Processing method and system for automobile steel production line

By obtaining steel images in automotive steel production lines and dynamically adjusting the heating temperature using feature extraction and comparison models, the temperature control problems caused by material differences in traditional heat treatment processes are solved, and the geometric consistency and production efficiency of steel are improved.

CN118521581BActive Publication Date: 2024-11-15NANJING WANXI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410982101.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-15
Estimated Expiration
2044-07-22

AI Technical Summary

Technical Problem

Traditional heat treatment processes are difficult to adapt to the material differences of different batches of steel, resulting in excessive heat treatment temperature, resulting in excessive geometric deformation of steel, affecting the quality and manufacturing efficiency of automotive steel.

Method used

By obtaining the images of steel at different stations, using preset feature extraction models to generate a shape feature set, and dynamically adjusting the heating temperature with geometric feature comparison models, real-time monitoring and precise control of the steel heating process in automotive steel production lines, avoiding Overheating or insufficient heating.

Benefits of technology

It improves the efficiency and accuracy of the heat treatment process, ensures the geometric consistency of steel, improves the overall quality and manufacturing efficiency of automotive steel, and reduces energy waste and steel appearance deterioration.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118521581B_ABST
    Figure CN118521581B_ABST
Patent Text Reader

Abstract

The present application provides a processing method and system for an automobile steel production line. The method obtains a first image of the steel to be detected at the first station of the production line, uses a preset feature extraction model, and generates a first shape feature set based on the first image, then obtains a second image of the steel to be detected at the second station of the production line, and uses a preset feature extraction model to generate a second shape feature set based on the second image, and then uses a preset geometric feature comparison model to determine a temperature adjustment instruction based on the first shape feature set and the second shape feature set, thereby using the temperature adjustment instruction to adjust the heating temperature of the heating station between the first station of the production line and the second station of the production line, so as to efficiently and accurately monitor and adjust the temperature during the heat treatment process to ensure the consistency of the geometric properties of the steel and optimize the heat treatment effect.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to data processing technology, and more particularly to a processing method and system for an automobile steel production line. Background Art

[0002] With the rapid development of the automobile industry, the quality requirements for automobile steel are increasing day by day. As the main material for manufacturing the main structure and key components of automobiles, the performance of automobile steel directly affects the safety, durability and lightweight level of the vehicle.

[0003] In the production process of automotive steel, heat treatment is a crucial step. It not only affects the mechanical properties of steel, such as strength, toughness, and plasticity, but is also directly related to the difficulty of subsequent processing and forming and the dimensional accuracy of the final product.

[0004] However, traditional heat treatment processes often rely on a fixed temperature control mode, which is difficult to adapt to the material differences between different batches of steel and can easily lead to excessive heat treatment temperatures, which in turn causes excessive geometric deformation of different batches of steel. Summary of the invention

[0005] The present application provides a processing method and system for an automobile steel production line, which is used to solve the technical problem that a fixed temperature control mode is difficult to adapt to the material differences of different batches of steel, which easily leads to excessive heat treatment temperature and thus causes excessive geometric deformation of different batches of steel.

[0006] In a first aspect, the present application provides a processing method for an automobile steel production line, comprising:

[0007] Acquire a first image of a steel material to be inspected at a first workstation of a production line, wherein the steel material to be inspected is at a first temperature at the first workstation of the production line;

[0008] Using a preset feature extraction model, and generating a first shape feature set according to the first image, the first shape feature set is used to characterize the geometric features of the steel to be tested at the first temperature;

[0009] Acquire a second image of the steel to be inspected at a second workstation of the production line, where the steel to be inspected is at a second temperature at the second workstation of the production line, and the second temperature is higher than the first temperature;

[0010] Using the preset feature extraction model, and generating a second shape feature set according to the second image, the second shape feature set is used to characterize the geometric features of the steel to be tested at the second temperature;

[0011] A preset geometric feature comparison model is used to determine a temperature adjustment instruction based on the first shape feature set and the second shape feature set. The temperature adjustment instruction is used to adjust the heating temperature of a heating station between the first station of the production line and the second station of the production line.

[0012] In this embodiment, by acquiring images of the steel to be tested at different stations, generating a shape feature set in combination with a preset feature extraction model, and comparing the geometric features based on these feature sets, the temperature of the heating station is dynamically adjusted. This enables real-time monitoring and precise control of the steel heating process in the automotive steel production line, improving the efficiency and accuracy of the heat treatment process. By automatically identifying changes in the shape of steel through image analysis, the heating temperature is adjusted in a timely manner to avoid excessive deformation of the steel due to overheating or insufficient heating, which in turn leads to a decrease in material properties, thereby ensuring the geometric consistency of the steel during the heating process and improving the overall quality and manufacturing efficiency of automotive steel.

[0013] Optionally, the steel material to be detected is a steel beam structure, the first image and the second image are images of the steel beam structure at a first viewing angle, and the first viewing angle is a viewing direction perpendicular to a longitudinal section of the steel beam structure; correspondingly, using a preset feature extraction model and generating a first shape feature set according to the first image includes:

[0014] Performing grayscale conversion processing on the first image to generate a first grayscale image;

[0015] Performing edge detection on the first grayscale image using an edge detection model to determine a first edge set of the steel beam structure in the first grayscale image, wherein a contour formed by a plurality of edge lines included in the first edge set corresponds to an outer contour of the beam structure in the first grayscale image, and the first shape feature set includes the first edge set and a first edge endpoint set formed by endpoints of each edge line in the first edge set;

[0016] Correspondingly, the using the preset feature extraction model and generating a second shape feature set according to the second image includes:

[0017] Performing grayscale conversion processing on the second image to generate a second grayscale image;

[0018] The second grayscale image is subjected to edge detection using an edge detection model to determine a second edge set of the steel beam structure in the second grayscale image, wherein a contour formed by a plurality of edge lines included in the second edge set corresponds to an outer contour of the steel beam structure in the second grayscale image, and the second shape feature set includes the second edge set and a second edge endpoint set formed by endpoints of each edge line in the second edge set.

[0019] In this scheme, for the steel beam structure, a viewing angle image perpendicular to the longitudinal section is used. The outer contour of the steel beam is clarified through grayscale conversion and edge detection to form a set of shape features, thereby ensuring the accuracy and robustness of shape feature extraction. Among them, the vertical viewing angle can effectively reduce the shape judgment errors caused by viewing angle deviation, making feature extraction more reliable and improving the accuracy of shape comparison in subsequent processing steps.

[0020] Optionally, using a preset geometric feature comparison model and determining the temperature adjustment instruction according to the first shape feature set and the second shape feature set includes:

[0021] Determine a first feature vector according to the first shape feature set , the first feature vector is a unit vector corresponding to the first vector, the first starting point of the first vector is the first feature point in the first shape feature set, the first ending point of the first vector is the second feature point in the first shape feature set, and the first vector is a vector from the first starting point to the first ending point;

[0022] Determine a second feature vector according to the first shape feature set , the second feature vector is the unit vector corresponding to the second vector, the second starting point of the second vector is the third feature point in the second shape feature set, the second ending point of the second vector is the fourth feature point in the second shape feature set, the second vector is a vector pointing from the second starting point to the second ending point, the first feature point and the third feature point correspond to the same point on the steel beam structure, and the second feature point and the fourth feature point correspond to the same point on the steel beam structure;

[0023] Using formula 1, and according to the first eigenvector And the second eigenvector Determine the first temperature adjustment amount , the formula 1 is:

[0024]

[0025] in, is the first calibration constant, is the second calibration constant;

[0026] According to the first temperature adjustment amount The temperature adjustment instruction is generated.

[0027] In this embodiment, the shape differences are quantified by determining characteristic vectors, and the temperature adjustment amount is calculated based on these vectors, so that the generation of temperature adjustment instructions is more accurate, which can accurately reflect the impact of steel shape changes on temperature requirements, ensure the efficiency and accuracy of the heating process, and avoid energy waste caused by excessive heating and degradation of the steel appearance.

[0028] Optionally, in determining the first feature vector according to the first shape feature set Previously, it also included:

[0029] Determine a first feature vector set based on the first shape feature set ,in, The first feature vector set Middle feature vectors, is a feature vector formed by two edge endpoints in the first edge endpoint set;

[0030] Using formula 2, and according to the first feature vector set Determine the first eigenvector , the formula 2 is:

[0031]

[0032] Correspondingly, in determining the second feature vector according to the first shape feature set Previously, it also included:

[0033] Determine a second feature vector set based on the second shape feature set ,in, The second feature vector set Middle feature vectors, is a feature vector formed by two edge endpoints in the second edge endpoint set;

[0034] According to the first feature vector And the second feature vector set Determine the second eigenvector , where the second eigenvector The corresponding The two edge endpoints of The two edge endpoints of the steel beam structure are corresponding edge endpoints.

[0035] In this embodiment, by constructing a feature vector set and selecting representative feature vectors therefrom, the shape feature comparison process is further refined and optimized, thereby enhancing the comprehensiveness and accuracy of the shape feature comparison. Through the comprehensive evaluation of multiple feature vectors in the set, the reliability of the temperature adjustment instructions is improved, ensuring the optimization and control of the geometric characteristics of the steel during the heat treatment process.

[0036] Optionally, the using a preset geometric feature comparison model and determining the temperature adjustment instruction according to the first shape feature set and the second shape feature set further includes:

[0037] The second temperature adjustment amount is determined by using formula 3 and the fully connected layer in the neural network model , the formula 3 is:

[0038]

[0039] in, is the activation function of the neural network model, is the first feature vector set, is the second eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, is the bias term in the fully connected layer;

[0040] Correspondingly, the first temperature adjustment amount Generating the temperature adjustment instruction includes:

[0041] Using Equation 4, and adjusting the amount according to the first temperature and the second temperature adjustment amount Determine the temperature adjustment amount , the formula 4 is:

[0042]

[0043] in, is the first weight, is the second weight, ,and, ;

[0044] According to the temperature adjustment amount Determine to generate the temperature adjustment instruction.

[0045] In this embodiment, the fully connected layer in the neural network model is used to further refine the calculation of the temperature adjustment amount. Through the nonlinear learning ability of the neural network, the nonlinear relationship between complex shape changes and temperature adjustment can be captured, thereby based on a comprehensive evaluation of multiple eigenvectors. In addition, the maximum shape difference is quantified by combining the eigenvectors corresponding to the two endpoints with the farthest distance. Then, by setting the weights as described above, the above-mentioned maximum shape difference is combined with the results of the comprehensive evaluation, so that the temperature adjustment can not only avoid excessive overall deformation of the steel, but also control the maximum deformation of the steel within a reasonable range, thereby adapting to a wider range of working conditions and improving the adaptability and accuracy of the system.

[0046] Optionally, after using the preset feature extraction model and generating a second shape feature set according to the second image, the method further includes:

[0047] Acquire a third image of the steel to be inspected at a third workstation of the production line, where the steel to be inspected is at a third temperature at the third workstation of the production line;

[0048] Using the preset feature extraction model, and generating a third shape feature set according to the third image, the third shape feature set is used to characterize the geometric features of the steel to be tested at the third temperature, the third temperature being lower than the second temperature;

[0049] Correspondingly, the using of the preset geometric feature comparison model and determining the temperature adjustment instruction according to the first shape feature set and the second shape feature set includes:

[0050] The preset geometric feature comparison model is used to determine the temperature adjustment instruction according to the first shape feature set, the second shape feature set and the third shape feature set.

[0051] In this embodiment, image processing and temperature adjustment instruction generation for the third workstation are added, and geometric feature comparisons of more temperature points are taken into consideration. Through this multi-point temperature monitoring and adjustment mechanism, the temperature control of the entire heat treatment process is made more refined. By considering the data of more workstations, the heat treatment process of steel can be more accurately controlled, reducing the error caused by a single node of temperature control, thereby improving the quality and consistency of the final product.

[0052] Optionally, using a preset geometric feature comparison model and determining the temperature adjustment instruction according to the first shape feature set, the second shape feature set, and the third shape feature set includes:

[0053] Performing grayscale conversion processing on the third image to generate a third grayscale image;

[0054] Performing edge detection on the third grayscale image using an edge detection model to determine a third edge set of the steel beam structure in the third grayscale image, wherein a contour formed by a plurality of edge lines included in the third edge set corresponds to an outer contour of the beam structure in the third grayscale image, and the third shape feature set includes the third edge set and a third edge endpoint set formed by endpoints of each edge line in the third edge set;

[0055] Determine a third feature vector set according to the third shape feature set ,in, is the third eigenvector set Middle feature vectors, is a feature vector formed by two edge endpoints in the third edge endpoint set;

[0056] The third temperature adjustment amount is determined by using Formula 5 and the fully connected layer in the neural network model , the formula 5 is:

[0057]

[0058] in, is the activation function of the neural network model, is the second eigenvector set, is the third eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, is the bias term in the fully connected layer;

[0059] Using Equation 6, and adjusting the amount according to the first temperature and the third temperature adjustment amount Determine the temperature adjustment amount , the formula 6 is:

[0060]

[0061] in, is the third weight, is the fourth weight, and ;

[0062] According to the temperature adjustment amount Determine to generate the temperature adjustment instruction.

[0063] In this step, the temperature management of the entire heat treatment process is further improved. Through the monitoring and adjustment of multiple temperature points at multiple workstations, the whole process of steel heat treatment process is intelligently controlled, which ensures the continuity and accuracy of heat treatment and improves the automation level and product quality of the entire production line. In addition, combined with the different characteristics of the heating and cooling workstations, two weight values ​​with opposite signs are set to realize multi-workstation-based steel heat treatment temperature control.

[0064] In a second aspect, the present application provides a control system for an automobile steel production line, comprising:

[0065] An acquisition module, used to acquire a first image of a steel material to be inspected at a first workstation of a production line, wherein the steel material to be inspected is at a first temperature at the first workstation of the production line;

[0066] A processing module, used for generating a first shape feature set according to the first image by using a preset feature extraction model, wherein the first shape feature set is used for characterizing geometric features of the steel material to be tested at the first temperature;

[0067] The acquisition module is further used to acquire a second image of the steel to be inspected at a second workstation of the production line, where the steel to be inspected is at a second temperature at the second workstation of the production line, and the second temperature is higher than the first temperature;

[0068] The processing module is further used to use the preset feature extraction model to generate a second shape feature set according to the second image, wherein the second shape feature set is used to characterize the geometric features of the steel to be tested at the second temperature;

[0069] The processing module is also used to utilize a preset geometric feature comparison model and determine a temperature adjustment instruction based on the first shape feature set and the second shape feature set. The temperature adjustment instruction is used to adjust the heating temperature of a heating station between the first station of the production line and the second station of the production line.

[0070] Optionally, the steel material to be detected is a steel beam structure, the first image and the second image are images of the steel beam structure at a first viewing angle, and the first viewing angle is a viewing direction perpendicular to a longitudinal section of the steel beam structure; the processing module is specifically used to:

[0071] Performing grayscale conversion processing on the first image to generate a first grayscale image;

[0072] Performing edge detection on the first grayscale image using an edge detection model to determine a first edge set of the steel beam structure in the first grayscale image, wherein a contour formed by a plurality of edge lines included in the first edge set corresponds to an outer contour of the beam structure in the first grayscale image, and the first shape feature set includes the first edge set and a first edge endpoint set formed by endpoints of each edge line in the first edge set;

[0073] Performing grayscale conversion processing on the second image to generate a second grayscale image;

[0074] The second grayscale image is subjected to edge detection using an edge detection model to determine a second edge set of the steel beam structure in the second grayscale image, wherein a contour formed by a plurality of edge lines included in the second edge set corresponds to an outer contour of the steel beam structure in the second grayscale image, and the second shape feature set includes the second edge set and a second edge endpoint set formed by endpoints of each edge line in the second edge set.

[0075] Optionally, the processing module is specifically used to:

[0076] Determine a first feature vector according to the first shape feature set , the first feature vector is a unit vector corresponding to the first vector, the first starting point of the first vector is the first feature point in the first shape feature set, the first ending point of the first vector is the second feature point in the first shape feature set, and the first vector is a vector from the first starting point to the first ending point;

[0077] Determine a second feature vector according to the first shape feature set , the second feature vector is the unit vector corresponding to the second vector, the second starting point of the second vector is the third feature point in the second shape feature set, the second ending point of the second vector is the fourth feature point in the second shape feature set, the second vector is a vector pointing from the second starting point to the second ending point, the first feature point and the third feature point correspond to the same point on the steel beam structure, and the second feature point and the fourth feature point correspond to the same point on the steel beam structure;

[0078] Using formula 1, and according to the first eigenvector And the second eigenvector Determine the first temperature adjustment amount , the formula 1 is:

[0079]

[0080] in, is the first calibration constant, is the second calibration constant;

[0081] According to the first temperature adjustment amount The temperature adjustment instruction is generated.

[0082] Optionally, the processing module is specifically used to:

[0083] Determine a first feature vector set based on the first shape feature set ,in, The first feature vector set Middle feature vectors, is a feature vector formed by two edge endpoints in the first edge endpoint set;

[0084] Using formula 2, and according to the first feature vector set Determine the first eigenvector , the formula 2 is:

[0085]

[0086] Correspondingly, in determining the second feature vector according to the first shape feature set Previously, it also included:

[0087] Determine a second feature vector set based on the second shape feature set ,in, The second feature vector set Middle feature vectors, is a feature vector formed by two edge endpoints in the second edge endpoint set;

[0088] According to the first feature vector And the second feature vector set Determine the second eigenvector , where the second eigenvector The corresponding The two edge endpoints of The two edge endpoints of the steel beam structure are corresponding edge endpoints.

[0089] Optionally, the processing module is specifically used to:

[0090] The second temperature adjustment amount is determined by using formula 3 and the fully connected layer in the neural network model , the formula 3 is:

[0091]

[0092] in, is the activation function of the neural network model, is the first feature vector set, is the second eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, is the bias term in the fully connected layer;

[0093] Correspondingly, the first temperature adjustment amount Generating the temperature adjustment instruction includes:

[0094] Using Equation 4, and adjusting the amount according to the first temperature and the second temperature adjustment amount Determine the temperature adjustment amount , the formula 4 is:

[0095]

[0096] in, is the first weight, is the second weight, ,and, ;

[0097] According to the temperature adjustment amount Determine to generate the temperature adjustment instruction.

[0098] Optionally, the processing module is specifically used to:

[0099] Acquire a third image of the steel to be inspected at a third workstation of the production line, where the steel to be inspected is at a third temperature at the third workstation of the production line;

[0100] Using the preset feature extraction model, and generating a third shape feature set according to the third image, the third shape feature set is used to characterize the geometric features of the steel to be tested at the third temperature, the third temperature being lower than the second temperature;

[0101] Correspondingly, the using of the preset geometric feature comparison model and determining the temperature adjustment instruction according to the first shape feature set and the second shape feature set includes:

[0102] The preset geometric feature comparison model is used to determine the temperature adjustment instruction according to the first shape feature set, the second shape feature set and the third shape feature set.

[0103] Optionally, the processing module is specifically used to:

[0104] Performing grayscale conversion processing on the third image to generate a third grayscale image;

[0105] Performing edge detection on the third grayscale image using an edge detection model to determine a third edge set of the steel beam structure in the third grayscale image, wherein a contour formed by a plurality of edge lines included in the third edge set corresponds to an outer contour of the beam structure in the third grayscale image, and the third shape feature set includes the third edge set and a third edge endpoint set formed by endpoints of each edge line in the third edge set;

[0106] Determine a third feature vector set according to the third shape feature set ,in, is the third eigenvector set Middle feature vectors, is a feature vector formed by two edge endpoints in the third edge endpoint set;

[0107] The third temperature adjustment amount is determined by using Formula 5 and the fully connected layer in the neural network model , the formula 5 is:

[0108]

[0109] in, is the activation function of the neural network model, is the second eigenvector set, is the third eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, is the bias term in the fully connected layer;

[0110] Using Equation 6, and adjusting the amount according to the first temperature and the third temperature adjustment amount Determine the temperature adjustment amount , the formula 6 is:

[0111]

[0112] in, is the third weight, is the fourth weight;

[0113] According to the temperature adjustment amount Determine to generate the temperature adjustment instruction.

[0114] In a third aspect, the present application provides an electronic device, including:

[0115] processor; and,

[0116] A memory, configured to store executable instructions of the processor;

[0117] The processor is configured to perform any possible method described in the first aspect by executing the executable instructions.

[0118] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement any possible method described in the first aspect.

[0119] The processing method and system for an automobile steel production line provided in the present application obtain a first image of the steel to be inspected at the first station of the production line, use a preset feature extraction model, and generate a first shape feature set based on the first image, then obtain a second image of the steel to be inspected at the second station of the production line, use a preset feature extraction model, and generate a second shape feature set based on the second image, then use a preset geometric feature comparison model, and determine a temperature adjustment instruction based on the first shape feature set and the second shape feature set, so as to use the temperature adjustment instruction to adjust the heating temperature of the heating station between the first station of the production line and the second station of the production line, thereby accurately obtaining the changes in the geometric characteristics of the steel after heating, providing a data basis for the dynamic adjustment of the heating process parameters, thereby efficiently and accurately monitoring and adjusting the temperature during the heat treatment process to ensure the consistency of the geometric properties of the steel and optimize the heat treatment effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0120] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0121] Figure 1 is a schematic flow chart of a processing method for an automotive steel production line according to an exemplary embodiment of the present application;

[0122] Figure 2 is a flow chart of a processing method for an automotive steel production line according to another exemplary embodiment of the present application;

[0123] Figure 3 is a schematic structural diagram of a control system for an automobile steel production line according to an exemplary embodiment of the present application;

[0124] Figure 4 It is a schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present application.

[0125] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0126] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0127] During the production process of automotive steel, the steel undergoes temperature changes at different heating stations, which directly affects its geometric shape and subsequent processing performance. Traditional methods rely on fixed heating parameters and manual visual inspection, which makes it difficult to achieve precise temperature control and shape monitoring, and can easily lead to steel deformation or uneven heat treatment, affecting the dimensional accuracy and mechanical properties of the final product. The goal of this application is to achieve automatic detection and analysis of the geometric characteristics of steel before and after heating on the automotive steel production line, and dynamically adjust the heating temperature to maintain the consistency of the steel shape, thereby improving overall production efficiency and product quality.

[0128] In order to achieve the above-mentioned objectives, in the embodiment provided in the present application, first, images of the steel beam structure are obtained at the first and second stations of the production line respectively. The image is subjected to grayscale conversion processing to facilitate subsequent image analysis. Then, the edge detection model is used to identify the outline of the steel beam, form an edge set, and the edge endpoints are recorded at the same time to form a shape feature set. In this way, we can extract the geometric features of the steel at different temperatures from the image. Then, the key edge endpoint pairs are selected from the shape feature set to construct feature vectors, which can characterize the geometric state of the steel at a specific position. By comparing the feature vectors corresponding to the first station and the second station, the difference in the shape of the steel due to temperature changes can be quantified. Then, based on the comparison of the feature vectors, a preset geometric feature comparison model is used to determine the amount of heating temperature that needs to be adjusted. In this process, a linear or nonlinear calculation model and a neural network model are involved to improve the accuracy of the temperature adjustment instruction. Among them, the qualified neural network model analyzes the feature vector through a fully connected layer, combined with an activation function, a weight matrix and a bias term, and outputs a temperature adjustment amount. In order to further improve the accuracy of control, the present application also considers acquiring images at a third workstation to form a third shape feature set, analyzing through similar steps, and finally integrating the information of the three workstations to determine the most appropriate temperature adjustment instructions and achieve more precise temperature control.

[0129] In addition, for the above method steps, the present application also designs a corresponding control system, electronic equipment and computer-readable storage medium to ensure the substantive application of the method. The control system integrates an acquisition module and a processing module, and can automatically perform tasks such as image acquisition, feature extraction, temperature calculation and instruction generation, thereby improving the automation level of the production line.

[0130] It can be seen that the embodiments provided in this application can improve the degree of automation and heat treatment accuracy of the automobile steel production line. By real-time monitoring of changes in the geometric characteristics of steel and dynamically adjusting the heating temperature according to these changes, the deformation of steel caused by improper temperature control is effectively avoided, and the stability of the finished product size and mechanical properties are improved. In addition, the solutions provided by the above embodiments can reduce manual intervention, improve production efficiency, and reduce energy consumption, which is a key technological advancement for the automobile steel manufacturing industry to move towards intelligent manufacturing.

[0131] Figure 1 FIG. 1 is a flow chart of a processing method for an automobile steel production line according to an exemplary embodiment of the present application. Figure 1 As shown, the method provided in this embodiment includes:

[0132] S101, obtaining a first image of the steel to be inspected at the first workstation of the production line.

[0133] In this step, a first image of the steel to be inspected at the first station of the production line may be obtained, and the steel to be inspected is at a first temperature at the first station of the production line. Specifically, a high-resolution camera or other imaging equipment may be arranged at the first station of the production line to ensure that a clear and complete image of the steel to be inspected is captured.

[0134] S102: Utilize a preset feature extraction model and generate a first shape feature set according to the first image.

[0135] In this step, a preset feature extraction model is used to generate a first shape feature set according to the first image, and the first shape feature set is used to characterize the geometric features of the steel to be tested at the first temperature.

[0136] Specifically, if the steel material to be detected is a steel beam structure, the first image is an image of the steel beam structure at a first viewing angle, and the first viewing angle is a viewing direction perpendicular to the longitudinal section of the steel beam structure. The first image may be subjected to grayscale conversion processing to generate a first grayscale image. Then, the edge detection model is used to perform edge detection on the first grayscale image to determine the first edge set of the steel beam structure in the first grayscale image. The contour formed by the multiple edge lines included in the first edge set corresponds to the outer contour of the beam structure in the first grayscale image. The first shape feature set includes the first edge set and the first edge endpoint set formed by the endpoints of each edge line in the first edge set. The above-mentioned edge detection model may be a Canny edge detection, Sobel operator or a more advanced edge detection algorithm to process the grayscale image to accurately identify the contour edge of the steel beam structure. This step generates an edge set, including multiple edge lines that constitute the outer contour of the steel beam.

[0137] Furthermore, from the above edge detection results, the endpoints of all edge lines are extracted to form edge endpoint sets. These endpoints represent the geometric feature points of the steel beam structure, and they are organized to form a first shape feature set, including an edge set and an edge endpoint set. This step is helpful for subsequent analysis of changes in steel shape.

[0138] S103, obtaining a second image of the steel to be inspected at the second workstation of the production line.

[0139] In this step, a second image of the steel to be tested at the second station of the production line may be obtained, where the steel to be tested is at a second temperature at the second station of the production line, and the second temperature is higher than the first temperature. Specifically, when the steel is heated and reaches the second station, the image acquisition and preprocessing steps are performed again. At this time, the steel is at a second temperature, which is higher than the first temperature, and the shape may change. The same image processing technology is used for grayscale conversion and edge detection to generate a second shape feature set, reflecting the geometric characteristics of the steel at a higher temperature.

[0140] S104: Utilize a preset feature extraction model and generate a second shape feature set according to the second image.

[0141] In this step, a preset feature extraction model is used to generate a second shape feature set according to the second image, and the second shape feature set is used to characterize the geometric features of the steel to be tested at the second temperature.

[0142] Specifically, if the steel material to be detected is a steel beam structure, the second image is an image of the steel beam structure at a first viewing angle, and the first viewing angle is a viewing direction perpendicular to the longitudinal section of the steel beam structure. The second image may be subjected to grayscale conversion processing to generate a second grayscale image. The second grayscale image is subjected to edge detection using an edge detection model to determine a second edge set of the steel beam structure in the second grayscale image, the contour formed by the multiple edge lines included in the second edge set corresponds to the outer contour of the steel beam structure in the second grayscale image, and the second shape feature set includes the second edge set and a second edge endpoint set formed by the endpoints of each edge line in the second edge set.

[0143] S105 , using a preset geometric feature comparison model, and determining a temperature adjustment instruction according to the first shape feature set and the second shape feature set.

[0144] In this step, a preset geometric feature comparison model is used to determine a temperature adjustment instruction based on the first shape feature set and the second shape feature set. The temperature adjustment instruction is used to adjust the heating temperature of the heating station between the first station of the production line and the second station of the production line. Optionally, the preset geometric feature comparison model is used to compare the first shape feature set and the second shape feature set to analyze the shape change. Specifically, the distance change and angle change between the feature points can be calculated, and then a specific temperature adjustment instruction is generated based on the calculated temperature adjustment amount. The instruction will be sent to the control system of the heating station to automatically adjust the heating temperature to ensure the geometric dimensional stability and quality consistency of the steel during continuous processing.

[0145] Optionally, the first feature vector may be determined according to the first shape feature set. , the first eigenvector is the unit vector corresponding to the first vector, the first starting point of the first vector is the first feature point in the first shape feature set, the first ending point of the first vector is the second feature point in the first shape feature set, and the first vector is the vector pointing from the first starting point to the first ending point.

[0146] Determine a second feature vector based on the first shape feature set , the second eigenvector is the unit vector corresponding to the second vector, the second starting point of the second vector is the third feature point in the second shape feature set, the second ending point of the second vector is the fourth feature point in the second shape feature set, the second vector is a vector pointing from the second starting point to the second ending point, the first feature point and the third feature point correspond to the same point on the steel beam structure, and the second feature point and the fourth feature point correspond to the same point on the steel beam structure.

[0147] Using formula 1, and according to the first eigenvector and the second eigenvector Determine the first temperature adjustment amount , Formula 1 is:

[0148]

[0149] in, is the first calibration constant, is the second calibration constant.

[0150] Finally, according to the first temperature adjustment amount Generate a temperature adjustment instruction, wherein the temperature adjustment instruction is used to adjust the heating temperature of the heating station. It is worth noting that when the above steps are adapted to the situation where the temperature of the heating station is too high, resulting in excessive deformation of the steel to be tested during the heating process, the above temperature adjustment instruction can be used to lower the heating temperature of the heating station.

[0151] Further, the first feature vector is determined according to the first shape feature set. Before, the first feature vector set can also be determined according to the first shape feature set ,in, is the first eigenvector set Middle feature vectors, is a feature vector formed by two edge endpoints in the first edge endpoint set;

[0152] Using formula 2, and according to the first eigenvector set Determine the first eigenvector , Formula 2 is:

[0153]

[0154] It is worth noting that the first eigenvector set mentioned above Middle feature vector is the first eigenvector set The eigenvector with the largest modulus.

[0155] Correspondingly, the second feature vector is determined according to the first shape feature set. Before, the second feature vector set can also be determined according to the second shape feature set ,in, is the second eigenvector set Middle feature vectors, is a feature vector formed by two edge endpoints in the second edge endpoint set.

[0156] According to the first eigenvector And the second eigenvector set Determine the second eigenvector , where the second eigenvector The corresponding The two edge endpoints of The two edge endpoints of the steel beam structure are corresponding edge endpoints.

[0157] Furthermore, in order to make the temperature adjustment avoid excessive overall deformation of the steel and control the maximum deformation of the steel within a reasonable range, so as to adapt to a wider range of working conditions, the second temperature adjustment amount can be determined by using Formula 3 and the fully connected layer in the neural network model. , Formula 3 is:

[0158]

[0159] in, is the activation function of the neural network model, is the first eigenvector set, is the second eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, It is a bias term in the fully connected layer. It is worth noting that the above-mentioned vector comparison function can be a combination of one or more functions of the Euclidean distance calculation function, the Manhattan distance calculation function, the cosine similarity calculation function, the Pearson correlation coefficient calculation function and the Hamming distance calculation function.

[0160] Then, using Equation 4, adjust the amount according to the first temperature and the second temperature adjustment amount Determine the temperature adjustment amount , Formula 4 is:

[0161]

[0162] in, is the first weight, is the second weight, ,and, ;

[0163] Adjust the amount according to the temperature Determine and generate temperature adjustment instructions. The fully connected layer in the neural network model is used to further refine the calculation of the temperature adjustment amount. Through the nonlinear learning ability of the neural network, the nonlinear relationship between complex shape changes and temperature adjustment can be captured, thereby based on a comprehensive evaluation of multiple eigenvectors. In addition, the maximum shape difference is quantified by combining the eigenvectors corresponding to the two farthest endpoints. Then, through the above-mentioned method of setting weights, the above-mentioned maximum shape difference is combined with the results of the comprehensive evaluation, so that the temperature adjustment can not only avoid excessive overall deformation of the steel, but also control the maximum deformation of the steel within a reasonable range, thereby adapting to a wider range of working conditions and improving the adaptability and accuracy of the system.

[0164] In this embodiment, by obtaining a first image of the steel to be inspected at the first station of the production line, using a preset feature extraction model, and generating a first shape feature set based on the first image, then obtaining a second image of the steel to be inspected at the second station of the production line, and using a preset feature extraction model, and generating a second shape feature set based on the second image, then using a preset geometric feature comparison model, and determining a temperature adjustment instruction based on the first shape feature set and the second shape feature set, thereby using the temperature adjustment instruction to adjust the heating temperature of the heating station between the first station of the production line and the second station of the production line, thereby accurately obtaining the changes in the geometric characteristics of the steel after heating, providing a data basis for the dynamic adjustment of the heating process parameters, thereby efficiently and accurately monitoring and adjusting the temperature during the heat treatment process to ensure the consistency of the geometric properties of the steel and optimize the heat treatment effect.

[0165] Figure 2 FIG. 1 is a flow chart of a processing method for an automobile steel production line according to another exemplary embodiment of the present application. Figure 2 As shown, the processing method for an automobile steel production line provided in this embodiment includes:

[0166] S201. Acquire a first image of the steel to be inspected at the first workstation of the production line.

[0167] In this step, a first image of the steel to be inspected at the first station of the production line may be obtained, and the steel to be inspected is at a first temperature at the first station of the production line. Specifically, a high-resolution camera or other imaging equipment may be arranged at the first station of the production line to ensure that a clear and complete image of the steel to be inspected is captured.

[0168] S202: Utilize a preset feature extraction model and generate a first shape feature set according to the first image.

[0169] In this step, a preset feature extraction model is used to generate a first shape feature set according to the first image, and the first shape feature set is used to characterize the geometric features of the steel to be tested at the first temperature.

[0170] Specifically, if the steel material to be detected is a steel beam structure, the first image is an image of the steel beam structure at a first viewing angle, and the first viewing angle is a viewing direction perpendicular to the longitudinal section of the steel beam structure. The first image may be subjected to grayscale conversion processing to generate a first grayscale image. Then, the edge detection model is used to perform edge detection on the first grayscale image to determine the first edge set of the steel beam structure in the first grayscale image. The contour formed by the multiple edge lines included in the first edge set corresponds to the outer contour of the beam structure in the first grayscale image. The first shape feature set includes the first edge set and the first edge endpoint set formed by the endpoints of each edge line in the first edge set. The above-mentioned edge detection model may be a Canny edge detection, Sobel operator or a more advanced edge detection algorithm to process the grayscale image to accurately identify the contour edge of the steel beam structure. This step generates an edge set, including multiple edge lines that constitute the outer contour of the steel beam.

[0171] Furthermore, from the above edge detection results, the endpoints of all edge lines are extracted to form edge endpoint sets. These endpoints represent the geometric feature points of the steel beam structure, and they are organized to form a first shape feature set, including an edge set and an edge endpoint set. This step is helpful for subsequent analysis of changes in steel shape.

[0172] S203, obtaining a second image of the steel to be inspected at the second workstation of the production line.

[0173] In this step, a second image of the steel to be tested at the second station of the production line may be obtained, where the steel to be tested is at a second temperature at the second station of the production line, and the second temperature is higher than the first temperature. Specifically, when the steel is heated and reaches the second station, the image acquisition and preprocessing steps are performed again. At this time, the steel is at a second temperature, which is higher than the first temperature, and the shape may change. The same image processing technology is used for grayscale conversion and edge detection to generate a second shape feature set, reflecting the geometric characteristics of the steel at a higher temperature.

[0174] S204: Utilize a preset feature extraction model and generate a second shape feature set according to the second image.

[0175] In this step, a preset feature extraction model is used to generate a second shape feature set according to the second image, and the second shape feature set is used to characterize the geometric features of the steel to be tested at the second temperature.

[0176] Specifically, if the steel material to be detected is a steel beam structure, the second image is an image of the steel beam structure at a first viewing angle, and the first viewing angle is a viewing direction perpendicular to the longitudinal section of the steel beam structure. The second image may be subjected to grayscale conversion processing to generate a second grayscale image. The second grayscale image is subjected to edge detection using an edge detection model to determine a second edge set of the steel beam structure in the second grayscale image, the contour formed by the multiple edge lines included in the second edge set corresponds to the outer contour of the steel beam structure in the second grayscale image, and the second shape feature set includes the second edge set and a second edge endpoint set formed by the endpoints of each edge line in the second edge set.

[0177] S205. Acquire a third image of the steel to be inspected at the third workstation of the production line.

[0178] In this step, a third image of the steel to be inspected at the third station of the production line may be obtained, and the steel to be inspected is at a third temperature at the third station of the production line. The second station and the third station may be stations for heating, for example, annealing stations.

[0179] S206: Utilize a preset feature extraction model and generate a third shape feature set according to the third image.

[0180] A preset feature extraction model is used to generate a third shape feature set according to the third image. The third shape feature set is used to characterize the geometric features of the steel to be tested at a third temperature, which is lower than the second temperature.

[0181] Specifically, the third image may be subjected to grayscale conversion processing to generate a third grayscale image. The third grayscale image is subjected to edge detection using an edge detection model to determine a third edge set of the steel beam structure in the third grayscale image. The contour formed by the multiple edge lines included in the third edge set corresponds to the outer contour of the beam structure in the third grayscale image. The third shape feature set includes the third edge set and a third edge endpoint set formed by the endpoints of each edge line in the third edge set. Then, the third feature vector set is determined based on the third shape feature set. ,in, is the third eigenvector set Middle feature vectors, is a feature vector formed by two edge endpoints in the third edge endpoint set;

[0182] S207 , using a preset geometric feature comparison model, and determining a temperature adjustment instruction according to the first shape feature set, the second shape feature set, and the third shape feature set.

[0183] Specifically, formula 1 can be used, and according to the first eigenvector and the second eigenvector Determine the first temperature adjustment amount , Formula 1 is:

[0184]

[0185] in, is the first calibration constant, is the second calibration constant.

[0186] Finally, according to the first temperature adjustment amount A temperature adjustment instruction is generated, wherein the temperature adjustment instruction is used to control the heating temperature of the heating station.

[0187] Then, using Formula 5, the third temperature adjustment amount is determined by the fully connected layer in the neural network model , Formula 5 is:

[0188]

[0189] in, is the activation function of the neural network model, is the second eigenvector set, is the third eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, is the bias term in the fully connected layer;

[0190] Using Equation 6, adjust the amount based on the first temperature and the third temperature adjustment amount Determine the temperature adjustment amount , Formula 6 is:

[0191]

[0192] in, is the third weight, is the fourth weight, and ;

[0193] Adjust the amount according to the temperature It is worth noting that when the above steps are adapted to the situation where the temperature of the heating station is too high and the deformation of the steel to be tested is too large during the heating process, the above temperature adjustment instructions can be used to lower the heating temperature of the heating station, that is, Set to a weight greater than 0. Through the monitoring and adjustment of multiple temperature points at multiple stations, the whole process of steel heat treatment is intelligently controlled, ensuring the continuity and accuracy of heat treatment, and improving the automation level and product quality of the entire production line. In addition, combined with the different characteristics of the heating and cooling stations, two weight values ​​with opposite signs are set to achieve temperature control of steel heat treatment based on multiple stations.

[0194] Figure 3 1 is a schematic diagram of a control system for an automobile steel production line according to an exemplary embodiment of the present application. Figure 3 As shown, the automotive steel production line control system 300 provided in this embodiment includes:

[0195] An acquisition module 310 is used to acquire a first image of a steel material to be inspected at a first station of a production line, where the steel material to be inspected is at a first temperature at the first station of the production line;

[0196] A processing module 320, configured to use a preset feature extraction model and generate a first shape feature set according to the first image, wherein the first shape feature set is used to characterize geometric features of the steel material to be tested at the first temperature;

[0197] The acquisition module 310 is further used to acquire a second image of the steel to be inspected at a second workstation of the production line, where the steel to be inspected is at a second temperature at the second workstation of the production line, and the second temperature is higher than the first temperature;

[0198] The processing module 320 is further used to use the preset feature extraction model to generate a second shape feature set according to the second image, wherein the second shape feature set is used to characterize the geometric features of the steel to be tested at the second temperature;

[0199] The processing module 320 is also used to utilize a preset geometric feature comparison model and determine a temperature adjustment instruction based on the first shape feature set and the second shape feature set. The temperature adjustment instruction is used to adjust the heating temperature of a heating station between the first station of the production line and the second station of the production line.

[0200] Optionally, the steel material to be detected is a steel beam structure, the first image and the second image are images of the steel beam structure at a first viewing angle, and the first viewing angle is a viewing direction perpendicular to a longitudinal section of the steel beam structure; the processing module 320 is specifically used to:

[0201] Performing grayscale conversion processing on the first image to generate a first grayscale image;

[0202] Performing edge detection on the first grayscale image using an edge detection model to determine a first edge set of the steel beam structure in the first grayscale image, wherein a contour formed by a plurality of edge lines included in the first edge set corresponds to an outer contour of the beam structure in the first grayscale image, and the first shape feature set includes the first edge set and a first edge endpoint set formed by endpoints of each edge line in the first edge set;

[0203] Performing grayscale conversion processing on the second image to generate a second grayscale image;

[0204] The second grayscale image is subjected to edge detection using an edge detection model to determine a second edge set of the steel beam structure in the second grayscale image, wherein a contour formed by a plurality of edge lines included in the second edge set corresponds to an outer contour of the steel beam structure in the second grayscale image, and the second shape feature set includes the second edge set and a second edge endpoint set formed by endpoints of each edge line in the second edge set.

[0205] Optionally, the processing module 320 is specifically configured to:

[0206] Determine a first feature vector according to the first shape feature set , the first feature vector is a unit vector corresponding to the first vector, the first starting point of the first vector is the first feature point in the first shape feature set, the first ending point of the first vector is the second feature point in the first shape feature set, and the first vector is a vector from the first starting point to the first ending point;

[0207] Determine a second feature vector according to the first shape feature set , the second feature vector is the unit vector corresponding to the second vector, the second starting point of the second vector is the third feature point in the second shape feature set, the second ending point of the second vector is the fourth feature point in the second shape feature set, the second vector is a vector pointing from the second starting point to the second ending point, the first feature point and the third feature point correspond to the same point on the steel beam structure, and the second feature point and the fourth feature point correspond to the same point on the steel beam structure;

[0208] Using formula 1, and according to the first eigenvector And the second eigenvector Determine the first temperature adjustment amount , the formula 1 is:

[0209]

[0210] in, is the first calibration constant, is the second calibration constant;

[0211] According to the first temperature adjustment amount The temperature adjustment instruction is generated.

[0212] Optionally, the processing module 320 is specifically configured to:

[0213] Determine a first feature vector set based on the first shape feature set ,in, The first feature vector set Middle feature vectors, is a feature vector formed by two edge endpoints in the first edge endpoint set;

[0214] Using formula 2, and according to the first feature vector set Determine the first eigenvector , the formula 2 is:

[0215]

[0216] Correspondingly, in determining the second feature vector according to the first shape feature set Previously, it also included:

[0217] Determine a second feature vector set based on the second shape feature set ,in, The second feature vector set Middle feature vectors, is a feature vector formed by two edge endpoints in the second edge endpoint set;

[0218] According to the first feature vector And the second feature vector set Determine the second eigenvector , where the second eigenvector The corresponding The two edge endpoints of The two edge endpoints of the steel beam structure are corresponding edge endpoints.

[0219] Optionally, the processing module 320 is specifically configured to:

[0220] The second temperature adjustment amount is determined by using formula 3 and the fully connected layer in the neural network model , the formula 3 is:

[0221]

[0222] in, is the activation function of the neural network model, is the first feature vector set, is the second eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, is the bias term in the fully connected layer;

[0223] Correspondingly, the first temperature adjustment amount Generating the temperature adjustment instruction includes:

[0224] Using Equation 4, and adjusting the amount according to the first temperature and the second temperature adjustment amount Determine the temperature adjustment amount , the formula 4 is:

[0225]

[0226] in, is the first weight, is the second weight, ,and, ;

[0227] According to the temperature adjustment amount Determine to generate the temperature adjustment instruction.

[0228] Optionally, the processing module 320 is specifically configured to:

[0229] Acquire a third image of the steel to be inspected at a third workstation of the production line, where the steel to be inspected is at a third temperature at the third workstation of the production line;

[0230] Using the preset feature extraction model, and generating a third shape feature set according to the third image, the third shape feature set is used to characterize the geometric features of the steel to be tested at the third temperature, the third temperature being lower than the second temperature;

[0231] Correspondingly, the using of the preset geometric feature comparison model and determining the temperature adjustment instruction according to the first shape feature set and the second shape feature set includes:

[0232] The preset geometric feature comparison model is used, and the temperature adjustment instruction is determined according to the first shape feature set, the second shape feature set, and the third shape feature set.

[0233] Optionally, the processing module 320 is specifically configured to:

[0234] Performing grayscale conversion processing on the third image to generate a third grayscale image;

[0235] Performing edge detection on the third grayscale image using an edge detection model to determine a third edge set of the steel beam structure in the third grayscale image, wherein a contour formed by a plurality of edge lines included in the third edge set corresponds to an outer contour of the beam structure in the third grayscale image, and the third shape feature set includes the third edge set and a third edge endpoint set formed by endpoints of each edge line in the third edge set;

[0236] Determine a third feature vector set according to the third shape feature set ,in, is the third eigenvector set Middle feature vectors, is a feature vector formed by two edge endpoints in the third edge endpoint set;

[0237] The third temperature adjustment amount is determined by using Formula 5 and the fully connected layer in the neural network model , the formula 5 is:

[0238]

[0239] in, is the activation function of the neural network model, is the second eigenvector set, is the third eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, is the bias term in the fully connected layer;

[0240] Using Equation 6, and adjusting the amount according to the first temperature and the third temperature adjustment amount Determine the temperature adjustment amount , the formula 6 is:

[0241]

[0242] in, is the third weight, is the fourth weight;

[0243] According to the temperature adjustment amount Determine to generate the temperature adjustment instruction.

[0244] Figure 4 is a schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present application. Figure 4 As shown, an electronic device 400 provided in this embodiment includes: a processor 401 and a memory 402; wherein:

[0245] The memory 402 is used to store computer programs, and the memory may also be a flash memory.

[0246] The processor 401 is used to execute the execution instructions stored in the memory to implement each step in the above method. For details, please refer to the relevant description in the above method embodiment.

[0247] Optionally, the memory 402 may be independent or integrated with the processor 401 .

[0248] When the memory 402 is a device independent of the processor 401, the electronic device 400 may further include:

[0249] The bus 403 is used to connect the memory 402 and the processor 401 .

[0250] This embodiment further provides a readable storage medium, in which a computer program is stored. When at least one processor of an electronic device executes the computer program, the electronic device executes the methods provided in the above-mentioned various implementation modes.

[0251] This embodiment also provides a program product, which includes a computer program stored in a readable storage medium. At least one processor of the electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device implements the methods provided in the above various embodiments.

[0252] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the claims.

[0253] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A processing method for an automobile steel production line, characterized in that: include: Acquire a first image of a steel material to be inspected at a first workstation of a production line, wherein the steel material to be inspected is at a first temperature at the first workstation of the production line; Using a preset feature extraction model, and generating a first shape feature set according to the first image, the first shape feature set is used to characterize the geometric features of the steel to be tested at the first temperature; Acquire a second image of the steel to be inspected at a second workstation of the production line, where the steel to be inspected is at a second temperature at the second workstation of the production line, and the second temperature is higher than the first temperature; Using the preset feature extraction model, and generating a second shape feature set according to the second image, the second shape feature set is used to characterize the geometric features of the steel to be tested at the second temperature; Using a preset geometric feature comparison model, and determining a temperature adjustment instruction according to the first shape feature set and the second shape feature set, the temperature adjustment instruction is used to adjust the heating temperature of a heating station between the first station of the production line and the second station of the production line; The steel material to be detected is a steel beam structure, the first image and the second image are images of the steel beam structure at a first viewing angle, and the first viewing angle is a viewing direction perpendicular to a longitudinal section of the steel beam structure; Correspondingly, the using a preset feature extraction model and generating a first shape feature set according to the first image includes: Performing grayscale conversion processing on the first image to generate a first grayscale image; Performing edge detection on the first grayscale image using an edge detection model to determine a first edge set of the steel beam structure in the first grayscale image, wherein a contour formed by a plurality of edge lines included in the first edge set corresponds to an outer contour of the beam structure in the first grayscale image, and the first shape feature set includes the first edge set and a first edge endpoint set formed by endpoints of each edge line in the first edge set; Correspondingly, the using the preset feature extraction model and generating a second shape feature set according to the second image includes: Performing grayscale conversion processing on the second image to generate a second grayscale image; Performing edge detection on the second grayscale image using an edge detection model to determine a second edge set of the steel beam structure in the second grayscale image, wherein a contour formed by a plurality of edge lines included in the second edge set corresponds to an outer contour of the steel beam structure in the second grayscale image, and the second shape feature set includes the second edge set and a second edge endpoint set formed by endpoints of each edge line in the second edge set; The using a preset geometric feature comparison model and determining a temperature adjustment instruction according to the first shape feature set and the second shape feature set includes: Determine a first feature vector according to the first shape feature set , the first feature vector is a unit vector corresponding to the first vector, the first starting point of the first vector is the first feature point in the first shape feature set, the first ending point of the first vector is the second feature point in the first shape feature set, and the first vector is a vector from the first starting point to the first ending point; Determine a second feature vector according to the first shape feature set , the second feature vector is the unit vector corresponding to the second vector, the second starting point of the second vector is the third feature point in the second shape feature set, the second ending point of the second vector is the fourth feature point in the second shape feature set, the second vector is a vector pointing from the second starting point to the second ending point, the first feature point and the third feature point correspond to the same point on the steel beam structure, and the second feature point and the fourth feature point correspond to the same point on the steel beam structure; Using formula 1, and according to the first eigenvector And the second eigenvector Determine the first temperature adjustment amount , the formula 1 is: ; in, is the first calibration constant, is the second calibration constant; According to the first temperature adjustment amount The temperature adjustment instruction is generated.

2. The processing method for automobile steel production line according to claim 1, characterized in that: In determining a first feature vector according to the first shape feature set Previously, it also included: Determine a first feature vector set based on the first shape feature set ,in, The first feature vector set Middle feature vectors, is a feature vector formed by two edge endpoints in the first edge endpoint set; Using formula 2, and according to the first feature vector set Determine the first eigenvector , the formula 2 is: ; Correspondingly, in determining the second feature vector according to the first shape feature set Previously, it also included: Determine a second feature vector set based on the second shape feature set ,in, The second feature vector set Middle feature vectors, is a feature vector formed by two edge endpoints in the second edge endpoint set; According to the first feature vector And the second feature vector set Determine the second eigenvector , where the second eigenvector The corresponding The two edge endpoints of The two edge endpoints of the steel beam structure are corresponding edge endpoints.

3. The processing method for automobile steel production line according to claim 2, characterized in that: The method of using a preset geometric feature comparison model and determining a temperature adjustment instruction according to the first shape feature set and the second shape feature set further includes: The second temperature adjustment amount is determined by using formula 3 and the fully connected layer in the neural network model , the formula 3 is: ; in, is the activation function of the neural network model, is the first feature vector set, is the second eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, is the bias term in the fully connected layer; Correspondingly, the first temperature adjustment amount Generating the temperature adjustment instruction includes: Using Equation 4, and adjusting the amount according to the first temperature and the second temperature adjustment amount Determine the temperature adjustment amount , the formula 4 is: ; in, is the first weight, is the second weight, ,and, ; According to the temperature adjustment amount Determine to generate the temperature adjustment instruction.

4. The processing method for automobile steel production line according to claim 1 or 2, characterized in that: After using the preset feature extraction model and generating a second shape feature set according to the second image, the method further includes: Acquire a third image of the steel to be inspected at a third workstation of the production line, where the steel to be inspected is at a third temperature at the third workstation of the production line; Using the preset feature extraction model, and generating a third shape feature set according to the third image, the third shape feature set is used to characterize the geometric features of the steel to be tested at the third temperature, the third temperature being lower than the second temperature; Correspondingly, the using of the preset geometric feature comparison model and determining the temperature adjustment instruction according to the first shape feature set and the second shape feature set includes: The preset geometric feature comparison model is used, and the temperature adjustment instruction is determined according to the first shape feature set, the second shape feature set, and the third shape feature set.

5. The processing method for automobile steel production line according to claim 4, characterized in that: The using a preset geometric feature comparison model and determining the temperature adjustment instruction according to the first shape feature set, the second shape feature set, and the third shape feature set includes: Performing grayscale conversion processing on the third image to generate a third grayscale image; Performing edge detection on the third grayscale image using an edge detection model to determine a third edge set of the steel beam structure in the third grayscale image, wherein a contour formed by a plurality of edge lines included in the third edge set corresponds to an outer contour of the beam structure in the third grayscale image, and the third shape feature set includes the third edge set and a third edge endpoint set formed by endpoints of each edge line in the third edge set; Determine a third feature vector set according to the third shape feature set ,in, is the third eigenvector set Middle feature vectors, is a feature vector formed by two edge endpoints in the third edge endpoint set; The third temperature adjustment amount is determined by using Formula 5 and the fully connected layer in the neural network model , the formula 5 is: ; in, is the activation function of the neural network model, is the first feature vector set, is the third eigenvector set, is the vector comparison function in the fully connected layer, is the weight matrix in the fully connected layer, is the bias term in the fully connected layer; Using Equation 6, and adjusting the amount according to the first temperature And the third temperature adjustment amount Determine the temperature adjustment amount , the formula 6 is: ; in, is the third weight, is the fourth weight; According to the temperature adjustment amount Determine to generate the temperature adjustment instruction.

6. A control system for an automobile steel production line, characterized in that: include: An acquisition module, used to acquire a first image of a steel material to be inspected at a first station of a production line, wherein the steel material to be inspected is at a first temperature at the first station of the production line; A processing module, used for generating a first shape feature set according to the first image by using a preset feature extraction model, wherein the first shape feature set is used for characterizing geometric features of the steel material to be tested at the first temperature; The acquisition module is further used to acquire a second image of the steel to be inspected at a second workstation of the production line, where the steel to be inspected is at a second temperature at the second workstation of the production line, and the second temperature is higher than the first temperature; The processing module is further used to use the preset feature extraction model to generate a second shape feature set according to the second image, wherein the second shape feature set is used to characterize the geometric features of the steel to be tested at the second temperature; The processing module is further used to use a preset geometric feature comparison model and determine a temperature adjustment instruction according to the first shape feature set and the second shape feature set, wherein the temperature adjustment instruction is used to adjust a heating temperature of a heating station between the first station of the production line and the second station of the production line; The steel material to be detected is a steel beam structure, the first image and the second image are images of the steel beam structure at a first viewing angle, and the first viewing angle is a viewing direction perpendicular to a longitudinal section of the steel beam structure; the processing module is specifically used to: Performing grayscale conversion processing on the first image to generate a first grayscale image; Performing edge detection on the first grayscale image using an edge detection model to determine a first edge set of the steel beam structure in the first grayscale image, wherein a contour formed by a plurality of edge lines included in the first edge set corresponds to an outer contour of the beam structure in the first grayscale image, and the first shape feature set includes the first edge set and a first edge endpoint set formed by endpoints of each edge line in the first edge set; Performing grayscale conversion processing on the second image to generate a second grayscale image; Performing edge detection on the second grayscale image using an edge detection model to determine a second edge set of the steel beam structure in the second grayscale image, wherein a contour formed by a plurality of edge lines included in the second edge set corresponds to an outer contour of the steel beam structure in the second grayscale image, and the second shape feature set includes the second edge set and a second edge endpoint set formed by endpoints of each edge line in the second edge set; Optionally, the processing module is specifically used to: Determine a first feature vector according to the first shape feature set , the first feature vector is a unit vector corresponding to the first vector, the first starting point of the first vector is the first feature point in the first shape feature set, the first ending point of the first vector is the second feature point in the first shape feature set, and the first vector is a vector from the first starting point to the first ending point; Determine a second feature vector according to the first shape feature set , the second feature vector is the unit vector corresponding to the second vector, the second starting point of the second vector is the third feature point in the second shape feature set, the second ending point of the second vector is the fourth feature point in the second shape feature set, the second vector is a vector pointing from the second starting point to the second ending point, the first feature point and the third feature point correspond to the same point on the steel beam structure, and the second feature point and the fourth feature point correspond to the same point on the steel beam structure; Using formula 1, and according to the first eigenvector And the second eigenvector Determine the first temperature adjustment amount , the formula 1 is: ; in, is the first calibration constant, is the second calibration constant; According to the first temperature adjustment amount The temperature adjustment instruction is generated.

7. An electronic device, characterized in that: include: processor; as well as, A memory, configured to store executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 5 by executing the executable instructions.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 5 when executed by a processor.