn the relationship between image features and surface quality, and deploys the trained model to actual production to achieve real-time online detection and evaluation of the surface quality of castings. The model is regularly updated and maintained to adapt to changes in the production process and improve the robustness of the model.

[0022] As an optional solution of the technical solution of the present application document, the liquid level simulation control system module adopts a PID (proportional, integral, differential) control algorithm to automatically adjust the liquid level according to the feedback of the liquid level sensor, integrates the surface quality detection model of the casting and the liquid level control system, and uses the result of the surface quality detection as a feedback signal to adjust the liquid level control system to ensure that the appropriate liquid level is maintained while maintaining a high-quality surface, so as to realize automatic monitoring and control of the surface quality of the casting and the aluminum liquid level.

[0023] A casting and rolling mill aluminum liquid level simulation control system and control method, comprising the following steps:

[0024] S1. During the production of cast-rolling parts by a casting-rolling mill, a high-resolution camera is installed on the casting-rolling mill through a casting-rolling part image acquisition module to capture the surface image of the casting-rolling part in real time and transmit it to the system for storage;

[0025] S2. Preprocess the image captured by the camera using the cast image preprocessing module, including grayscale, edge detection, etc. Grayscale: The image captured by the camera is usually in color. Grayscale is the process of converting a color image into a grayscale image, which reduces the dimension of the image data, simplifies subsequent processing, and improves computational efficiency.

[0026] Edge detection: Edge detection is an important step in image processing, which is used to identify the boundaries and contours in the image. In the detection of castings, edge detection can help highlight the contours of the castings, making it easier to identify and segment the target objects.

[0027] S3, then the casting image feature extraction module extracts features related to surface quality from the image, such as surface flatness, bubbles, cracks, etc. Surface flatness: the texture feature information of the image is calculated by the gray level co-occurrence matrix method to evaluate the surface flatness;

[0028] Bubbles: Bubbles usually appear as bright areas in an image. Bubble features can be detected and extracted by setting morphological operations and other methods.

[0029] Cracks: Cracks usually appear as long, continuous dark areas in the image, and their features can be extracted by edge detection algorithms (such as Canny edge detection);

[0030] S4. Train a convolutional neural network machine learning model through the casting surface quality detection model training module to learn the relationship between image features and surface quality. Deploy the trained model to actual production to achieve real-time online detection and evaluation of the casting surface quality. Regularly update and maintain the model to adapt to changes in the production process and improve the robustness of the model.

[0031] S5. According to the liquid level simulation control system module, a PID (proportional, integral, differential) control algorithm is adopted to automatically adjust the liquid level according to the feedback of the liquid level sensor, and the surface quality detection model of the casting and the liquid level control system are integrated. The result of the surface quality detection is used as a feedback signal to adjust the liquid level control system to ensure that the appropriate liquid level is maintained while maintaining a high-quality surface, so as to realize automatic monitoring and control of the surface quality of the casting and the aluminum liquid level.

[0032] 3. Beneficial effects

[0033] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0034] 1. The present application adopts technical means such as a casting image acquisition module, thereby effectively solving the technical problems in the prior art, and further realizing that in the production process of castings by a casting and rolling machine, by collecting image data of the castings, the surface quality of the castings is detected using a trained model, so as to facilitate timely detection of defects on the products and ensure the quality of the cast and rolled products. The results of the surface quality detection are used as feedback signals to adjust the liquid level control system to ensure that while maintaining a high-quality surface, an appropriate liquid level is maintained, so as to achieve the effect of automatic monitoring and control of the surface quality of the castings and the aluminum liquid level.

[0035] 2. The present application sets up a turning rack, thereby realizing that after the image of one side of the cast is collected, the cast can be turned over by the turning rack to collect images on the other side, thereby improving the efficiency of the device. At the same time, by setting up a casting limit mechanism, the casting to be turned over can be limited and fixed, making the turning smoother and improving the practicality of the device.

[0036] 3. The present application provides a guide plate to facilitate the guidance of the castings, and provides a classification rack to facilitate the device to classify and guide the castings with surface quality problems and the castings with good quality, facilitate classification and placement, and further improve the practicality of the device. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A schematic diagram of the overall structure of the casting and rolling mill aluminum liquid level simulation control system and control method disclosed in a preferred embodiment of the present application, and a flow chart of the surface quality detection process of the casting and rolling part;

[0038] Figure 2 This is a schematic diagram of the overall structure of a casting and rolling mill aluminum liquid level simulation control system and control method disclosed in a preferred embodiment of the present application;

[0039] Figure 3 A schematic diagram of the structure of a feed frame of a casting and rolling mill aluminum liquid level simulation control system and control method disclosed in a preferred embodiment of the present application;

[0040] Figure 4 A schematic diagram of the structure of the feeding mechanism of the aluminum liquid level simulation control system and control method of the casting and rolling mill disclosed in a preferred embodiment of the present application;

[0041] Figure 5 A schematic diagram of the structure of a material turning rack of a casting and rolling mill aluminum liquid level simulation control system and control method disclosed in a preferred embodiment of the present application;

[0042] Figure 6 A schematic cross-sectional view of the structure of the casting and rolling part limiting mechanism, etc. of the casting and rolling mill aluminum liquid level simulation control system and control method disclosed in a preferred embodiment of the present application;

[0043] Explanation of the numbers in the figure: 1. Casting and rolling machine body; 2. Feeding rack; 3. Feeding mechanism; 4. Turning rack; 5. Casting and rolling part limiting mechanism; 201. Guide plate; 202. Electric push rod A; 203. Classification rack; 301. Conveying shaft; 302. Gear; 303. Transmission wheel; 304. Motor; 401. Fixed rack; 402. Electric push rod B; 403. Rack; 404. Adjusting wheel; 501. Electric push rod C; 502. Inclined panel; 503. Limiting plate; 504. Tension spring. DETAILED DESCRIPTION

[0044] The present application is further described in detail below in conjunction with the accompanying drawings.

[0045] Reference Figure 1 A casting and rolling mill aluminum liquid level simulation control system and control method, comprising: a casting image acquisition module, a casting image preprocessing module, a casting image feature extraction module, a casting surface quality detection model training module, and a liquid level simulation control system module.

[0046] Casting image acquisition module: The casting image acquisition module installs a high-resolution camera on the casting mill to capture the surface image of the casting;

[0047] Casting image preprocessing module: The casting image preprocessing module preprocesses the image captured by the camera, including grayscale, edge detection, etc.

[0048] Casting image feature extraction module: The casting image feature extraction module extracts features related to surface quality from images, such as surface flatness, bubbles, cracks, etc.

[0049] Casting surface quality detection model training module: The casting surface quality detection model training module trains a machine learning model to learn the relationship between image features and surface quality;

[0050] Liquid level simulation control system module: The liquid level simulation control system module adopts PID proportional, integral and differential control algorithms to automatically adjust the liquid level according to the feedback from the liquid level sensor.

[0051] The casting image acquisition module installs a high-resolution camera on the casting and rolling mill to capture the surface image of the casting in real time and transmit it to the system for storage.

[0052] The casting image preprocessing module preprocesses the images captured by the camera, including grayscale, edge detection, etc. Grayscale: The images captured by the camera are usually in color. Grayscale is the process of converting color images into grayscale images. It reduces the dimension of image data, simplifies subsequent processing, and improves computing efficiency.

[0053] Edge detection: Edge detection is an important step in image processing, which is used to identify boundaries and contours in images. In casting inspection, edge detection can help highlight the contours of castings, making it easier to identify and segment target objects.

[0054] The casting image feature extraction module extracts features related to surface quality from the image, such as surface flatness, bubbles, cracks, etc. Surface flatness: The surface flatness is evaluated by calculating the texture feature information of the image using the gray level co-occurrence matrix method;

[0055] Bubbles: Bubbles usually appear as bright areas in an image. Bubble features can be detected and extracted by setting morphological operations and other methods.

[0056] Cracks: Cracks usually appear as long, continuous dark areas in the image, and crack features can be extracted using edge detection algorithms such as the Canny edge detection method.

[0057] The casting surface quality detection model training module trains a convolutional neural network machine learning model to learn the relationship between image features and surface quality, deploys the trained model to actual production, realizes real-time online detection and evaluation of the casting surface quality, and regularly updates and maintains the model to adapt to changes in the production process and improve the robustness of the model.

[0058] The liquid level simulation control system module adopts PID proportional, integral and differential control algorithms, automatically adjusts the liquid level according to the feedback of the liquid level sensor, integrates the surface quality detection model of the casting and the liquid level control system, and uses the results of the surface quality detection as a feedback signal to adjust the liquid level control system to ensure that the appropriate liquid level is maintained while maintaining a high-quality surface, so as to realize automatic monitoring and control of the surface quality of the casting and the aluminum liquid level.

[0059] A casting and rolling mill aluminum liquid level simulation control system and control method, comprising the following steps:

[0060] S1. During the production of cast-rolling parts by a casting-rolling mill, a high-resolution camera is installed on the casting-rolling mill through a casting-rolling part image acquisition module to capture the surface image of the casting-rolling part in real time and transmit it to the system for storage;

[0061] S2. Preprocess the image captured by the camera using the cast image preprocessing module, including grayscale, edge detection, etc. Grayscale: The image captured by the camera is usually in color. Grayscale is the process of converting a color image into a grayscale image, which reduces the dimension of the image data, simplifies subsequent processing, and improves computational efficiency.

[0062] Edge detection: Edge detection is an important step in image processing, which is used to identify the boundaries and contours in the image. In the detection of castings, edge detection can help highlight the contours of the castings, making it easier to identify and segment the target objects.

[0063] S3, then the casting image feature extraction module extracts features related to surface quality from the image, such as surface flatness, bubbles, cracks, etc. Surface flatness: the texture feature information of the image is calculated by the gray level co-occurrence matrix method to evaluate the surface flatness;

[0064] Bubbles: Bubbles usually appear as bright areas in an image. Bubble features can be detected and extracted by setting morphological operations and other methods.

[0065] Cracks: Cracks usually appear as long, continuous dark areas in the image, and their features can be extracted by edge detection algorithms such as the Canny edge detection method;

[0066] S4. Train a convolutional neural network machine learning model through the casting surface quality detection model training module to learn the relationship between image features and surface quality. Deploy the trained model to actual production to achieve real-time online detection and evaluation of the casting surface quality. Regularly update and maintain the model to adapt to changes in the production process and improve the robustness of the model.

[0067] S5. According to the liquid level simulation control system module, PID proportional, integral and differential control algorithms are adopted to automatically adjust the liquid level according to the feedback of the liquid level sensor, integrate the surface quality detection model of the casting and the liquid level control system, and use the result of the surface quality detection as a feedback signal to adjust the liquid level control system to ensure that the appropriate liquid level is maintained while maintaining a high-quality surface, so as to realize automatic monitoring and control of the surface quality of the casting and the aluminum liquid level.

[0068] Reference Figure 2 A casting and rolling mill aluminum liquid level simulation control system and control method, comprising a casting and rolling mill body 1, a feeding frame 2, a feeding mechanism 3, a turning frame 4, and a casting and rolling piece limiting mechanism 5,

[0069] A feed frame 2 is fixedly connected to the casting and rolling machine body 1, a feed mechanism 3 is rotatably connected to the feed frame 2, a tipping frame 4 is fixedly connected to the feed frame 2, and a plurality of casting and rolling part limiting mechanisms 5 are fixedly connected to the tipping frame 4.

[0070] Reference Figure 3 A plurality of guide plates 201 are fixedly connected to the feed frame 2, an electric push rod A202 is fixedly connected to one end of the feed frame 2 away from the casting and rolling machine body 1, and a classification frame 203 is fixedly connected to the electric push rod A202.

[0071] Reference Figure 4 The feeding mechanism 3 includes a plurality of conveying shafts 301, which are rotatably connected to the feeding frame 2. A plurality of gears 302 are fixedly connected to one end of the conveying shaft 301, and a transmission wheel 303 is meshed and transmitted between two adjacent gears 302. The transmission wheel 303 is rotatably connected to the feeding frame 2, and a motor 304 is fixedly connected to one of the conveying shafts 301.

[0072] Reference Figure 5 The material turning frame 4 is rotatably connected with a fixed frame 401 at one end, the fixed frame 401 is fixedly connected with the feeding frame 2, a plurality of cameras are mounted on the fixed frame 401, an electric push rod B402 is fixedly connected with the fixed frame 401, the electric push rod B402 is connected with a rack 403, and the rack 403 is meshed with a regulating wheel 404 connected to one end of the material turning frame 4. The material turning frame 4 realizes the turning of the casting.

[0073] Reference Figure 6 The casting limit mechanism 5 includes an electric push rod C501, which is connected and fixed with an inclined plate 502, and a limit plate 503 is provided on one side of the inclined plate 502 in sliding contact, and the limit plate 503 is slidably matched with the material turning frame 4, and a tension spring 504 is connected and fixed between the limit plate 503 and the material turning frame 4. The inclined plate 502 realizes the function of lifting the limit plate 503.

[0074] The implementation principle of a casting and rolling mill aluminum liquid level simulation control system and control method in the embodiment of the present application is as follows: when the casting and rolling mill main body 1 is producing castings, the castings will fall onto the feeding mechanism 3, at which time the motor 304 will drive the conveying shaft 301 to rotate, and then drive the transmission wheel 303 to rotate through the gear 302, thereby driving multiple conveying shafts 301 to rotate and drive the casting to move, and then the guide plate 201 guides the casting, and then when the casting moves to the tipping rack 4, the electric push rod B402 drives the rack 403 to move, so that the rack 403 drives the tipping rack 4 connected to the adjusting wheel 404 to flip, and at this time the electric push rod C501 moves away, the inclined plate 502 does not limit the limit plate 503, and the casting is clamped under the action of the tension spring 504, and then the tipping rack 4 drives the casting to flip;

[0075] At this time, a high-resolution camera is installed on the casting and rolling mill through a casting and rolling piece image acquisition module to capture the surface image of the casting and rolling piece in real time and transmit it to the system for storage;

[0076] The image captured by the camera is preprocessed using the cast image preprocessing module, including grayscale, edge detection, etc. Grayscale: The image captured by the camera is usually in color. Grayscale is the process of converting a color image into a grayscale image, which reduces the dimension of the image data, simplifies subsequent processing, and improves computational efficiency.

[0077] Edge detection: Edge detection is an important step in image processing, which is used to identify the boundaries and contours in the image. In the detection of castings, edge detection can help highlight the contours of the castings, making it easier to identify and segment the target objects.

[0078] Then the casting image feature extraction module extracts features related to surface quality from the image, such as surface flatness, bubbles, cracks, etc. Surface flatness: The surface flatness is evaluated by calculating the texture feature information of the image through the gray level co-occurrence matrix method;

[0079] Bubbles: Bubbles usually appear as bright areas in an image. Bubble features can be detected and extracted by setting morphological operations and other methods.

[0080] Cracks: Cracks usually appear as long, continuous dark areas in the image, and their features can be extracted by edge detection algorithms such as the Canny edge detection method;

[0081] A convolutional neural network machine learning model is trained through the casting surface quality detection model training module to learn the relationship between image features and surface quality. The trained model is deployed in actual production to achieve real-time online detection and evaluation of the casting surface quality. The model is regularly updated and maintained to adapt to changes in the production process and improve the robustness of the model.

[0082] According to the liquid level simulation control system module, the PID proportional, integral and differential control algorithms are adopted to automatically adjust the liquid level according to the feedback of the liquid level sensor, integrate the surface quality detection model of the casting and the liquid level control system, and use the result of the surface quality detection as a feedback signal to adjust the liquid level control system to ensure that the appropriate liquid level is maintained while maintaining a high-quality surface, so as to realize automatic monitoring and control of the surface quality of the casting and the aluminum liquid level;

[0083] According to the output results of the model, the classification rack 203 is moved by the electric push rod A202, so that the castings are pushed to the appropriate position for transportation.

Claims

1. A casting and rolling mill aluminum liquid level simulation control system and control method, comprising: a casting image acquisition module, a casting image preprocessing module, a casting image feature extraction module, a casting surface quality detection model training module, and a liquid level simulation control system module, characterized in that: Casting image acquisition module: The casting image acquisition module installs a high-resolution camera on the casting mill to capture the surface image of the casting; Casting image preprocessing module: the casting image preprocessing module preprocesses the image captured by the camera, including grayscale, edge detection, etc.; Casting image feature extraction module: The casting image feature extraction module extracts features related to surface quality from the image, such as surface flatness, bubbles, cracks, etc.; Casting surface quality detection model training module: the casting surface quality detection model training module trains a machine learning model for learning the relationship between image features and surface quality; Liquid level simulation control system module: The liquid level simulation control system module adopts PID (proportional, integral, differential) control algorithm to automatically adjust the liquid level according to the feedback of the liquid level sensor.

2. The casting and rolling mill aluminum liquid level simulation control system and control method according to claim 1, characterized in that: The casting image acquisition module installs a high-resolution camera on the casting and rolling machine to capture the surface image of the casting in real time and transmit it to the system for storage.

3. The casting and rolling mill aluminum liquid level simulation control system and control method according to claim 1, characterized in that: The cast image preprocessing module preprocesses the image captured by the camera, including grayscale, edge detection, etc. The grayscale: The image captured by the camera is usually in color, and grayscale is the process of converting a color image into a grayscale image, which reduces the dimension of the image data, simplifies subsequent processing, and improves calculation efficiency; The edge detection is an important step in image processing, which is used to identify the boundaries and contours in the image. In the detection of castings, edge detection can help highlight the contours of the castings, thereby making it easier to identify and segment the target object.

4. The casting and rolling mill aluminum liquid level simulation control system and control method according to claim 1, characterized in that: The casting image feature extraction module extracts features related to surface quality from the image, such as surface flatness, bubbles, cracks, etc. The surface flatness is evaluated by calculating the texture feature information of the image by using the gray level co-occurrence matrix method; The bubbles: bubbles are usually shown as bright areas in the image, and bubble features can be detected and extracted by setting morphological operations and other methods; The crack: The crack is usually manifested as an elongated, continuous dark area in the image, and the crack features can be extracted by an edge detection algorithm (such as Canny edge detection) method.

5. The casting and rolling mill aluminum liquid level simulation control system and control method according to claim 1, characterized in that: The casting surface quality detection model training module trains a convolutional neural network machine learning model to learn the relationship between image features and surface quality, deploys the trained model to actual production, realizes real-time online detection and evaluation of the casting surface quality, and regularly updates and maintains the model to adapt to changes in the production process and improve the robustness of the model.

6. The casting and rolling mill aluminum liquid level simulation control system and control method according to claim 1, characterized in that: The liquid level simulation control system module adopts a PID (proportional, integral, differential) control algorithm to automatically adjust the liquid level according to the feedback of the liquid level sensor, integrates the surface quality detection model of the casting and the liquid level control system, and uses the result of the surface quality detection as a feedback signal to adjust the liquid level control system to ensure that the appropriate liquid level is maintained while maintaining a high-quality surface, so as to realize automatic monitoring and control of the surface quality of the casting and the aluminum liquid level.

7. A casting and rolling mill aluminum liquid level simulation control system and control method, comprising the following steps: S1. During the production of cast-rolling parts by a casting-rolling mill, a high-resolution camera is installed on the casting-rolling mill through a casting-rolling part image acquisition module to capture the surface image of the casting-rolling part in real time and transmit it to the system for storage; S2. Preprocess the image captured by the camera using the cast image preprocessing module, including grayscale, edge detection, etc. Grayscale: The image captured by the camera is usually in color. Grayscale is the process of converting a color image into a grayscale image, which reduces the dimension of the image data, simplifies subsequent processing, and improves computational efficiency. Edge detection: Edge detection is an important step in image processing, which is used to identify the boundaries and contours in the image. In the detection of castings, edge detection can help highlight the contours of the castings, making it easier to identify and segment the target objects. S3, then the casting image feature extraction module extracts features related to surface quality from the image, such as surface flatness, bubbles, cracks, etc. Surface flatness: the texture feature information of the image is calculated by the gray level co-occurrence matrix method to evaluate the surface flatness; Bubbles: Bubbles usually appear as bright areas in an image. Bubble features can be detected and extracted by setting morphological operations and other methods. Cracks: Cracks usually appear as long, continuous dark areas in the image, and their features can be extracted by edge detection algorithms (such as Canny edge detection); S4. Train a convolutional neural network machine learning model through the casting surface quality detection model training module to learn the relationship between image features and surface quality. Deploy the trained model to actual production to achieve real-time online detection and evaluation of the casting surface quality. Regularly update and maintain the model to adapt to changes in the production process and improve the robustness of the model. S5. According to the liquid level simulation control system module, a PID (proportional, integral, differential) control algorithm is adopted to automatically adjust the liquid level according to the feedback of the liquid level sensor, and the surface quality detection model of the casting and the liquid level control system are integrated. The result of the surface quality detection is used as a feedback signal to adjust the liquid level control system to ensure that the appropriate liquid level is maintained while maintaining a high-quality surface, so as to realize automatic monitoring and control of the surface quality of the casting and the aluminum liquid level.