Method and system for dynamically regulating and controlling technological parameters of aluminum alloy plate

Abstract

The invention belongs to the technical field of industrial control, and discloses an aluminum alloy plate process parameter dynamic regulation and control method and system, which combines an offline established model library with online collected real-time data by introducing model optimization in a test stage and a model online adjustment mechanism in a formal straightening stage. Therefore, dynamic and self-adaptive regulation and control of straightening process parameters of aluminum alloy plates of different types and specifications are achieved, the limitation caused by the fact that a traditional method depends on a fixed parameter table or an inaccurate model is overcome, and the straightening precision and the production efficiency are improved.

Description

Technical Field

[0001] The present application relates to the field of industrial control technology, and in particular to a method and system for dynamically controlling process parameters of aluminum alloy plates. Background Art

[0002] Rolling is a key process in continuous aluminum alloy sheet production lines. After rolling, the sheets typically exhibit various shape defects, such as bends, waves, and twists, and also generate residual stresses. To eliminate these defects and improve sheet quality, the sheets undergo a straightening process. A multi-roller straightener is a commonly used straightening machine. It repeatedly bends and deforms the sheet between multiple rows of staggered straightening rollers, eliminating these defects and reducing residual stresses. Straightening machine process parameters, such as the relative vertical position (i.e., reduction) between the upper and lower straightening rollers in each row and the straightening speed (sheet conveying speed), play a decisive role in the straightening effect. Precise control of these parameters allows for the regulation of sheet plastic deformation, ultimately resulting in a flat, low-stress sheet.

[0003] A single production line often needs to process aluminum alloy plates with varying alloy compositions, heat treatment states, and specifications (thickness, width). These plates exhibit fundamentally different plastic deformation behaviors and responses to straightening parameters (particularly the roll reduction combination) in a multi-roll straightening machine.

[0004] To achieve effective straightening of different plates, it is necessary to set corresponding straightening process parameters based on the type and specifications of the plates currently being produced. Traditional methods for setting straightening parameters rely mainly on two approaches: one is to pre-establish a fixed set of process parameter tables for each common alloy-specification combination. This method requires determining the optimal set of straightening parameters for each target plate combination through extensive offline experiments or complex numerical simulations based on precise physical models before actual production. This requires an extremely large investment of time, manpower, and computing resources, is costly and inefficient, and in practice cannot cover all potential production needs, especially for uncommon alloys or special specifications.

[0005] The second method is to calculate parameters based on simplified physical models. These models are usually constructed based on idealized assumptions and have large deviations from the actual situation. This will cause the preset model parameters to deviate from the actual situation, resulting in a discrepancy between the straightening effect predicted by the model and the actual plate shape, affecting the straightening accuracy.

[0006] When a production line needs to switch to a new alloy type or specification combination, there are often no ready-made accurate models or reliable process parameters to refer to. In this case, it is usually necessary to rely on experienced operators to perform debugging. However, the accuracy and efficiency of such debugging are highly dependent on the operator's level of experience, making it difficult to ensure consistency.

[0007] Therefore, there is an urgent need for an advanced method that can rapidly assess or identify the actual physical and mechanical properties of the plate as it passes through the straightener, based on information about the plate type and specifications, as well as real-time online inspection data. Based on this identification, the straightening process parameters can be quickly determined and dynamically adjusted to improve straightening accuracy and efficiency, reduce trial and error, and reduce scrap, to accommodate different plate types and production conditions.

[0008] In view of the above problems, the existing technology is in urgent need of improvement. Summary of the Invention

[0009] The purpose of this application is to provide a method and system for dynamically controlling the process parameters of aluminum alloy plates, which can quickly determine and dynamically adjust the straightening process parameters based on the basic information and real-time data of the plates, improve the straightening accuracy and efficiency, reduce trial and error and waste, and adapt to different plates and production conditions.

[0010] In a first aspect, the present application provides a method for dynamically controlling process parameters of an aluminum alloy plate, which is used to adjust the process parameters of an aluminum alloy plate when a multi-roll straightening machine is used to straighten the aluminum alloy plate. The method comprises the following steps: A1. Obtain basic information of the aluminum alloy plate, as well as real-time online detection data during the testing phase; during the testing phase, straighten the aluminum alloy plate using preset process parameters; A2. Based on the basic information, a set of candidate prediction models is determined from a preset prediction model library; each prediction model in the prediction model library is used to predict the deformation response of the aluminum alloy plate based on process parameters and real-time online detection data; A3. Based on the real-time online detection data of the test phase, the optimal prediction model with the best prediction accuracy for the actual deformation response of the aluminum alloy plate is determined among the candidate prediction models; A4. Performing formal straightening of the aluminum alloy plate using the optimal prediction model, and adjusting the optimal prediction model online based on real-time online detection data during formal straightening; A5. Based on the current real-time online detection data, use the optimal prediction model after online adjustment to adjust the process parameters online.

[0011] Preferably, the basic information includes alloy grade, specification, thickness, width, length and surface condition; The real-time online detection data includes input data and response data; the input data includes flatness data, thickness data and temperature data before straightening; the response data includes flatness data after straightening; The process parameters include a pressing amount and a straightening speed; the preset process parameters include a preset pressing amount and a preset straightening speed.

[0012] Preferably, step A1 includes: A101 obtain the basic information of the aluminum alloy plate; A102. Use the preset process parameters to test straightening of the aluminum alloy plate, and during the test straightening process, continuously collect the real-time online detection data to form a real-time online detection data sequence; A103. Calculate the deformation response characteristic parameters of the aluminum alloy plate within the current test length range based on the real-time online detection data sequence; A104. If the deformation response characteristic parameter does not meet the preset test end threshold condition, continue to collect the real-time online detection data and update the real-time online detection data sequence, and return to step A103; otherwise, end the test phase and use the current real-time online detection data sequence as the real-time online detection data of the test phase.

[0013] Preferably, step A2 includes: A201. Based on the basic information, navigate along a hierarchical index structure in the preset prediction model library to locate a subset of prediction models related to the basic information; the hierarchical index structure is pre-constructed based on multiple information dimensions of the basic information; A202. Based on the preset matching rules between the basic information and each prediction model in the prediction model subset, determine a group of prediction models whose matching degree with the basic information is higher than a preset matching degree threshold as the candidate prediction models.

[0014] Preferably, step A202 includes: Determining a basic matching degree between the basic information and each prediction model in the prediction model subset based on a preset matching rule between the basic information and each prediction model in the prediction model subset; Analyzing actual deformation response characteristics of the aluminum alloy plate based on real-time online detection data during the testing phase; For each prediction model in the prediction model subset, based on the actual deformation response characteristics, calculating a prediction matching degree of the prediction model to the actual deformation response; For each prediction model in the prediction model subset, the corresponding basic matching degree is combined with the corresponding predicted matching degree to obtain a comprehensive matching degree; A group of prediction models whose comprehensive matching degree is higher than a preset matching degree threshold is determined as the candidate prediction models.

[0015] Preferably, step A3 includes: A301. Based on the preset process parameters used in the test phase and the input data in the real-time online detection data of the test phase, using each of the candidate prediction models, predict the deformation response of the aluminum alloy plate; A302. Compare the deformation response predicted by each candidate prediction model with the actual deformation response reflected by the real-time online detection data of the test phase, and calculate the prediction error of each candidate prediction model; A303. Determine the candidate prediction model with the smallest prediction error as the optimal prediction model.

[0016] Preferably, step A4 includes: A401. After initiating the formal straightening phase, determine initial process parameters using the optimal prediction model based on the initially detected input data, and collect real-time online detection data during the formal straightening process to obtain the real-time online detection data for the formal straightening process; A402. Extracting the current moment's real-time online detection data from the real-time online detection data during the formal straightening to determine the actual deformation response of the aluminum alloy plate at the current moment; A403 based on the current moment of the process parameters and the current moment of the real-time online detection data input data, using the current optimal prediction model, predicting the deformation response of the aluminum alloy plate, to obtain the current moment of the predicted deformation response; A404 compares the actual deformation response at the current moment and the predicted deformation response at the current moment, and calculates the predicted deviation at the current moment; A405. Based on the current prediction deviation and the preset adjustment rules, calculate the adjustment amount of the internal parameters of the current optimal prediction model; A406. Update the internal parameters of the current optimal prediction model according to the adjustment amount to obtain the optimal prediction model after online adjustment.

[0017] Preferably, step A405 includes: Determine the current production stage based on the predicted deviation at the current moment and the real-time online detection data during the formal straightening; the production stage includes an initial transient stage and a steady-state stage; According to the determined production stage, a corresponding adjustment rule is selected from a preset adjustment rule library; the adjustment rule library includes at least one adjustment rule applicable to the initial transient stage and at least one adjustment rule applicable to the steady-state stage; Based on the prediction deviation at the current moment and the selected adjustment rule, a coordinated adjustment amount of multiple related parameters within the current optimal prediction model is calculated as the adjustment amount.

[0018] Preferably, step A5 includes: A501. According to the real-time online detection data and the preset straightening requirements at the current moment, determine the target deformation response; A502. Using the optimal prediction model after online adjustment, and under the input data conditions of the real-time online detection data at the current moment, solving for process parameters that can make the predicted deformation response of the optimal prediction model after online adjustment reach the target deformation response, and using these as target process parameters; A503. Adjust the process parameters according to the target process parameters.

[0019] In a second aspect, the present application provides a dynamic control system for aluminum alloy plate process parameters, which is used to adjust the process parameters of a multi-roll straightening machine when straightening an aluminum alloy plate. The system includes: An information acquisition module is used to obtain basic information of the aluminum alloy plate and real-time online detection data during the testing phase; during the testing phase, the aluminum alloy plate is straightened using preset process parameters; a candidate model screening module, configured to determine a set of candidate prediction models from a preset prediction model library based on the basic information; each prediction model in the prediction model library is configured to predict the deformation response of the aluminum alloy plate based on process parameters and real-time online detection data; An optimal model determination module is used to determine, from among the candidate prediction models, an optimal prediction model with the best prediction accuracy for the actual deformation response of the aluminum alloy plate based on the real-time online detection data during the test phase; A model adjustment module, configured to perform formal straightening on the aluminum alloy plate using the optimal prediction model, and to perform online adjustment on the optimal prediction model based on real-time online detection data during formal straightening; The process parameter adjustment module is used to adjust the process parameters online according to the real-time online detection data at the current moment and use the optimal prediction model after online adjustment.

[0020] Beneficial effects: The present application provides a method and system for dynamically controlling the process parameters of aluminum alloy plates. By introducing a model optimization mechanism in the testing phase and an online model adjustment mechanism in the formal straightening phase, the offline established model library is combined with the real-time data collected online, thereby realizing dynamic and adaptive control of the straightening process parameters of aluminum alloy plates of different types and specifications. This overcomes the limitations of traditional methods that rely on fixed parameter tables or inaccurate models, and improves straightening accuracy and production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Flowchart of the method for dynamically controlling process parameters of aluminum alloy plates provided in an embodiment of the present application.

[0022] Figure 2 This is a schematic diagram of the structure of the dynamic control system of aluminum alloy plate process parameters provided in an embodiment of the present application.

[0023] Explanation of reference numbers: 1. Information acquisition module; 2. Candidate model screening module; 3. Optimal model determination module; 4. Model adjustment module; 5. Process parameter adjustment module. DETAILED DESCRIPTION

[0024] The technical model in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. The components of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of this application.

[0025] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0026] refer to Figure 1 The present application proposes a method for dynamically controlling the process parameters of an aluminum alloy plate, which is used to adjust the process parameters of an aluminum alloy plate when a multi-roll straightening machine is used to straighten the aluminum alloy plate. The method comprises the following steps: A1. Obtain basic information of the aluminum alloy plate, as well as real-time online detection data during the testing phase; during the testing phase, straighten the aluminum alloy plate using preset process parameters; A2. Based on the basic information, a set of candidate prediction models is determined from a preset prediction model library; each prediction model in the prediction model library is used to predict the deformation response of the aluminum alloy plate based on process parameters and real-time online detection data; A3. Based on the real-time online detection data of the test phase, the optimal prediction model with the best prediction accuracy for the actual deformation response of the aluminum alloy plate is determined among the candidate prediction models; A4. Performing formal straightening of the aluminum alloy plate using the optimal prediction model, and adjusting the optimal prediction model online based on real-time online detection data during formal straightening; A5. Based on the current real-time online detection data, use the optimal prediction model after online adjustment to adjust the process parameters online.

[0027] Among them, basic information refers to the inherent properties of aluminum alloy plates, such as alloy grade, specification, thickness, width, length and surface condition, which can be obtained by manual input, reading from the production management system or automatic identification by sensors. Its main purpose is to provide a preliminary basis for subsequent model selection.

[0028] Among them, real-time online detection data refers to the data collected in real time during the straightening process to reflect the plate state and straightening effect, such as plate shape data, thickness data, temperature data before straightening (as input data) and plate shape data after straightening (as response data). It can be continuously collected using online plate shape meters, thickness gauges, infrared thermometers and other equipment. Its main purpose is to obtain data reflecting the actual deformation characteristics of the current plate and provide an objective basis for model evaluation, selection and adjustment.

[0029] Among them, process parameters refer to the set values ​​used to control the straightening process, including the amount of pressure reduction (the relative position of each roller) and the straightening speed (the plate conveying speed). They can be set and adjusted through the straightening machine control system. Their main purpose is to achieve plastic deformation by regulating the stress state of the plate.

[0030] Among them, the prediction model library refers to a pre-established collection of various different types of prediction models. Each prediction model is used to predict the deformation response of aluminum alloy plates based on process parameters and real-time online detection data. It can include models based on physical principles, data-driven machine learning models (such as neural networks, support vector machines), or a combination of the two. Its main purpose is to cover the deformation laws under different plate characteristics and straightening conditions.

[0031] Among them, the candidate prediction model refers to a group of prediction models that may be applicable to the current plate and are preliminarily screened from the prediction model library based on basic information. They are determined from the model library through indexing or matching rules based on basic information. Their main purpose is to narrow the scope of model evaluation and improve the efficiency of model selection.

[0032] Among them, the optimal prediction model refers to the model with the highest prediction accuracy for the actual deformation response of the aluminum alloy plate after verification with actual data among the candidate prediction models. It is determined by comparing the errors between the prediction results of each candidate model in the test phase and the actual detection data. Its main purpose is to select the model that can most accurately describe the current deformation behavior of the plate.

[0033] Among them, deformation response refers to the plate shape change or residual stress state of the aluminum alloy plate under the action of specific process parameters. It is usually measured by the plate shape data after straightening, etc. It is the core target that needs to be predicted and controlled in the straightening process.

[0034] Among them, online adjustment of the optimal prediction model refers to the dynamic correction of the internal parameters of the optimal prediction model according to the prediction deviation fed back by real-time detection data during the formal straightening process. It can be implemented by using an algorithm based on error back propagation, Kalman filtering or other adaptive control algorithms. Its main purpose is to enable the model to adapt to small fluctuations and changes in the production process and improve the real-time accuracy and robustness of the model.

[0035] Among them, online adjustment of process parameters refers to the use of the optimal prediction model after online adjustment to solve the process parameters for straightening aluminum alloy plates during the formal straightening process.

[0036] The core innovation of this application lies in that by introducing the model optimization in the testing phase and the model online adjustment mechanism in the formal straightening phase, the offline established model library is combined with the real-time data collected online, thereby realizing dynamic and adaptive control of the straightening process parameters of aluminum alloy plates of different types and specifications, overcoming the limitations of traditional methods that rely on fixed parameter tables or inaccurate models, and improving straightening accuracy and production efficiency.

[0037] Specifically, the method first obtains basic information about the aluminum alloy sheet to be straightened and then conducts a test straightening test using pre-set process parameters. During this process, real-time online test data is continuously collected. This data reflects the sheet's actual deformation characteristics under actual straightening conditions. Based on this basic information, the system initially selects a set of candidate prediction models from a pre-set prediction model library. These models are pre-built based on different sheet characteristics. Subsequently, the prediction accuracy of each candidate model is evaluated using the real-time online test data collected during the test phase, and the model with the highest prediction accuracy is selected as the optimal prediction model. During the actual straightening phase, the system uses this optimal prediction model to straighten the aluminum alloy sheet while continuously collecting real-time online test data. Crucially, based on the real-time data collected during the actual straightening process, the system adjusts the internal parameters of the optimal prediction model online to adapt it to actual changes in the production process. Finally, based on the current real-time online test data, the system uses the optimal prediction model, which has been adjusted online, to dynamically calculate and adjust the current process parameters to achieve precise straightening control. This entire process forms a data-driven, model-adaptive closed-loop control process.

[0038] Through the above scheme, the present application can dynamically select and adjust the prediction model based on the actual characteristics of the aluminum alloy plate and the real-time status of the production process, thereby achieving precise and adaptive control of the straightening process parameters. This significantly improves the straightening accuracy and adaptability for plates of different alloys and specifications, reduces the reliance on extensive offline experiments and empirical trial and error, shortens the format switching time, reduces scrap rate and production costs, and improves the flexibility and efficiency of the production line.

[0039] In some embodiments, the basic information includes alloy grade, specification, thickness, width, length and surface condition; The real-time online detection data includes input data and response data; the input data includes flatness data, thickness data and temperature data before straightening; the response data includes flatness data after straightening; The process parameters include a pressing amount and a straightening speed; the preset process parameters include a preset pressing amount and a preset straightening speed.

[0040] Among them, the alloy grade, specification, thickness, width, length, and surface condition are the inherent properties of aluminum alloy plates, which affect the plate's mechanical behavior and deformation response during the straightening process. Clarifying this basic information allows, in subsequent steps, to filter candidate models relevant to the current plate's characteristics from the prediction model library based on these properties, laying the foundation for determining the optimal model and performing predictions.

[0041] Pre-straightening shape data, thickness data, and temperature data are input variables that influence the straightening process and results. Pre-straightening shape represents the initial state to be corrected; thickness is a fundamental geometric parameter of the sheet; and temperature affects the material's mechanical properties. These input data are captured and used as inputs to the prediction model in subsequent steps, enabling the model to make predictions and adjustments based on the current sheet state and environmental conditions. Post-straightening shape data is the result of the straightening process, reflecting the sheet's deformation response and the straightening effect. These response data are captured and compared with the model's predictions to evaluate the model's prediction accuracy and determine the optimal model.

[0042] The reduction and straightening speed are the primary control parameters of a multi-roller straightening machine, determining the degree and rate of bending deformation in the sheet metal. The prediction model is used to predict the sheet shape response for a given reduction and straightening speed, or to determine the reduction and straightening speed that achieve the target sheet shape. By defining these process parameters, the method can directly apply to the control variables of the straightening machine.

[0043] This clear information and parameters provide an accurate data foundation for the establishment, selection, evaluation, and online adjustment of prediction models, enabling them to more accurately reflect the actual deformation response of aluminum alloy plates. This helps select the model with the best prediction accuracy for the actual deformation response and enables effective adjustment of the model based on actual data, thereby improving the accuracy of the final process parameter adjustment and overcoming the limitations of dynamic control effects caused by incomplete information.

[0044] In some embodiments, step A1 comprises: A101 obtain the basic information of the aluminum alloy plate; A102. Use the preset process parameters to test straightening of the aluminum alloy plate, and during the test straightening process, continuously collect the real-time online detection data to form a real-time online detection data sequence; A103. Calculate the deformation response characteristic parameters of the aluminum alloy plate within the current test length range based on the real-time online detection data sequence; A104. If the deformation response characteristic parameter does not meet the preset test end threshold condition, continue to collect the real-time online detection data and update the real-time online detection data sequence, and return to step A103; otherwise, end the test phase and use the current real-time online detection data sequence as the real-time online detection data of the test phase.

[0045] Deformation response characteristic parameters refer to key indicators extracted or calculated from the real-time online test data sequence that characterize the actual deformation behavior of the aluminum alloy plate under the currently preset process parameters. These parameters can be statistics, rates of change, or specific pattern recognition results. For example, the plate's waviness, residual stress level, and plate shape change trend can be calculated. These parameters abstract and refine the original data sequence, reflecting the plate's mechanical response characteristics in a more fundamental way. The preset test end threshold condition refers to a pre-set criterion for determining whether the test straightening phase can be terminated. This condition is typically set based on the deformation response characteristic parameters. For example, the test can be terminated when the mean waviness value stabilizes within a certain range or the estimated residual stress value falls below a certain threshold. This condition ensures that the test is terminated promptly after sufficient data reflecting the plate's stable deformation characteristics has been obtained. Determining the end of the test based on characteristic parameters and looping data collection refers to comparing the calculated deformation response characteristic parameters with the preset test end threshold condition. If the condition is not met, the system automatically controls the continuation of the test straightening and data collection, updating the data sequence, and then recalculating the characteristic parameters and making another judgment, forming a feedback loop. The test phase is terminated only when the characteristic parameters meet the termination conditions. This mechanism enables the test process to adaptively determine the test length based on the actual deformation response of the plate, avoiding the shortcomings that may be caused by fixed-length testing.

[0046] In this application, basic plate information is first acquired, providing a foundation for subsequent testing and model selection. Then, the plate is tested and straightened using preset process parameters. The key to the testing process is the continuous acquisition of real-time online inspection data and the sequential organization of this data into a data sequence. This continuous acquisition and sequence formation method enables the system to capture the dynamic deformation process of the plate within a certain length range as it passes through the straightener, rather than just the instantaneous state at a specific point. Next, based on this data sequence, the system calculates deformation response characteristic parameters that reflect the actual deformation characteristics of the plate within the current test length range. These characteristic parameters are an effective refinement of the original data sequence, providing a more fundamental characterization of the plate's behavior under the preset parameters. The system then enters a judgment loop based on data feedback. It compares the calculated deformation response characteristic parameters with preset test termination thresholds. If the characteristic parameters have not yet reached a preset stable or converged state, indicating that the currently acquired data may not fully reflect the plate's stable deformation characteristics, the system will control the continuation of test straightening and data acquisition, update the data sequence, and return to step A103 to recalculate the characteristic parameters and make another judgment. This cycle continues until the deformation response characteristic parameters meet the preset threshold conditions. Once the conditions are met, the system considers that sufficiently stable and representative data has been acquired, and the test phase ends. The current complete data sequence is used as the real-time online detection data of the test phase.

[0047] This adaptive test termination mechanism based on deformation response characteristic parameters means that the test process no longer relies on a fixed test length or time, but instead dynamically adjusts the test range based on the actual deformation response of the plate. This ensures that the test time is shortened as much as possible while obtaining valid and stable data, thereby improving test efficiency. At the same time, the real-time online detection data sequence obtained, including the dynamic process and steady state, and the deformation response characteristic parameters extracted from it, can more accurately characterize the actual behavior of the current plate under the preset process parameters. These high-quality data serve as input for subsequent steps, especially for model selection, optimal model determination, and subsequent online model adjustment, significantly improving the accuracy and robustness of the entire process parameter dynamic control method.

[0048] In some embodiments, step A2 comprises: A201. Based on the basic information, navigate along a hierarchical index structure in the preset prediction model library to locate a subset of prediction models related to the basic information; the hierarchical index structure is pre-constructed based on multiple information dimensions of the basic information; A202. Based on the preset matching rules between the basic information and each prediction model in the prediction model subset, determine a group of prediction models whose matching degree with the basic information is higher than a preset matching degree threshold as the candidate prediction models.

[0049] The hierarchical index structure is a tree-like or hierarchically organized data structure, whose nodes and levels are pre-built based on different dimensions of basic information (such as alloy grade, thickness range, width range, etc.). Navigation involves searching down the hierarchical index structure layer by layer based on the specific basic information of the aluminum alloy plate to be straightened, until the leaf node or node set corresponding to this basic information is located. The prediction model subset refers to the set of prediction models located by navigating along the hierarchical index structure. The models in this subset have a high correlation with the basic information of the current aluminum alloy plate.

[0050] The preset matching rules are criteria or algorithms used to evaluate the correlation between the basic information of the aluminum alloy plate and each prediction model in the prediction model subset. The matching rules can be formulated based on factors related to the basic information, such as the data characteristics used during model training, the model type, and the model complexity.

[0051] The matching degree is a value calculated based on preset matching rules, reflecting the correlation or applicability between the basic information of the current aluminum alloy plate and a prediction model. The higher the matching degree, the more relevant the model is to the current plate.

[0052] The preset matching threshold is a pre-set numerical limit. Only when the calculated matching degree exceeds this threshold will the corresponding prediction model be considered sufficiently relevant to the current plate and thus be selected into the candidate prediction model set.

[0053] Specifically, this solution provides a method for identifying candidate prediction models from a pre-set prediction model library. First, basic information about the aluminum alloy plate to be straightened is obtained, describing the plate's inherent properties and specifications. This basic information is then used to navigate within a hierarchical index structure pre-built based on the basic information dimensions. This hierarchical index structure organizes the vast model library by dimensions such as alloy type and specification range. By navigating along the path corresponding to the current plate's basic information, a subset of prediction models relevant to the current plate's characteristics can be quickly located. This process avoids traversing the entire model library, significantly improving model location efficiency. Next, within the located prediction model subset, the degree of match between the current plate's basic information and each prediction model in the subset is evaluated based on pre-set matching rules, and a matching score is calculated. The matching rules take into account factors such as the similarity between the model training data and the current plate's basic information. Finally, a pre-set matching threshold is set, and only prediction models with a matching score exceeding this threshold are selected as candidate prediction models. These candidate models are considered to have a high correlation with the current plate's actual properties and can more accurately predict its deformation response.

[0054] By combining hierarchical navigation with matching screening, the solution solves the problem of inefficiency in directly screening from a huge model library, while ensuring that the screened models are highly relevant to the current plate, laying the foundation for subsequent accurate prediction and dynamic control of process parameters.

[0055] Preferably, step A202 may include: Determining a basic matching degree between the basic information and each prediction model in the prediction model subset based on a preset matching rule between the basic information and each prediction model in the prediction model subset; Analyzing actual deformation response characteristics of the aluminum alloy plate based on real-time online detection data during the testing phase; For each prediction model in the prediction model subset, based on the actual deformation response characteristics, calculating a prediction matching degree of the prediction model to the actual deformation response; For each prediction model in the prediction model subset, the corresponding basic matching degree is combined with the corresponding predicted matching degree to obtain a comprehensive matching degree; A group of prediction models whose comprehensive matching degree is higher than a preset matching degree threshold is determined as the candidate prediction models.

[0056] Among them, the preset matching rules can adopt rule-based matching, such as setting matching conditions according to dimensions such as alloy type, thickness range, width range, etc.; or adopt a method based on similarity calculation, such as encoding basic information into a vector and calculating the distance or similarity between the model-associated vector.

[0057] The basic matching degree refers to the degree of match between the basic information of the aluminum alloy plate and the prediction model, calculated according to the preset matching rules. It can be expressed as a percentage, score, or other numerical form, with a higher value indicating a higher matching degree.

[0058] The actual deformation response characteristics are those extracted or calculated from real-time online test data during the testing phase, reflecting the actual deformation behavior of the aluminum alloy plate during the test straightening process. Specifically, these characteristics can be characterized by calculating plate shape change, residual stress distribution, and characteristic parameters of the load-displacement curve. For example, the difference in plate shape data before and after straightening, curvature changes at specific locations, or stress-strain data can be analyzed.

[0059] The prediction match is a numerical value that evaluates the prediction model's ability to predict the actual deformation response characteristics of aluminum alloy plates. Specifically, the prediction model can be applied to the input data from the testing phase to obtain a predicted deformation response. The predicted result is then compared with the actual deformation response characteristics, and the prediction error or similarity is calculated and converted into a match value. For example, the mean squared error between the predicted and actual plate shapes can be calculated; the smaller the mean squared error, the higher the prediction match.

[0060] Fusion refers to combining the basic and predicted matching scores to form a comprehensive evaluation metric. This can be achieved through weighted summation, product, fuzzy logic-based fusion, or other multi-criteria decision-making methods. For example, weights can be set, multiplying the basic matching score by one weight and the predicted matching score by another, and then summing the results to obtain the overall matching score. Weights can be determined empirically or through training.

[0061] The preset matching threshold is the minimum overall matching requirement used to screen candidate prediction models. Only models with an overall matching score above this threshold are selected as candidate prediction models. This threshold can be determined based on experience or through offline testing.

[0062] Specifically, during the straightening test phase of an aluminum alloy plate, basic information such as the plate's alloy grade and specifications is first obtained. Based on this basic information, a hierarchical index structure is used to quickly locate a relevant subset of prediction models within a pre-set prediction model library. For each prediction model in this subset, the degree of match between the model's scope of application and the current plate's basic information is calculated according to pre-set rules or methods to obtain a basic matching degree. Simultaneously, during the straightening test, real-time online inspection data of the plate is continuously collected, such as plate shape data before and after straightening. Using this real-time data, the actual deformation response characteristics of the aluminum alloy plate under the test conditions, such as shape change and residual stress characteristics, are analyzed and extracted. Then, for each model in the prediction model subset, the deformation response of the plate is predicted using the input data from the test phase. The predicted results are compared with the actual deformation response characteristics, and the accuracy or similarity of the model's prediction to the actual deformation is calculated to obtain a predicted matching degree. Next, the calculated basic matching degree is combined with the predicted matching degree, for example, using a weighted average, to obtain a comprehensive matching degree for each model. Finally, a comprehensive matching threshold is set, and all prediction models with a matching degree above this threshold are screened and assembled into a set of candidate prediction models. The models in this set are not only relevant to the basic properties of the sheet metal but also demonstrate good predictive capabilities in actual testing, making them more likely to be the most suitable prediction model for the sheet metal. This approach overcomes the limitations of relying solely on static basic information for matching, improves the quality of the candidate model set, and lays the foundation for subsequent selection of the optimal model and precise control of process parameters.

[0063] Through the above technical solution, this application solves the problem that relying solely on the basic information of the plate for matching cannot fully reflect the actual deformation characteristics, resulting in an inaccurate set of candidate models. By introducing an analysis of the actual deformation response characteristics during the test phase and integrating it with the matching degree based on basic information, the applicability of the prediction model can be evaluated more comprehensively and accurately. As a result, the selected candidate prediction model set can more effectively reflect the model's ability to predict the actual deformation behavior of the current plate, improve the quality of the candidate models, provide a more reliable basis for the subsequent selection of the optimal model, and thus improve the accuracy and robustness of the entire process parameter dynamic control method.

[0064] In some embodiments, step A3 comprises: A301. Based on the preset process parameters used in the test phase and the input data in the real-time online detection data of the test phase, using each of the candidate prediction models, predict the deformation response of the aluminum alloy plate; A302. Compare the deformation response predicted by each candidate prediction model with the actual deformation response reflected by the real-time online detection data of the test phase, and calculate the prediction error of each candidate prediction model; A303. Determine the candidate prediction model with the smallest prediction error as the optimal prediction model.

[0065] Among them, predicting the deformation response of the aluminum alloy plate refers to the state of the aluminum alloy plate after straightening or the change in its state calculated by the prediction model based on the input process parameters and plate state data, such as the predicted plate shape data after straightening or the change in the plate shape data after straightening relative to the plate shape data before straightening.

[0066] Among them, calculating the prediction error refers to the process of quantifying the difference between the model prediction results and the actual observation results. Indicators such as root mean square error, mean absolute error, and maximum absolute error can be used.

[0067] This solution addresses the problem of objectively selecting the optimal model from multiple candidate prediction models based on test data by providing specific implementation steps for determining the optimal prediction model. Specifically, step A301 utilizes the preset process parameters and collected input data from the actual testing phase to input these actual test conditions into each candidate prediction model, enabling each model to output its predicted deformation response for the same actual operating conditions. This step provides a unified set of prediction results based on actual test data for subsequent evaluation of the prediction capabilities of each model. Step A302 compares the deformation responses predicted by each candidate model in step A301 with the actual deformation responses of the aluminum alloy plate as reflected in the real-time online test data from the testing phase. This comparison quantifies the difference between each model's prediction results and the observed results, thereby calculating the prediction error for each model. This step provides an objective metric for evaluating model prediction accuracy, allowing the performance of different models to be directly compared. Step A303, based on the prediction error calculated in step A302, selects the candidate prediction model with the smallest prediction error as the optimal prediction model. The smallest prediction error indicates that the model achieved the highest prediction accuracy for the actual deformation response of the aluminum alloy plate during the testing phase. By selecting the model with the highest prediction accuracy, it can be ensured that the model used in the subsequent formal straightening stage can best reflect the actual deformation characteristics of the current plate, laying the foundation for accurate process parameter control.

[0068] Combining the testing phase environment and candidate model set provided by previous solutions, this solution can identify the model that most accurately predicts the deformation behavior of the specific plate being processed from a preselected library. This overcomes the inaccuracies of traditional methods that rely on fixed parameter tables or simplified models, improves the model's adaptability to actual plate properties, and thus provides a more reliable foundation for subsequent online control. This mechanism of model evaluation and selection based on actual test data is the foundation of the entire dynamic control method's ability to effectively respond to different plate properties and production fluctuations.

[0069] In one embodiment, the step of determining the optimal prediction model can be implemented as follows: First, in step A301, the preset reduction, preset straightening speed, pre-straightening flatness data, thickness data, and temperature data collected during the testing phase are input to each candidate prediction model. Each model outputs predicted post-straightening flatness data based on its internal algorithm and parameters. Next, in step A302, the predicted post-straightening flatness data output by each candidate model is compared point by point with the actual post-straightening flatness data detected during the testing phase, and the difference between the two is calculated. The root mean square error (RMSE) can be used as a metric for prediction error: the square root of the average of the squares of the differences between the predicted and actual values ​​is calculated. Finally, in step A303, after calculating the RMSE values ​​of all candidate prediction models, the model with the smallest RMSE value is found and determined as the optimal prediction model.

[0070] Through the above steps, this solution objectively and quantitatively evaluates the prediction accuracy of each candidate prediction model under actual test conditions. Selecting the model with the smallest prediction error ensures that the selected model most accurately reflects the deformation characteristics of the plate under actual working conditions. This provides a more accurate prediction basis for online control of process parameters based on this model during the subsequent formal straightening stage, thereby improving straightening results and product quality.

[0071] In some embodiments, step A4 comprises: A401. After initiating the formal straightening phase, determine initial process parameters using the optimal prediction model based on the initially detected input data, and collect real-time online detection data during the formal straightening process to obtain the real-time online detection data for the formal straightening process; A402. Extracting the current moment's real-time online detection data from the real-time online detection data during the formal straightening to determine the actual deformation response of the aluminum alloy plate at the current moment; A403 based on the current moment of the process parameters and the current moment of the real-time online detection data input data, using the current optimal prediction model, predicting the deformation response of the aluminum alloy plate, to obtain the current moment of the predicted deformation response; A404 compares the actual deformation response at the current moment and the predicted deformation response at the current moment, and calculates the predicted deviation at the current moment; A405. Based on the current prediction deviation and the preset adjustment rules, calculate the adjustment amount of the internal parameters of the current optimal prediction model; A406. Update the internal parameters of the current optimal prediction model according to the adjustment amount to obtain the optimal prediction model after online adjustment.

[0072] Among them, the internal parameters of the optimal prediction model refer to the variable values ​​or weights in the mathematical expression or algorithm structure that constitutes the optimal prediction model. These parameters determine how the model maps input data and process parameters to predicted deformation responses. They can be implemented using the connection weights and biases of the neural network, the coefficients of the regression model, or the material property parameters in the physical model.

[0073] The adjustment rule refers to the pre-defined logic or algorithm that guides the system in calculating the adjustment amount for the model's internal parameters based on the prediction deviation. This can be implemented using an optimization algorithm based on error gradients (such as gradient descent), a rule-based expert system, or an adaptive control law. The adjustment amount is the specific numerical correction required to the internal parameters of the optimal prediction model, calculated according to the adjustment rule. It can be represented in the form of a vector or matrix, corresponding to each parameter in the model that needs to be updated.

[0074] This solution details how to perform online adjustments to the optimal prediction model determined during the testing phase during the actual straightening phase. This ensures that the model consistently and accurately predicts the deformation response of aluminum alloy plates during production, thereby maintaining or improving the straightening effect. Specifically, upon initiating the actual straightening phase, the system first calculates initial process parameters based on the initially detected input data using the optimal prediction model determined during the previous testing phase. This provides a starting point for the final production phase based on the optimal model. Simultaneously, real-time online inspection data is continuously collected throughout the actual straightening process. This data serves as the basis for subsequent online model adjustments and dynamic control of process parameters. The system extracts the current real-time online inspection data from the continuously collected actual straightening data stream. This data contains information about the plate shape after straightening and directly reflects the actual deformation result of the aluminum alloy plate under the current process parameters, i.e., the actual deformation response at the current moment. The system then uses the current actual process parameters and the current input data (such as pre-straightening plate shape, thickness, and temperature) to feed this information into the current (possibly adjusted) optimal prediction model. The prediction model then predicts the expected deformation response of the plate under these conditions, resulting in the model-predicted deformation response at the current moment. By comparing the actual deformation response at the current moment with the model's predicted deformation response at the current moment, the system calculates the difference between the two, known as the current prediction deviation. This deviation quantifies the prediction error of the current optimal prediction model under the current production conditions. Based on the calculated current prediction deviation and in combination with preset adjustment rules, the system calculates the specific adjustments required to the internal parameters of the optimal prediction model. These adjustment rules are pre-defined strategies that guide the system on how to modify the model based on the prediction error. For example, larger errors may lead to larger adjustments, or different adjustment methods may be used depending on the nature of the error (such as systematic deviation or random fluctuation). This step converts the prediction error into correction instructions for the model parameters. Finally, the system updates the internal parameters of the optimal prediction model based on the calculated adjustments. By updating the model parameters, the model better fits the actual deformation response at the current moment, thereby reducing future prediction deviations. The updated model becomes the "online adjusted optimal prediction model" and is used for subsequent predictions and process parameter calculations, forming a closed loop of continuous learning and adaptation.

[0075] By looping through the above steps, this solution implements an online model adjustment mechanism based on real-time prediction deviations. This allows the optimal prediction model to dynamically adapt to the various uncertainties and changes in the actual production process, maintaining high prediction accuracy, thereby providing a more accurate prediction basis for subsequent model-based dynamic control of process parameters. This online adjustment mechanism, combined with the initial process of screening the optimal model through test data, enables the entire process parameter control method to balance the generalization capabilities of offline optimization with the real-time nature of online adaptation, improving the system's adaptability to different plate materials and changing working conditions.

[0076] Preferably, step A405 may include: Determine the current production stage based on the predicted deviation at the current moment and the real-time online detection data during the formal straightening; the production stage includes an initial transient stage and a steady-state stage; According to the determined production stage, a corresponding adjustment rule is selected from a preset adjustment rule library; the adjustment rule library includes at least one adjustment rule applicable to the initial transient stage and at least one adjustment rule applicable to the steady-state stage; Based on the prediction deviation at the current moment and the selected adjustment rule, a coordinated adjustment amount of multiple related parameters within the current optimal prediction model is calculated as the adjustment amount.

[0077] The production stage refers to the different time or spatial intervals during the formal straightening process of aluminum alloy plates in a multi-roll straightener, divided according to the plate deformation behavior and model prediction deviation characteristics. It can be defined based on indicators such as the processing length and processing time after the start of straightening, and the size or rate of change of the prediction deviation. The initial transient stage refers to the early stage at the beginning of the straightening process, when the plate shape changes rapidly and the prediction deviation may be large. The steady-state stage refers to the later stage after the straightening process has been carried out for a period of time, when the plate shape tends to be stable and the fluctuation of the prediction deviation is small.

[0078] The adjustment rule base refers to a pre-stored set of algorithms, formulas, or lookup tables used to calculate model parameter adjustments based on forecast deviations. It can include multiple adjustment strategies optimized for different production stages. An adjustment rule is a specific strategy or algorithm within the adjustment rule base that guides how to convert forecast deviations into model parameter adjustments. It can be a mathematical function, a set of parameters, a decision tree, or a small controller.

[0079] Among them, the collaborative adjustment amount refers to the simultaneous adjustment of multiple interrelated parameters within the optimal prediction model. The adjustment amounts of these parameters are not calculated in isolation, but take into account the mutual influence between them to achieve the overall optimal adjustment effect. It can be a vector or a set of related numerical values.

[0080] This solution addresses the issue of the adjustment rules used to calculate model parameter adjustments during online adjustment of the optimal prediction model, which may not adapt to different production stages. Therefore, a method is proposed to dynamically select adjustment rules based on the current production stage. Specifically, the current production stage is determined based on the current prediction deviation and real-time online inspection data during the actual straightening process. By analyzing the magnitude and trend of the prediction deviation, as well as the shape changes reflected in the real-time online inspection data, the system can identify whether the straightening process is in the initial transient stage, where the shape is rapidly converging, or the steady-state stage, where the shape is stable and fluctuations are minimal. This determination is necessary because the deformation behavior of the sheet metal and the characteristics of the model prediction deviation vary at different stages, requiring different model adjustment strategies. Based on the determined production stage, the corresponding adjustment rule is then selected from a pre-set adjustment rule library. This library contains adjustment rules optimized for different production stages. For example, a rule for the initial transient stage may prioritize rapidly reducing large prediction deviations, while a rule for the steady-state stage may prioritize maintaining model stability and suppressing small fluctuations. By selecting the most appropriate adjustment rule based on the current stage, the effectiveness and pertinence of the model adjustment strategy can be ensured. Finally, based on the current forecast deviation and the selected adjustment rule, the coordinated adjustment amount for multiple related parameters within the optimal forecast model is calculated as the adjustment amount. Here, the current forecast deviation is used as the basis for adjustment, combined with the adjustment rule selected for the current production stage, to calculate the required adjustment amount for the internal parameters of the optimal forecast model. The emphasis on the coordinated adjustment of multiple related parameters means that the adjustment process takes into account the mutual influence between different parameters within the model and performs holistic optimization adjustments, rather than adjusting individual parameters in isolation.

[0081] By determining the current production stage based on the current prediction deviation and real-time online detection data during formal straightening, and selecting the corresponding adjustment rule from a preset adjustment rule library based on the determined production stage, and then calculating the coordinated adjustment amount for multiple related parameters within the optimal prediction model based on the current prediction deviation and the selected adjustment rule, this solution can dynamically adapt the online adjustment strategy of the optimal prediction model to the different stages of the straightening process. This overcomes the problem that using fixed adjustment rules cannot effectively respond to the deformation characteristics and model deviation characteristics of the plate at different production stages, and improves the accuracy, stability, and adaptability of the model's online adjustment. As a result, the optimal prediction model after online adjustment can more accurately predict the plate's deformation response, providing a more reliable basis for subsequent online adjustment of process parameters.

[0082] In some embodiments, step A5 comprises: A501. According to the real-time online detection data and the preset straightening requirements at the current moment, determine the target deformation response; A502. Using the optimal prediction model after online adjustment, and under the input data conditions of the real-time online detection data at the current moment, solving for process parameters that can make the predicted deformation response of the optimal prediction model after online adjustment reach the target deformation response, and using these as target process parameters; A503. Adjust the process parameters according to the target process parameters.

[0083] Among them, the target deformation response refers to the deformation state or plate quality index that the aluminum alloy plate is expected to achieve after straightening, which is determined based on the current state of the plate reflected by the real-time online detection data at the current moment and the preset straightening requirements. It can be characterized by the target plate flatness, target residual stress level or a specific plate curve.

[0084] Among them, solving the process parameters that can make the predicted deformation response of the optimal prediction model after online adjustment reach the target deformation response refers to using mathematical or computational methods, based on the optimal prediction model after online adjustment, given the input data in the real-time online detection data at the current moment as part of the model input, and setting the target deformation response as the expected output of the model, and reversely calculating the process parameter combination that can make the model output reach the expected value. This can be achieved using numerical optimization algorithms, model inversion technology, or a lookup table-based method. The target process parameters refer to the process parameter values ​​calculated through the above-mentioned solution process and used to guide the parameter setting of the straightening machine at the current moment. They represent the precise setting values ​​of parameters such as the reduction and straightening speed required to achieve the target deformation response.

[0085] This solution introduces the concept of a target deformation response and uses an online-adjusted prediction model for inverse analysis to accurately determine the process parameters required to achieve this goal. Specifically, based on current real-time online inspection data and pre-set straightening requirements, a specific desired deformation state, known as the target deformation response, is determined. This step provides a clear control target for subsequent parameter calculations. Determining the target deformation response based on current real-time online inspection data ensures that the target more accurately reflects current production conditions, rather than relying solely on pre-set ideal requirements. Next, an inverse analysis is performed using the optimal prediction model, which has been adjusted online to more accurately reflect current sheet material characteristics and equipment status, using the input data from the current real-time online inspection data as the model input conditions. Prediction models are typically used to predict the deformation response based on process parameters and input data. This step involves determining the process parameters that produce this desired deformation response, given the desired deformation response (target deformation response) and the current input data. Using the online-adjusted optimal prediction model for the solution ensures that the calculated target process parameters are based on the most accurate model available, thereby improving parameter reliability. The solution is performed under the input data conditions of the real-time online detection data at the current moment, ensuring that the calculation results are applicable to the current actual production status. The process parameters obtained by the solution are the target process parameters. Finally, according to the calculated target process parameters, the process parameters of the straightening machine are actually adjusted. This is the process of converting theoretical calculation results into actual control instructions. Adjustment according to the target process parameters means that the adjustment is based on precise calculation results, rather than simple empirical adjustments or fixed rules. This can more effectively achieve the set target deformation response, improve straightening accuracy and effect, and better cope with fluctuations in plate performance and changes in equipment status. By combining real-time detection data, online adjustment models and goal-oriented inverse solutions, this solution can achieve refined and adaptive dynamic control of the aluminum alloy plate straightening process.

[0086] In one specific embodiment, based on current real-time online detection data, such as the current sheet material's real-time flatness data (e.g., wave height, edge wave, or mid-wave extent), and preset straightening requirements (e.g., final flatness level), a target deformation response can be determined, such as reducing the current wave height to a specific value or achieving a target residual stress distribution. Then, using an online-adjusted optimal prediction model, which may be a neural network model calibrated with real-time data, an optimization algorithm, such as a gradient-based optimization method, is run to iteratively adjust the model's process parameter inputs (e.g., roll reduction and straightening speed) based on the input data from the current real-time online detection data, such as the current sheet material's thickness, width, and temperature, as well as the pre-straightening flatness data, until the error between the model's predicted deformation response (e.g., post-straightening flatness data) and the set target deformation response is less than a preset threshold. The process parameter inputs at this point are then determined as the target process parameters. Finally, the calculated target reduction amount and straightening speed are sent to the control system of the straightening machine, which drives the actuator (such as a hydraulic cylinder or motor) to accurately adjust the relative position of the straightening rollers and the conveying speed of the plate, thereby realizing the adjustment of process parameters.

[0087] This technical solution accurately calculates the process parameters required to achieve the desired straightening target based on the real-time plate state and straightening objectives, overcoming the limitations of traditional methods that rely on fixed parameter tables or simple feedback control. Utilizing an online, adjusted predictive model for reverse engineering ensures the accuracy of parameter calculations, better adapting to fluctuations in plate properties and changes in equipment status. This improves the dynamic adaptability and control precision of the straightening process, resulting in smoother, more stable aluminum alloy plates.

[0088] refer to Figure 2 The present application provides a dynamic control system for aluminum alloy plate process parameters, which is used to adjust the process parameters of a multi-roll straightening machine when straightening an aluminum alloy plate. The system includes: Information acquisition module 1 is used to obtain basic information of the aluminum alloy plate and real-time online detection data during the testing phase; during the testing phase, the aluminum alloy plate is straightened using preset process parameters (the specific process can be referred to step A1 above); a candidate model screening module 2, configured to determine a set of candidate prediction models from a preset prediction model library based on the basic information; each prediction model in the prediction model library is configured to predict the deformation response of the aluminum alloy plate based on process parameters and real-time online detection data (for details, refer to step A2 above); The optimal model determination module 3 is used to determine the optimal prediction model with the best prediction accuracy for the actual deformation response of the aluminum alloy plate from among the candidate prediction models based on the real-time online detection data in the test phase (the specific process can be referred to step A3 above); Model adjustment module 4, configured to perform formal straightening of the aluminum alloy plate using the optimal prediction model, and to perform online adjustment of the optimal prediction model based on real-time online detection data during formal straightening (for the specific process, please refer to step A4 above); The process parameter adjustment module 5 is used to adjust the process parameters online according to the current real-time online detection data using the optimal prediction model adjusted online (the specific process can be referred to step A5 above).

[0089] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for dynamically controlling the process parameters of an aluminum alloy plate, which is used to adjust the process parameters of an aluminum alloy plate when a multi-roll straightening machine is used to straighten the aluminum alloy plate, characterized in that: The steps of the method include: A1. Obtain basic information of the aluminum alloy plate, as well as real-time online detection data during the testing phase; during the testing phase, straighten the aluminum alloy plate using preset process parameters; A2. Based on the basic information, a set of candidate prediction models is determined from a preset prediction model library; each prediction model in the prediction model library is used to predict the deformation response of the aluminum alloy plate based on process parameters and real-time online detection data; A3. Based on the real-time online detection data of the test phase, the optimal prediction model with the best prediction accuracy for the actual deformation response of the aluminum alloy plate is determined among the candidate prediction models; A4. Performing formal straightening of the aluminum alloy plate using the optimal prediction model, and adjusting the optimal prediction model online based on real-time online detection data during formal straightening; A5. Based on the current real-time online detection data, use the optimal prediction model after online adjustment to adjust the process parameters online.

2. The method for dynamically controlling process parameters of an aluminum alloy plate according to claim 1, characterized in that: The basic information includes alloy grade, specification, thickness, width, length and surface condition; The real-time online detection data includes input data and response data; the input data includes flatness data, thickness data and temperature data before straightening; the response data includes flatness data after straightening; The process parameters include a pressing amount and a straightening speed; the preset process parameters include a preset pressing amount and a preset straightening speed.

3. The method for dynamically controlling the process parameters of an aluminum alloy plate according to claim 2, wherein: Step A1 includes: A101 obtain the basic information of the aluminum alloy plate; A102. Use the preset process parameters to test straightening of the aluminum alloy plate, and during the test straightening process, continuously collect the real-time online detection data to form a real-time online detection data sequence; A103. Calculate the deformation response characteristic parameters of the aluminum alloy plate within the current test length range based on the real-time online detection data sequence; A104. If the deformation response characteristic parameter does not meet the preset test end threshold condition, continue to collect the real-time online detection data and update the real-time online detection data sequence, and return to step A103; otherwise, end the test phase and use the current real-time online detection data sequence as the real-time online detection data of the test phase.

4. The method for dynamically controlling process parameters of an aluminum alloy plate according to claim 2, wherein: Step A2 includes: A201. Based on the basic information, navigate along a hierarchical index structure in the preset prediction model library to locate a subset of prediction models related to the basic information; the hierarchical index structure is pre-constructed based on multiple information dimensions of the basic information; A202. Based on the preset matching rules between the basic information and each prediction model in the prediction model subset, determine a group of prediction models whose matching degree with the basic information is higher than a preset matching degree threshold as the candidate prediction models.

5. The method for dynamically controlling process parameters of an aluminum alloy plate according to claim 4, characterized in that: Step A202 includes: Determining a basic matching degree between the basic information and each prediction model in the prediction model subset based on a preset matching rule between the basic information and each prediction model in the prediction model subset; Analyzing actual deformation response characteristics of the aluminum alloy plate based on real-time online detection data during the testing phase; For each prediction model in the prediction model subset, based on the actual deformation response characteristics, calculating a prediction matching degree of the prediction model to the actual deformation response; For each prediction model in the prediction model subset, the corresponding basic matching degree is combined with the corresponding predicted matching degree to obtain a comprehensive matching degree; A group of prediction models whose comprehensive matching degree is higher than a preset matching degree threshold is determined as the candidate prediction models.

6. The method for dynamically controlling process parameters of an aluminum alloy plate according to claim 2, characterized in that: Step A3 includes: A301. Based on the preset process parameters used in the test phase and the input data in the real-time online detection data of the test phase, using each of the candidate prediction models, predict the deformation response of the aluminum alloy plate; A302. Compare the deformation response predicted by each candidate prediction model with the actual deformation response reflected by the real-time online detection data of the test phase, and calculate the prediction error of each candidate prediction model; A303. Determine the candidate prediction model with the smallest prediction error as the optimal prediction model.

7. The method for dynamically controlling process parameters of an aluminum alloy plate according to claim 2, characterized in that: Step A4 includes: A401. After initiating the formal straightening phase, determine initial process parameters using the optimal prediction model based on the initially detected input data, and collect real-time online detection data during the formal straightening process to obtain the real-time online detection data for the formal straightening process; A402. Extracting the current moment's real-time online detection data from the real-time online detection data during the formal straightening to determine the actual deformation response of the aluminum alloy plate at the current moment; A403 based on the current moment of the process parameters and the current moment of the real-time online detection data input data, using the current optimal prediction model, predicting the deformation response of the aluminum alloy plate, to obtain the current moment of the predicted deformation response; A404 compares the actual deformation response at the current moment and the predicted deformation response at the current moment, and calculates the predicted deviation at the current moment; A405. Based on the current prediction deviation and the preset adjustment rules, calculate the adjustment amount of the internal parameters of the current optimal prediction model; A406. Update the internal parameters of the current optimal prediction model according to the adjustment amount to obtain the optimal prediction model after online adjustment.

8. The method for dynamically controlling process parameters of an aluminum alloy plate according to claim 7, characterized in that: Step A405 includes: Determine the current production stage based on the predicted deviation at the current moment and the real-time online detection data during the formal straightening; the production stage includes an initial transient stage and a steady-state stage; According to the determined production stage, a corresponding adjustment rule is selected from a preset adjustment rule library; the adjustment rule library includes at least one adjustment rule applicable to the initial transient stage and at least one adjustment rule applicable to the steady-state stage; Based on the prediction deviation at the current moment and the selected adjustment rule, a coordinated adjustment amount of multiple related parameters within the current optimal prediction model is calculated as the adjustment amount.

9. The method for dynamically controlling process parameters of an aluminum alloy plate according to claim 2, wherein: Step A5 includes: A501. According to the real-time online detection data and the preset straightening requirements at the current moment, determine the target deformation response; A502. Using the optimal prediction model after online adjustment, and under the input data conditions of the real-time online detection data at the current moment, solving for process parameters that can make the predicted deformation response of the optimal prediction model after online adjustment reach the target deformation response, and using these as target process parameters; A503. Adjust the process parameters according to the target process parameters.

10. A dynamic control system for process parameters of aluminum alloy plates, used to adjust the process parameters of aluminum alloy plates when straightening by a multi-roller straightening machine, characterized in that: The system includes: An information acquisition module is used to obtain basic information of the aluminum alloy plate and real-time online detection data during the testing phase; during the testing phase, the aluminum alloy plate is straightened using preset process parameters; a candidate model screening module, configured to determine a set of candidate prediction models from a preset prediction model library based on the basic information; each prediction model in the prediction model library is configured to predict the deformation response of the aluminum alloy plate based on process parameters and real-time online detection data; An optimal model determination module is used to determine, from among the candidate prediction models, an optimal prediction model with the best prediction accuracy for the actual deformation response of the aluminum alloy plate based on the real-time online detection data during the test phase; A model adjustment module, configured to perform formal straightening on the aluminum alloy plate using the optimal prediction model, and to perform online adjustment on the optimal prediction model based on real-time online detection data during formal straightening; The process parameter adjustment module is used to adjust the process parameters online according to the real-time online detection data at the current moment and use the optimal prediction model after online adjustment.

Images