Cold working process parameter optimization and plate rolling control method and system based on artificial intelligence

Abstract

The invention belongs to the technical field of industrial control methods and systems, particularly discloses a cold working process parameter optimization and plate rolling control method based on artificial intelligence, and relates to the technical field of cold working forming and intelligent control. The method comprises the steps that historical process parameters and corresponding plate quality data are collected, an evaluation function is constructed, and a process weight value is output; optimal parameters are screened according to the process weight and imported into cold working equipment; the curvature distribution of the plate in the plate rolling process is monitored in real time, and the plate rolling state is recognized; and executing a maintenance strategy or a correction strategy according to the state, and fusing correction feedback to optimize subsequent parameters. According to the method, the plate rolling precision and the forming consistency can be remarkably improved, the abnormal state self-adaptive correction capacity is achieved, dependence on manual experience is reduced, and the method is suitable for various plate rolling devices and plate types and has good universality and popularization value.

Description

Technical Field

[0001] The present invention belongs to the technical field of industrial control methods and systems, and specifically relates to a cold working process parameter optimization and plate rolling control method and system based on artificial intelligence. Background Art

[0002] Cold working is a key process in sheet metal processing and is widely used in automotive, home appliance, building structure, aerospace, and mechanical equipment. In particular, in processes such as cold bending, cold rolling, and coiling, the proper setting of process parameters is directly related to the product's forming quality, dimensional accuracy, and material utilization. However, traditional cold working processes, which mostly rely on empirical rules or process settings based on physical models, often suffer from low parameter adjustment efficiency, poor product consistency, and high process uncontrollability in actual production.

[0003] Taking the sheet metal rolling process as an example, its control process involves multiple variables, such as sheet thickness, yield strength, initial stress state, bend radius, coiling speed, and coiling tension. These variables are strongly nonlinear and multi-coupled. Traditional control methods typically use linear models, lookup tables, or simple control strategies. These methods struggle to dynamically respond to complex working conditions and the individual differences between batches of material. This results in uneven product curling, severe springback, and large processing errors, impacting product quality and subsequent assembly.

[0004] In recent years, artificial intelligence technologies, particularly machine learning, neural networks, and reinforcement learning, have been increasingly widely applied in industrial manufacturing. These technologies can establish a mapping relationship between input parameters and processing results through training on large-scale production data. They possess self-learning, self-adaptation, and self-optimization capabilities, providing a new technical approach to solving parameter optimization and control issues in cold working processes. For example, by introducing deep learning models, data samples from previous sheet metal rolling processes can be analyzed to uncover the influence of key parameters on forming quality. Combined with reinforcement learning algorithms, these models can also enable continuous iteration and optimization of control strategies, thereby improving process stability and efficiency.

[0005] However, the current application of artificial intelligence in the field of cold working is still in its early stages and faces the following problems: 1. The lack of an intelligent parameter optimization framework that is highly coupled with the cold working process makes it difficult to achieve real-time decision-making and precise control; 2. Most existing intelligent models are black box models and lack physical interpretability, which affects their credibility and scalability in industrial production; 3. Some systems only target a single piece of equipment or process parameters and lack the ability to fully control the entire cold working production chain.

[0006] Therefore, there is an urgent need to develop an AI-based cold working process parameter optimization and sheet rolling control method and system. By building a multi-dimensional data perception and real-time analysis mechanism, integrating data-driven models with physical mechanism models, this system can achieve intelligent prediction, adaptive parameter adjustment, and closed-loop control of the sheet rolling process under different working conditions, ensuring sheet forming quality and process stability. This system not only improves the automation and intelligence level of cold working processing, but also has significant economic benefits and industrial promotion value, and is an important direction for promoting the upgrading of intelligent manufacturing. Summary of the Invention

[0007] The purpose of the present invention is to provide a cold working process parameter optimization and rolling control method and system based on artificial intelligence, aiming to accurately optimize and control the cold working process parameters and the rolling process through intelligent means, which not only improves production efficiency and product quality, but also enhances process stability, enables real-time monitoring and feedback adjustment, and effectively solves the problems caused by traditional manual control methods.

[0008] The technical solution adopted by the present invention is as follows: a cold working process parameter optimization and plate rolling control method based on artificial intelligence, comprising the following steps: S1. Obtain historical process data: collect the actual process parameters of the cold working equipment in multiple past process batches and the corresponding plate quality data; S2. Constructing a process evaluation function: Taking process parameters and plate quality as input variables, construct an evaluation function to quantify the execution quality of each set of process parameters; S3. Generate process weight value: Output the execution quality of each set of process parameters as a process score, and determine the corresponding process weight value by sorting; S4. Screening and optimizing process parameters: Optimizing and screening the process parameter set according to all process weight values, and importing the screening results into the parameter control module; S5. Real-time monitoring of curvature distribution: During the plate rolling process, the plate edge image data is collected and the curvature distribution curve is identified through image processing and curvature calculation modules; S6. Identify the plate rolling state: compare the curvature distribution with the preset evaluation interval to identify the current plate rolling state, including normal state, over-bending state, and smooth transition state; S7. Output control strategy: Output control strategy based on the recognition result, wherein the normal state executes the maintenance strategy and the abnormal state triggers the correction strategy; S8, fusion correction feedback: collect edge data of the plate before and after correction in abnormal state, calculate the edge change curvature distribution value, and integrate it into the current optimized process parameters as a correction factor.

[0009] Furthermore, the process evaluation function is a multidimensional regression model, and its model weights are obtained by training dimensions such as quality grade, surface flatness, edge smoothness, and rolling deviation in historical process samples. Linear regression or neural network algorithm is used to achieve scoring output, and the output results are used to perform performance ranking and priority assignment on all parameter groups.

[0010] Furthermore, the step of screening the optimized process parameters includes: S31. Perform high-dimensional feature encoding on all process parameters according to process weight values; S32, using the encoding result as a query condition, searching for the corresponding parameter template in the optimization database; S33. Select the parameter template with the highest matching degree with the current plate feature through similarity calculation as the optimized process parameter.

[0011] Furthermore, the process of real-time monitoring of curvature distribution includes: using a high-resolution image sensor to obtain a real-time image of the edge of the plate during the rolling process, extracting the edge line through an edge detection algorithm, and using a sliding window method to obtain continuous edge coordinate points, calculating the curvature distribution map based on the angle change between discrete points, and inputting the result into a state recognition module for state judgment.

[0012] Furthermore, the correction strategy includes: jointly fine-tuning parameters such as the rolling force, rolling speed, and roller spacing position according to the current state type. Specifically, when the plate is in an excessively bent state, the rolling force is reduced and the feed speed is reduced; when the plate is in a smooth transition state, the roller pressure is appropriately increased and the rolling radius is reduced, while maintaining balanced pressure at the edge.

[0013] Furthermore, the correction feedback fusion step includes: S61, collecting edge data before correction in the abnormal state of plate rolling, and outputting its curvature distribution value; S62, after executing the correction strategy, collect the corrected edge data again and obtain the corresponding curvature distribution value; S63, comparing and analyzing the curvature distribution results before and after correction, and calculating the curvature change Δκ value; S64, inputting the Δκ value as a correction factor into the correction function model to generate a dynamic fine-tuning parameter; S65, updating and integrating the fine-tuning parameters into the current process parameter group to form updated optimized parameters for subsequent use; S66. Establish an optimized process parameter library, version all parameter groups that are fed back as "normal", and build a parameter library containing process number, execution time, plate information, final curvature results, etc., for subsequent process migration and rapid call, to achieve rapid optimization recommendations for the same type of plates.

[0014] Furthermore, each group of parameters in the optimized process parameter library is provided with a dynamic validity period label. The system regularly reviews and evaluates the usage frequency and stability index of the parameter group. When a group of parameters has a low usage frequency or an abnormal increase in feedback within a specified time, it is marked as a historical invalid group and removed from the default recommendation list.

[0015] The present application also provides an artificial intelligence-based cold working process parameter optimization and plate rolling control system, which is used to execute the artificial intelligence-based cold working process parameter optimization and plate rolling control method, including: Process data acquisition module, used to collect historical parameters and plate quality results and provide them to the evaluation model; Parameter evaluation module, used to construct evaluation functions and output process scores and process weight values; Process optimization module, used to screen optimal parameters and form control instructions; The curvature monitoring module is used to collect images, calculate curvature and determine the rolling status during the rolling process; A strategy control module, used to determine whether to execute a maintenance strategy or a correction strategy based on the rolling state; Correction fusion module, used to collect edge data before and after correction and output correction factors; The process parameter library management module is used to record, manage and eliminate optimization parameter groups.

[0016] The present invention also provides an electronic device comprising: at least one processor, and a memory, wherein the memory stores a computer program. When the program is executed by the processor, the processor executes the artificial intelligence-based cold working process parameter optimization and plate rolling control method, thereby realizing intelligent optimization, automatic control and correction feedback functions of the cold working process parameters.

[0017] Beneficial effects Compared with the existing technology, the artificial intelligence-based cold working process parameter optimization and plate rolling control method provided by the present invention has the following beneficial effects: 1. This invention uses an artificial intelligence model to evaluate and optimize historical process parameters, automatically screening out the optimal rolling parameter combination, effectively improving rolling accuracy and sheet forming consistency.

[0018] 2. Utilize real-time curvature monitoring and status recognition mechanisms to achieve rapid response and adaptive correction to abnormal coiling conditions, significantly reducing defective product rates and material waste.

[0019] 3. Build a process parameter knowledge base to achieve long-term accumulation and reuse of process data, support rapid adaptation of different plate types, improve production efficiency, and reduce reliance on manual experience.

[0020] In summary, the present invention deeply integrates artificial intelligence technology into the cold working process parameter optimization and rolling control process, which not only effectively improves the system intelligence level and rolling accuracy, but also greatly reduces the labor intensity and manufacturing costs of workers. It has broad application prospects and significant industrial promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Schematic diagram of the method of the present invention. DETAILED DESCRIPTION

[0022] In order to facilitate understanding of the present invention, the present invention is described in more detail below with reference to specific embodiments. It should be understood that the following description is only for illustrating specific embodiments of the present invention and is not intended to limit the scope of protection of the present invention.

[0023] Next, specific embodiments of the present invention will be described with reference to the accompanying drawings.

[0024] Example 1 like Figure 1 As shown, this embodiment provides a cold working process parameter optimization and plate rolling control method based on artificial intelligence, comprising the following steps: S1. Obtain historical process data: collect the actual process parameters of the cold working equipment in multiple past process batches and the corresponding plate quality data; S2. Constructing a process evaluation function: Taking process parameters and plate quality as input variables, construct an evaluation function to quantify the execution quality of each set of process parameters; S3. Generate process weight value: Output the execution quality of each set of process parameters as a process score, and determine the corresponding process weight value by sorting; S4. Screening and optimizing process parameters: Optimizing and screening the process parameter set according to all process weight values, and importing the screening results into the parameter control module; S5. Real-time monitoring of curvature distribution: During the plate rolling process, the plate edge image data is collected and the curvature distribution curve is identified through image processing and curvature calculation modules; S6. Identify the plate rolling state: compare the curvature distribution with the preset evaluation interval to identify the current plate rolling state, including normal state, over-bending state, and smooth transition state; S7. Output control strategy: Output control strategy based on the recognition result, wherein the normal state executes the maintenance strategy and the abnormal state triggers the correction strategy; S8, fusion correction feedback: collect edge data of the plate before and after correction in abnormal state, calculate the edge change curvature distribution value, and integrate it into the current optimized process parameters as a correction factor.

[0025] Example 2 This embodiment is further improved on the basis of embodiment 1: The process evaluation function is a multidimensional regression model, and its model weights are obtained by training dimensions such as quality grade, surface flatness, edge smoothness, and rolling deviation in historical process samples. Linear regression or neural network algorithm is used to achieve scoring output, and the output results are used to perform performance ranking and priority assignment for all parameter groups.

[0026] The screening step of optimizing process parameters includes: S31. Perform high-dimensional feature encoding on all process parameters according to process weight values; S32, using the encoding result as a query condition, searching for the corresponding parameter template in the optimization database; S33. Select the parameter template with the highest matching degree with the current plate feature through similarity calculation as the optimized process parameter.

[0027] The process of real-time monitoring of curvature distribution includes: using a high-resolution image sensor to obtain a real-time image of the edge of the plate during the rolling process, extracting the edge line through an edge detection algorithm, and using a sliding window method to obtain continuous edge coordinate points, calculating the curvature distribution map based on the angle change between discrete points, and inputting the result into a state recognition module for state judgment.

[0028] The correction strategy includes: jointly fine-tuning parameters such as the rolling force, rolling speed, and roller spacing position according to the current state type. Specifically, when the plate is in an excessively bent state, the rolling force is reduced and the feed speed is reduced; when the plate is in a smooth transition state, the roller pressure is appropriately increased and the rolling radius is reduced, while maintaining balanced pressure at the edge.

[0029] Example 3 This embodiment is further improved on the basis of embodiment 1: The correction feedback fusion step includes: S61, collecting edge data before correction in the abnormal state of plate rolling, and outputting its curvature distribution value; S62, after executing the correction strategy, collect the corrected edge data again and obtain the corresponding curvature distribution value; S63, comparing and analyzing the curvature distribution results before and after correction, and calculating the curvature change Δκ value; S64, inputting the Δκ value as a correction factor into the correction function model to generate a dynamic fine-tuning parameter; S65, updating and integrating the fine-tuning parameters into the current process parameter group to form updated optimized parameters for subsequent use; S66. Establish an optimized process parameter library, version all parameter groups that are fed back as "normal", and build a parameter library containing process number, execution time, plate information, final curvature results, etc., for subsequent process migration and rapid call, to achieve rapid optimization recommendations for the same type of plates.

[0030] Each group of parameters in the optimized process parameter library is provided with a dynamic validity period label. The system regularly reviews and evaluates the usage frequency and stability index of the parameter group. When a group of parameters has a low usage frequency or an abnormal increase in feedback within a specified time, it is marked as a historical invalid group and removed from the default recommendation list.

[0031] Example 4 This embodiment provides an artificial intelligence-based cold working process parameter optimization and plate rolling control method, which is suitable for four-roller hydraulic plate rolling machine equipment, especially for batch production of high-precision fan barrel-type plate parts.

[0032] First, execute step S1 to obtain historical process data. This example collects data from cold-rolling projects completed by a factory within the past year, extracting the actual operating parameters of each project, including 12 variables such as rolling force, feed speed, roll gap setting value, and upper roll initial position. Quality data for each batch of finished plates, such as curvature error, edge smoothness, and residual stress distribution, is also collected. Ultimately, a structured process data set with a sample size of no less than 500 groups is formed.

[0033] In step S2, a process evaluation function is constructed. This embodiment employs a multidimensional regression model based on a fusion of random forests and neural networks, using the aforementioned historical data as training samples, process parameters as input variables, and a plate quality score as output. The model is iteratively trained using cross-validation, ultimately yielding a scoring function that can output a corresponding score for any set of process parameters, reflecting the average execution quality of that parameter set in historical applications.

[0034] Execute step S3 to generate process weight values. After normalizing the scoring values ​​output by the model, a parameter priority list is established through sorting, and each group of parameters is assigned a unique weight value to achieve quantitative grading of parameter groups.

[0035] Proceed to step S4 to select and optimize process parameters. The system first performs high-dimensional encoding on all weight values ​​and constructs an index number using hash mapping. It then uses process conditions such as plate type (e.g., Q235, 16Mn), thickness, and target coil radius as search criteria to query the optimized process database and select the parameter group that best matches the current task as the initial process parameters for the current task.

[0036] In step S5, the curvature distribution of the plate during the rolling process is monitored in real time. This embodiment uses a high-resolution, bidirectional line-scanning industrial camera installed at the discharge end of the rolling equipment. An edge detection algorithm is used to extract the plate edge profile. A sliding window is used to extract the coordinates of discrete points every 5 mm. The local curvature is calculated using a three-point fitting method, ultimately generating a complete plate edge curvature distribution map. This graphical data is transmitted in real time to the control system for status recognition.

[0037] In step S6, the plate roll state is identified. The system compares the curvature distribution map with the preset curvature standard range. If the curvature fluctuation value is within the set range, it is identified as "normal state". If there is a significant high curvature concentration area, it is identified as "excessive curvature state". If the curvature of the continuous section is flat and exceeds the standard, it is identified as "transitional smooth state".

[0038] In step S7, a control strategy is output based on the identification results. In normal conditions, the system maintains the current process parameters; if an abnormal condition is identified, a correction strategy is triggered. For example, if excessive bending is identified, the system automatically reduces the upper roller feed pressure by 5% and reduces the rolling speed by 20%. If a smooth transition is identified, the system automatically tightens the roller gap by 0.8mm and appropriately increases the feed rate to enhance the bending force.

[0039] Step S8 integrates the correction feedback. The system collects two consecutive roll curvature distribution maps before and after the correction, compares and analyzes the difference, and calculates the edge curvature change Δκ(x). Δκ(x) is input into the correction function module, which outputs the roll speed correction factor and pressure correction gain. These are then integrated with the current parameters to form the next round of roll parameters.

[0040] To enhance process reusability, this example further establishes an optimized process parameter library, recording all parameter groups with "normal" feedback, including their process number, execution time, plate material and specifications, final curvature mean and variance, and other indicators. Each parameter group is labeled with a dynamic expiration date, and the system regularly evaluates its call frequency, feedback accuracy, and other factors. If a parameter group receives three consecutive abnormal feedbacks or has not been used within three months, it is marked as a "historically invalid group" and automatically removed from the recommended list.

[0041] The method described in this embodiment has been successfully applied to the cold rolling process of multiple batches of fan cylinders with different specifications. It has been verified that it has good automatic adjustment capabilities, parameter migration capabilities and a significant effect of improving rolling accuracy. The product qualification rate has been increased from the original 92.4% to 98.7%, and the dependence on operators has been significantly reduced.

[0042] Example 5 This embodiment provides an artificial intelligence-based cold working process parameter optimization and plate rolling control system, which is used to execute the artificial intelligence-based cold working process parameter optimization and plate rolling control method, including: Process data acquisition module, used to collect historical parameters and plate quality results and provide them to the evaluation model; Parameter evaluation module, used to construct evaluation functions and output process scores and process weight values; Process optimization module, used to screen optimal parameters and form control instructions; The curvature monitoring module is used to collect images, calculate curvature and determine the rolling status during the rolling process; A strategy control module, used to determine whether to execute a maintenance strategy or a correction strategy based on the rolling state; Correction fusion module, used to collect edge data before and after correction and output correction factors; The process parameter library management module is used to record, manage and eliminate optimization parameter groups.

[0043] Example 6 This embodiment provides an artificial intelligence-based cold working process parameter optimization and plate rolling control system, which is deployed on a set of four-roller hydraulic plate rolling machine equipment and can realize intelligent parameter optimization, status monitoring, strategy control and process knowledge management of the entire cold working plate rolling process.

[0044] The system includes the following modules: 1. Process data acquisition module This module connects the equipment's historical data interface with the quality inspection system database. It collects historical process parameters (such as plate thickness, width, rolling force, rolling speed, and roll gap settings) for multiple batches of plate rolling tasks, along with corresponding plate quality results (such as curvature error, edge flatness, and thickness consistency). It then structures and processes this data, outputting it to the subsequent evaluation model. In actual operation, this module synchronizes data every 24 hours, acquiring at least 100 sets of valid historical samples.

[0045] 2. Parameter Evaluation Module This module incorporates a neural network regression model as a process evaluation function, trained on pairs of process parameters and quality scores. The model is trained before deployment and can dynamically adjust after deployment by receiving new data in real time. The module outputs execution scores based on the input process parameters and generates process weights based on the ranking of scores, which guide parameter selection in the optimization module.

[0046] 3. Process optimization module This module sorts and hashes all optional process parameter groups based on the process weights generated by the evaluation module. It then matches the results with standard parameter templates in the process optimization database. This matching process calculates similarity and optimal matching based on parameters such as the current sheet material type, thickness, and target roll radius. Ultimately, it generates structured control instructions and sends them to the equipment control unit. This module supports online parameter fine-tuning and offers high-speed response capabilities.

[0047] 4. Curvature monitoring module A high-resolution industrial camera and edge recognition sensor are installed at the equipment's exit. Combined with image processing algorithms, this module captures real-time images of the plate edge during the rolling process, extracts edge lines, and calculates the discrete curvature distribution. This module utilizes a three-point method and curvature fitting technology, achieving an error control of ±0.5%. The system compares the identified curvature distribution against pre-set evaluation intervals to determine the current rolling status.

[0048] 5. Policy Control Module This module calls the control strategy mapping table based on the status output by the curvature monitoring module (including normal, excessive curvature, and smooth transition). If the module identifies a normal state, it maintains the current control instructions. If an abnormal state is identified, it immediately triggers the correction logic to adjust parameters such as rolling speed, pressure, and roll gap. The strategy control module supports real-time communication with the PLC system, with adjustment delays of no more than 200ms.

[0049] 6. Modify the fusion module After executing the correction command, this module uses the image sensor to recollect plate edge data, obtaining the curvature distribution of the corresponding area before and after correction, and calculating the change Δκ. This change is input into the correction function model, which outputs a correction factor for dynamic adjustment of optimization parameters. The module also supports recording and trend analysis of multiple correction actions, supporting parameter fine-tuning.

[0050] 7. Process parameter library management module This module maintains a structured database of optimized process parameters. Each record contains information such as parameter group ID, applicable plate type, execution time, usage frequency, and curvature feedback. This module implements a dynamic validity determination mechanism: if a parameter group continuously displays abnormal feedback or is called too infrequently within a set period, it is marked as historically invalid and removed from the recommended list. Furthermore, the system supports exporting recommended parameter combinations for rapid initialization of new tasks.

[0051] By deploying the system described in this embodiment, the traditional cold rolling process can be upgraded from "empirical" to "data-driven" intelligent control, significantly improving parameter selection efficiency, rolling accuracy, and equipment automation. System measurements have shown that using the system has increased the first-article pass rate from 92% to over 98%, reduced the frequency of manual intervention by over 60%, and significantly enhanced the system's ability to adapt to a wide range of plate specifications, demonstrating promising prospects for widespread application.

[0052] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A cold working process parameter optimization and plate rolling control method based on artificial intelligence, characterized in that: The following steps are involved: S1. Obtain historical process data: collect the actual process parameters of the cold working equipment in multiple past process batches and the corresponding plate quality data; S2. Constructing a process evaluation function: Taking process parameters and plate quality as input variables, construct an evaluation function to quantify the execution quality of each set of process parameters; S3. Generate process weight value: Output the execution quality of each set of process parameters as a process score, and determine the corresponding process weight value by sorting; S4. Screening and optimizing process parameters: Optimizing and screening the process parameter set according to all process weight values, and importing the screening results into the parameter control module; S5. Real-time monitoring of curvature distribution: During the plate rolling process, the plate edge image data is collected and the curvature distribution curve is identified through image processing and curvature calculation modules; S6. Identify the plate rolling state: compare the curvature distribution with the preset evaluation interval to identify the current plate rolling state, including normal state, over-bending state, and smooth transition state; S7. Output control strategy: Output control strategy based on the recognition result, wherein the normal state executes the maintenance strategy and the abnormal state triggers the correction strategy; S8, fusion correction feedback: collect edge data of the plate before and after correction in abnormal state, calculate the edge change curvature distribution value, and integrate it into the current optimized process parameters as a correction factor.

2. The method for cold working process parameter optimization and plate rolling control based on artificial intelligence according to claim 1, characterized in that: The process evaluation function is a multidimensional regression model, and its model weights are obtained by training dimensions such as quality grade, surface flatness, edge smoothness, and rolling deviation in historical process samples. Linear regression or neural network algorithm is used to achieve scoring output, and the output results are used to perform performance ranking and priority assignment for all parameter groups.

3. The method for cold working process parameter optimization and plate rolling control based on artificial intelligence according to claim 1, characterized in that: The screening step of optimizing process parameters includes: S31. Perform high-dimensional feature encoding on all process parameters according to process weight values; S32, using the encoding result as a query condition, searching for the corresponding parameter template in the optimization database; S33. Select the parameter template with the highest matching degree with the current plate feature as the optimized process parameter through similarity calculation.

4. The method for cold working process parameter optimization and plate rolling control based on artificial intelligence according to claim 1, characterized in that: The process of real-time monitoring of curvature distribution includes: using a high-resolution image sensor to obtain a real-time image of the edge of the plate during the rolling process, extracting the edge line through an edge detection algorithm, and using a sliding window method to obtain continuous edge coordinate points, calculating the curvature distribution map based on the angle change between discrete points, and inputting the result into a state recognition module for state judgment.

5. The method for cold working process parameter optimization and plate rolling control based on artificial intelligence according to claim 1, characterized in that: The correction strategy includes: jointly fine-tuning parameters such as the rolling force, rolling speed, and roller spacing position according to the current state type. Specifically, when the plate is in an excessively bent state, the rolling force is reduced and the feed speed is reduced; when the plate is in a smooth transition state, the roller pressure is appropriately increased and the rolling radius is reduced, while maintaining balanced pressure at the edge.

6. The method for cold working process parameter optimization and plate rolling control based on artificial intelligence according to claim 1, characterized in that: The correction feedback fusion step includes: S61, collecting edge data before correction in the abnormal state of plate rolling, and outputting its curvature distribution value; S62, after executing the correction strategy, collect the corrected edge data again and obtain the corresponding curvature distribution value; S63, comparing and analyzing the curvature distribution results before and after correction, and calculating the curvature change Δκ value; S64, inputting the Δκ value as a correction factor into the correction function model to generate a dynamic fine-tuning parameter; S65 , integrating the fine-tuning parameters into the current process parameter group to form updated optimized parameters for subsequent use.

7. The method for cold working process parameter optimization and plate rolling control based on artificial intelligence according to claim 1, characterized in that: The modified feedback fusion step further comprises the following steps: S66. Establish an optimized process parameter library, version all parameter groups that are fed back as "normal", and build a parameter library containing process number, execution time, plate information, final curvature results, etc. for subsequent process migration and rapid call, to achieve rapid optimization recommendations for the same type of plates.

8. The method for cold working process parameter optimization and plate rolling control based on artificial intelligence according to claim 7, characterized in that: Each group of parameters in the optimized process parameter library is provided with a dynamic validity period label. The system regularly reviews and evaluates the usage frequency and stability index of the parameter group. When a group of parameters has a low usage frequency or an abnormal increase in feedback within a specified time, it is marked as a historical invalid group and removed from the default recommendation list.

9. An artificial intelligence-based cold working process parameter optimization and plate rolling control system, used to execute the artificial intelligence-based cold working process parameter optimization and plate rolling control method according to any one of claims 1 to 8, characterized in that: include: Process data acquisition module, used to collect historical parameters and plate quality results and provide them to the evaluation model; Parameter evaluation module, used to construct evaluation functions and output process scores and process weight values; Process optimization module, used to screen optimal parameters and form control instructions; The curvature monitoring module is used to collect images, calculate curvature and determine the rolling status during the rolling process; A strategy control module, used to determine whether to execute a maintenance strategy or a correction strategy based on the rolling state; Correction fusion module, used to collect edge data before and after correction and output correction factors; The process parameter library management module is used to record, manage and eliminate optimization parameter groups.

10. An electronic device, characterized in that: include: At least one processor and a memory, wherein the memory stores a computer program, and when the program is executed by the processor, the processor executes the method according to any one of claims 1 to 8 to achieve intelligent optimization, automatic control and correction feedback functions of cold working process parameters.

Images