Thermal forming correction control system of non-standard mineral casting

Abstract

The invention discloses a thermal forming correction control system of a non-standard mineral casting, and belongs to the field of intelligent control, and the thermal forming correction control system comprises the following steps: obtaining design requirements of the non-standard mineral casting to obtain a casting segmentation result; according to material requirements, a sample data set is constructed through multi-dimensional experimental data, and learning is carried out to obtain a parameter control analysis model; multi-dimensional parameter analysis is carried out, and control parameters of each casting segmentation area are determined; based on the control parameters, acquiring an abnormal case set, performing abnormal probability calculation, converting the abnormal case set into a control coefficient, performing abnormal parameter tracing, and determining abnormal control parameters; and thermal forming correction control information is obtained and sent to the controller for control. According to the non-standard mineral casting thermal forming correction control method, the technical problem that the correction effect is poor due to the fact that uniform parameters are adopted for existing non-standard mineral casting thermal forming correction control is solved, and the technical effects that the non-standard casting is subjected to differentiated segmentation, different areas are correspondingly controlled and the overall correction effect is improved according to geometrical characteristics and material characteristics are achieved.

Description

Technical Field

[0001] The invention relates to the field of intelligent control, and in particular to a hot forming correction control system for non-standard mineral castings. Background Art

[0002] The existing hot forming correction control of non-standard mineral castings usually uses unified process parameters to process the entire non-standard mineral casting. However, since non-standard castings often have complex geometric features and material differences, it is difficult to achieve a satisfactory correction effect by using unified parameters. On the one hand, there may be obvious geometric and material differences in different areas of the casting. If unified parameters are used, it is difficult to take into account the characteristics of each area, which may easily lead to insufficient or excessive local correction, affecting the final correction effect; on the other hand, some abnormal conditions may occur during the hot forming process of the casting, such as insufficient heating temperature, improper control of correction force, etc. If they are not discovered and adjusted in time, it will also affect the final correction effect. Summary of the invention

[0003] The present application provides a hot forming correction control system for non-standard mineral castings, aiming to solve the technical problem that the existing hot forming correction control of non-standard mineral castings adopts unified parameters for control, resulting in poor correction effect.

[0004] In view of the above problems, the present application provides a hot forming correction control system for non-standard mineral castings, including: a casting segmentation module, which is used to obtain the design requirements of non-standard mineral castings, including design drawings, description parameters, and material requirements, and to segment the design drawings according to the identification difference points based on the description parameters. The segmentation nodes are the points with the smallest identification difference, and the segmentation parts are the areas with identification difference, so as to obtain the casting segmentation results; a model acquisition module, which is used to construct a sample data set according to the material requirements, through the material cumulative thickness, component shape, control temperature, correction force, correction time, and multi-dimensional experimental data, and to learn through the sample data set to obtain a parameter control analysis model; a control parameter determination module, which is used to determine the parameter control analysis model based on the casting segmentation results. According to the casting shape and description parameters, multi-dimensional parameter analysis is performed through the parameter control analysis model to determine the control parameters of each casting segmentation area. The control parameters of each casting segmentation area include control temperature, correction force, and correction time; the abnormality tracing module is used to obtain an abnormal case set based on the control parameters of each casting segmentation area, calculate the abnormal probability of each casting segmentation area using the abnormal case set, convert it into a control coefficient according to the abnormal probability, and trace the abnormal parameters to determine the abnormal control parameters; the correction control module is used to perform global optimization based on the control coefficient, the control parameters of each casting segmentation area, and the abnormal control parameters in combination with the casting segmentation area to obtain hot forming correction control information and send it to the controller for control.

[0005] The technical solution provided in this application has at least the following technical effects or advantages: By obtaining the design drawings, description parameters and material requirements of non-standard castings, segmenting the design drawings, determining the differentiated areas, and obtaining the casting segmentation results, the aim is to perform differentiated analysis and regional division on the geometric characteristics and material differences of castings, laying the foundation for subsequent targeted control; according to material requirements, a sample data set is constructed through multi-dimensional experimental data of material cumulative thickness, component shape, heating temperature, correction force and time, so as to establish a parameter control analysis model to provide a basis for subsequent parameter determination; using the parameter control analysis model, the corresponding control parameters are determined for different casting segmentation areas, including control temperature, correction force and correction time, so as to achieve differentiated control according to the specific characteristics of each area. Control measures are taken to improve the overall correction effect; the corresponding control coefficient is calculated according to the abnormal probability of each casting segmentation area, and the abnormal parameters are traced and analyzed to timely discover and adjust when abnormalities occur in the production process to ensure the overall correction quality; the control parameters, abnormal control parameters and control coefficients of each casting segmentation area are globally optimized to obtain hot forming correction control information, and the technical solution is sent to the controller for execution, which solves the technical problem that the existing hot forming correction control of non-standard mineral castings adopts unified parameters for control, resulting in poor correction effect, and achieves the technical effect of differentiated segmentation of non-standard castings according to geometric features and material properties, corresponding control of different areas, and improving the overall correction effect.

[0006] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 A structural schematic diagram of a hot forming correction control system for non-standard mineral castings is provided for an embodiment of the present application; Figure 2 A flow chart of a parameter control analysis model obtained in a hot forming correction control system for non-standard mineral castings is provided for an embodiment of the present application; Explanation of the reference numerals: casting segmentation module 11 , model acquisition module 12 , control parameter determination module 13 , abnormality tracing module 14 , correction control module 15 . DETAILED DESCRIPTION

[0008] The overall idea of ​​the technical solution provided by this application is as follows: The embodiment of the present application provides a hot forming correction control system for non-standard mineral castings. First, based on the design requirements of non-standard mineral castings, the non-standard mineral castings are differentiated and segmented to find the minimum differentiated points and differentiated areas. Then, a sample data set is constructed through a large amount of experimental data to obtain a parameter control analysis model to achieve data-driven intelligent parameter selection. On this basis, multi-dimensional parameter analysis is performed on each casting segmentation area to obtain the control parameters of each casting segmentation area, including control temperature, correction force, and correction time. At the same time, an abnormal case set and abnormal probability calculation are introduced to improve the robustness and adaptability of the correction process by tracing the control coefficient and abnormal parameters. Subsequently, the control coefficient, the control parameters of each casting segmentation area, and the abnormal control parameters are comprehensively considered to perform global optimization to obtain the hot forming correction control information of the entire casting, and hot forming correction control is performed on this basis.

[0009] In general, a complete non-standard mineral casting hot forming correction control system is formed through casting segmentation, data modeling, regional optimization control, exception handling and other means. Compared with the existing technology, it fully considers the local differences of castings, realizes precise control, and thus improves the correction effect.

[0010] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification.

[0011] Embodiment 1 like Figure 1 As shown, the embodiment of the present application provides a hot forming correction control system for non-standard mineral castings, the system comprising: The casting segmentation module 11 is used to obtain the design requirements of non-standard mineral castings, including design drawings, description parameters, and material requirements, and segment the design drawings according to the identification difference points based on the description parameters. The segmentation nodes are the points with the smallest identification difference, and the segmentation parts are the areas with identification difference, so as to obtain the casting segmentation results; Specifically, the casting segmentation module 11 first obtains the design requirements of non-standard mineral castings, including design drawings, description parameters and material requirements. Among them, the design drawing is the geometric shape drawing of the casting; the description parameters are the size, tolerance and other parameters of the casting; the material requirements are the types of materials and performance requirements required for the casting. After obtaining the design requirements of non-standard mineral castings, the design drawing is segmented based on the description parameters to identify the differentiated features of the castings in different regions. Among them, the principle of segmentation is to find the point with the smallest recognition difference as the segmentation node, and divide the area with recognition difference into different segmentation components. For example, using algorithms based on image processing, such as edge detection, regional growth, etc., it can also be segmented according to the geometric features of the casting, such as corners, holes, etc. Through segmentation, the non-standard mineral casting is divided into several relatively consistent sub-regions, and the casting segmentation results are obtained, so as to facilitate the targeted setting of the hot forming correction control parameters, and lay the foundation for the subsequent control parameter analysis and optimization.

[0012] The model acquisition module 12 is used to construct a sample data set according to the material requirements through the material cumulative thickness, component shape, control temperature, correction force, correction time, and multi-dimensional experimental data, and learn through the sample data set to obtain a parameter control analysis model; Specifically, in order to obtain the optimal hot forming correction control parameters, a parameter control analysis model is constructed through the model acquisition module 12. The model is based on multi-dimensional experimental data of material requirements and process parameters and is obtained through a machine learning method.

[0013] First, according to the material requirements, such as material type, hardness, toughness, etc., a series of experiments are designed to collect correction effect data under different combinations of process parameters. Among them, the process parameters include the cumulative thickness of the material, the shape of the casting component, the control temperature, the correction force, the correction time, etc. By changing the values ​​of the parameters, multiple experiments are carried out to obtain multi-dimensional experimental data. Then, the multi-dimensional experimental data is used as a sample to obtain a sample data set, and a machine learning algorithm is used for training and learning. For example, a regression model in supervised learning, such as support vector machine, neural network, etc., is used; or a strategy model in reinforcement learning, such as Q learning, strategy gradient, etc. is used. By learning the relationship between the parameters in the sample data set, a mathematical model that can predict the control temperature, correction force, and correction time according to the cumulative thickness of the material and the shape of the casting component is established, that is, a parameter control analysis model, which is used to guide the setting of hot forming correction control parameters in actual production.

[0014] A control parameter determination module 13 is used to perform multi-dimensional parameter analysis based on the casting segmentation result, according to the casting shape and description parameters, through the parameter control analysis model, to determine the control parameters of each casting segmentation area, wherein the control parameters of each casting segmentation area include control temperature, correction force, and correction time; Specifically, after obtaining the parameter control analysis model, the control parameter determination module 13 uses the parameter control analysis model to determine the control parameters of each casting segmentation area.

[0015] First, the casting segmentation results are obtained, including the geometric shape, size, and adjacent relationship of each casting segmentation area. Then, according to the characteristics of each casting segmentation area, that is, the cumulative thickness of the material and the component shape of the casting segmentation area, the feature extraction is performed to obtain the regional feature vector, which is input into the parameter control analysis model. Through the calculation and analysis of the model, the control parameters of the casting segmentation area are obtained, including the control temperature, correction force, correction time, etc. The above process is repeated until the control parameters of all casting segmentation areas are obtained as the basis for subsequent hot forming correction control.

[0016] The abnormality tracing module 14 is used to obtain an abnormal case set based on the control parameters of each casting segmentation area, calculate the abnormal probability of each casting segmentation area using the abnormal case set, convert it into a control coefficient according to the abnormal probability, and trace the abnormal parameters to determine the abnormal control parameters; Specifically, in the actual hot forming correction process, some abnormal situations may occur, such as material defects, equipment failures, etc., resulting in unsatisfactory correction effects in some casting segmentation areas. In order to improve the stability and reliability of the correction, these abnormal situations need to be predicted and processed.

[0017] First, the abnormal tracing module 14 retrieves historical correction data based on the control parameters of each casting segmentation area, and obtains an abnormal case set, which contains various abnormal situations that may occur in the corresponding casting segmentation area and control parameters and their. Then, using the abnormal case set, the abnormal probability of each casting segmentation area is calculated. For example, statistical methods such as Bayesian estimation, maximum likelihood estimation, etc. are used to analyze the frequency and distribution of various abnormal situations in the abnormal case set, and the probability value of abnormal occurrence in each segmentation area is obtained. Subsequently, the abnormal probability is converted into a control coefficient as the weight of the control parameters of each casting segmentation area, that is, the control coefficient. The higher the casting segmentation area with an abnormal probability, the greater its control coefficient, and it will be given a higher priority and stricter constraints in the subsequent control parameter optimization to reduce the risk of abnormal occurrence. Afterwards, for the casting segmentation area with a higher abnormal probability, the abnormal parameter is traced, the historical abnormal data of the area in the abnormal case set is analyzed, the control parameters that cause the abnormality are found, and they are marked as abnormal control parameters. In the subsequent production process, these abnormal control parameters should be avoided as much as possible to improve the success rate of correction.

[0018] By calculating the probability of abnormality and tracing abnormal parameters, the abnormal risks in the thermoforming correction process can be effectively predicted and avoided, and production efficiency and product quality can be improved. At the same time, converting the probability of abnormality into a control coefficient can take into account the importance and sensitivity of each area when optimizing the control parameters, and achieve more intelligent and refined correction control.

[0019] The correction control module 15 is used to perform global optimization based on the control coefficient, the control parameters of each casting segmentation area, and the abnormal control parameters in combination with the casting segmentation area to obtain hot forming correction control information and send it to the controller for control.

[0020] Specifically, the obtained control coefficients, control parameters of each casting segmentation area, and abnormal control parameters are all local controls for a single casting segmentation area, and do not consider the overall performance and coordination of the entire casting. The correction control module 15 performs global optimization based on the control coefficients, control parameters of each casting segmentation area, and abnormal control parameters to obtain hot forming correction control information.

[0021] The goal of global optimization is to maximize the correction quality and efficiency of the entire casting while meeting the control requirements of each casting segmentation area. For example, the smooth transition and coordination of parameters such as temperature, force, time, etc. between adjacent casting segmentation areas; reasonable arrangement of the correction sequence and timing between different casting segmentation areas to avoid mutual interference; key monitoring and control of areas with high abnormal probability, and adjustment of their control parameters when necessary.

[0022] In order to achieve global optimization, optimization algorithms such as genetic algorithm, particle swarm algorithm, simulated annealing algorithm, etc. can be used. Through intelligent search and iterative optimization, the optimal hot forming correction control information is found under multiple objectives and constraints. For example, the control parameters, abnormal probability, abnormal control parameters and other information of each casting segmentation area are encoded as the input of the optimization algorithm, and the objective function and constraints are set, such as minimizing the correction time, maximizing the correction accuracy, limiting the temperature gradient, etc. Through the iterative calculation of the optimization algorithm, the global optimal control parameter combination is searched as the hot forming correction control information. Subsequently, the optimized hot forming correction control information is sent to the controller, and the controller accurately controls the hot forming correction equipment according to the hot forming correction control information, so as to realize the automatic and intelligent correction of non-standard mineral castings and improve the correction effect.

[0023] Furthermore, the embodiment of the present application also includes: Locate and extract shape change points and size change points according to the design drawing and the description parameters; Clustering is performed according to the shape change points and the size change points to obtain a plurality of change clustering areas; Based on the multiple change clustering areas, the transition area between the multiple change clustering areas without identifiable difference points is obtained for midline segmentation to obtain the casting segmentation result.

[0024] In a feasible implementation manner, another method for obtaining casting segmentation results is provided.

[0025] First, obtain the design drawings and description parameters of non-standard mineral castings. The design drawing is a geometric shape drawing of non-standard mineral castings, which contains information such as the shape, size, structure, etc. of non-standard mineral castings; the description parameters are text descriptions of the characteristics of non-standard mineral castings, such as wall thickness, diameter, height, etc. Then, by analyzing the design drawings and description parameters, identify and extract the shape change points and size change points on non-standard mineral castings. Among them, the shape change point refers to the location where the geometric shape of the casting changes significantly, such as corners, holes, bosses, etc.; the size change point refers to the location where the cross-sectional size of the casting changes significantly, such as thickness mutation, diameter change, height difference, etc.

[0026] Subsequently, cluster analysis is performed on the shape change points and size change points, and the change points with similar spatial positions and similar characteristic attributes are classified into one category to form multiple change clusters. Among them, each change cluster represents a continuous area on the casting with similar change characteristics. In practical applications, appropriate clustering algorithms and parameters are selected according to the characteristics of the casting. For example, for castings with simple shapes and relatively regular changes, distance-based K-means clustering is used to obtain multiple change clusters by setting the number of clusters and iterative optimization; for castings with complex shapes and irregular changes, density-based DBSCAN clustering is used to adaptively discover change clusters of arbitrary shapes by setting the neighborhood radius and minimum number of points of the points, and obtain multiple change clusters. By dividing the change points on the casting into multiple change clusters, it provides an important reference for subsequent segmentation and control.

[0027] Next, based on multiple change clusters, the areas where transition unrecognized difference points are located are found between adjacent change clusters, and the castings are segmented using the midlines of these areas as boundaries. Transition unrecognized difference points refer to points in the transition area between two adjacent change clusters where the geometric and dimensional features change most gently and have the smallest difference from the clusters on both sides. They are located near the boundaries of the change clusters and represent the best positions for feature transition and connection between clusters. At these points, there is no obvious feature recognition difference compared to the points inside the clusters to which they belong, so they are transition unrecognized difference points. First, the transition area between adjacent change clusters is determined. In the transition area, by analyzing the gradient and continuity of feature changes, the areas where transition unrecognized difference points are densely distributed are extracted. The casting is divided into several sub-areas using the midline of the transition unrecognized difference point area as the dividing line to obtain the casting segmentation result.

[0028] Further, such as Figure 2 As shown, the embodiment of the present application also includes: According to the material requirements, the material composition is determined, and multiple groups of experimental samples are shaped according to the material casting thickness and component shape; In this way, the multiple groups of experimental samples are controlled by controlling the temperature, correction strength, and correction time, and the experimental data are recorded to construct the multi-dimensional experimental data, which includes the thickness and shape parameters of the multiple groups of experimental samples and the corresponding thermal imaging correction effects under different control temperatures, correction strengths, and correction times; A correction effect screening threshold is set, the multi-dimensional experimental data is screened, the sample data set is obtained, deep learning is performed on the sample data set, the parameter control relationship between the casting thickness, the component shape and the control temperature, the correction force, and the correction time is obtained, and the parameter control analysis model is obtained.

[0029] In a feasible implementation, in order to construct a parameter control analysis model, first, the material composition of non-standard mineral castings, such as alloy components, additives, etc., is determined according to the material requirements of non-standard mineral castings, such as strength, hardness, toughness, etc. Then, for different material compositions, according to different casting thicknesses and component shapes, multiple groups of experimental samples are prepared, which should cover the typical thickness range and shape characteristics of non-standard mineral castings. Subsequently, hot forming correction experiments are carried out on the prepared multiple groups of experimental samples. By setting different control temperatures, correction forces and correction times, each group of samples is corrected, and the corresponding experimental data are recorded, including the thickness and shape parameters of the samples (such as size, curvature, etc.), and the correction effects under different process parameters (such as deformation, stress distribution, etc.). Through a large number of experiments, multi-dimensional experimental data can be obtained to fully reflect the relationship between material, geometry, process parameters and correction effects.

[0030] According to the actual production requirements, the correction effect screening threshold is set, such as the allowable range of deformation, the uniformity of stress distribution, etc. Then, the correction effect screening threshold is used to screen the obtained multi-dimensional experimental data, and the data whose correction effect does not meet the correction effect screening threshold is eliminated to obtain a high-quality sample data set. After that, by constructing neural network models, such as convolutional neural networks, recursive neural networks, etc., deep learning is performed on the sample data set, and the casting thickness and component shape are used as inputs, and the process parameters such as control temperature, correction force, and correction time are used as outputs to train the network model. After multiple rounds of iterative optimization, the network model can learn the complex mapping relationship between the geometric characteristics of the casting and the hot forming correction process parameters, forming a parameter control analysis model, which provides an important basis for determining the control parameters of the segmented area of ​​the casting, and helps to improve the correction efficiency and quality.

[0031] Furthermore, the embodiment of the present application also includes: Based on the design requirements of the non-standard mineral castings and the control parameters of each casting segmentation area, a local abnormal case set is obtained, where the abnormal case set is a non-standard mineral casting case with overcorrection or undercorrection; Perform abnormal parameter tracing on the abnormal casting features of the abnormal case set, determine the abnormal control parameters, and use the abnormal control parameters as sensitive adjustment parameters; The probability of similar anomalies in the abnormal case set is calculated, converted into a control coefficient according to the abnormal probability, and the abnormal case set is used as the control weight of the sensitive adjustment parameter.

[0032] In a preferred embodiment, first, according to the design requirements of non-standard mineral castings and the control parameters of each casting segmentation area obtained, abnormal cases are retrieved and collected in the local production history data, that is, non-standard mineral castings that have been overcorrected (such as excessive deformation, stress concentration, etc.) or undercorrected (such as uneven deformation, excessive residual stress, etc.) in the past production process. By collecting and sorting these abnormal cases, an abnormal case set is formed to provide important data support for subsequent abnormal analysis and parameter optimization. Subsequently, for each abnormal casting in the abnormal case set, its geometric features, material properties, segmentation area and other information are analyzed, and the control parameters that produce the abnormality, such as control temperature, correction force, correction time, etc., are traced. By comparing the differences between abnormal castings and normal castings, as well as the deviation of the correction effect of abnormal castings, the key control parameters that have the greatest impact on the correction quality are identified, marked as abnormal control parameters, and the abnormal control parameters are used as sensitive adjustment parameters in the hot forming correction process. In the subsequent correction control optimization, these parameters need to be focused on and adjusted to reduce the risk of abnormality and improve the stability and reliability of correction.

[0033] Next, the abnormal case set is classified and statistically analyzed. The abnormal cases are divided into several similar abnormal groups according to the sensitive adjustment parameters. The probability of each similar abnormal group appearing in the abnormal case set is calculated as the abnormal occurrence probability of the sensitive adjustment parameter, and the similar abnormal probability is obtained. After that, the abnormal probability is converted into a control coefficient. The higher the abnormal probability, the greater the impact of this type of abnormality on the correction. A higher weight needs to be given during the correction control optimization as the control weight of the sensitive adjustment parameter. In the subsequent correction control optimization, for each casting segmentation area, the original control parameters and the sensitive adjustment parameters and their control weights provided by the abnormal case set are comprehensively considered, and the control parameters are dynamically adjusted and optimized. The larger the control weight of the sensitive adjustment parameter, the higher the adjustment range and priority of the sensitive adjustment parameter, so as to avoid the occurrence of abnormal situations as much as possible.

[0034] Furthermore, the embodiment of the present application also includes: According to the casting segmentation area, a partition topology structure is constructed, and a mapping relationship between the control parameters of each casting segmentation area and the corresponding topological node is established; Based on the partition topology, the influence relationship between the parameters of adjacent partition areas is analyzed through abnormal case data to determine the dependent influence relationship between the control parameters of each casting partition area; According to the dependent influence relationship between the control parameters of each casting segmentation area, an optimization space is established, and the sensitive adjustment parameters and their control coefficients are added to the optimization space as constraints to perform global optimization and obtain the hot forming correction control information; According to the thermoforming correction control information, the correction equipment of each divided area is allocated, and the generated equipment control execution is sent to the correction equipment controller.

[0035] In a feasible implementation, first, a partition topology is constructed according to the obtained casting segmentation results. The partition topology is a graphical representation method, in which nodes are used to represent each casting segmentation area, and edges are used to represent the connection relationship between adjacent areas. The partition topology can intuitively describe the partition layout and adjacency relationship of the casting. Then, a mapping relationship is established between the control parameters (control temperature, correction force, correction time) of each casting segmentation area and the corresponding topological nodes in the partition topology, and each topological node is attached with the control parameters of its corresponding casting segmentation area to form a partition control parameter topological network. Subsequently, the influence relationship between the control parameters of adjacent segmentation areas is analyzed using the collected abnormal case data. For example, if the control temperature of one area is abnormally high, it may cause the temperature of the adjacent area to rise abnormally, thereby affecting its correction effect; for example, if the correction force of one area is too large, it may cause additional stress and deformation to the adjacent area, resulting in uneven correction. By systematically analyzing the abnormal case data, the dependent influence relationship between the control parameters of each casting segmentation area can be summarized, and the interaction and constraint conditions between adjacent segmentation areas can be described, providing an important basis for correction control.

[0036] Then, based on the obtained dependency influence relationship, an optimization space is established. Each dimension of the optimization space corresponds to a control parameter of a casting segmentation area, and the dependency influence relationship between the control parameters of different casting segmentation areas forms a constraint condition in the optimization space. Afterwards, the identified sensitive adjustment parameters and their control coefficients are also added to the optimization space as constraints. Among them, the sensitive adjustment parameters are parameters that have a greater impact on the correction, and their value range and adjustment priority need to be used as important optimization constraints. After establishing a complete optimization space, optimization algorithms such as genetic algorithms, particle swarm algorithms, and simulated annealing algorithms are used to perform global optimization searches. Under the premise of meeting various constraints, the optimal control parameter combination is found to achieve the best correction effect of the entire casting, while minimizing the risk of abnormal situations. The result of the optimization is the control information of the hot forming correction, including the specific setting values ​​of the control parameters of each casting segmentation area. After that, according to the obtained hot forming correction control information, the correction equipment of each casting segmentation area is parameterized and task assigned. According to the control parameter requirements of each casting segmentation area, the appropriate heating device, pressure device, etc. are selected, and the corresponding control temperature, correction force, and correction time are input into the correction equipment controller to guide the equipment to perform accurate partition correction operations according to the hot forming correction control information. During the entire partition correction process, the deformation state of the casting is monitored and fed back in real time, and the control parameters of the equipment are dynamically adjusted according to the monitoring data to ensure that the correction effect meets expectations, and to respond to possible abnormal situations in a timely manner to avoid over-correction or under-correction.

[0037] Furthermore, the embodiment of the present application also includes: Based on the dependence and influence relationship between the control parameters of each segmented area of ​​the casting, construct the evaluation function of the control parameters of each segmented area, merge the evaluation functions of the control parameters of all segmented areas to obtain a comprehensive evaluation function, and construct the optimization space based on the comprehensive evaluation function and taking the maximum target result as the optimization goal; The sensitive adjustment parameters and control coefficients of each segmented area are added to the optimization space as optimization constraints to perform global iterative optimization, including randomly extracting control parameter schemes based on the control parameters of each segmented area and satisfying the constraints, evaluating using the comprehensive evaluation function to obtain the current optimal solution, then randomly extracting control parameter schemes and comparing the evaluation values ​​with the current optimal solution, replacing the current optimal solution with the one with the larger evaluation value, and repeating until convergence requirement or number of iterations are reached.

[0038] In a preferred embodiment, when performing global optimization, first, based on the obtained dependence and influence relationship between the control parameters of each casting segmentation area, a control parameter evaluation function is constructed for each casting segmentation area to measure the degree of influence of the control parameter setting of the area on the correction effect, for example, considering factors such as temperature, force, time, etc., and combining the geometric characteristics and material properties of the area for comprehensive evaluation. Then, the control parameter evaluation functions of all casting segmentation areas are merged to obtain a comprehensive evaluation function, which reflects the overall effect of the control parameter settings of each area of ​​the entire casting and evaluates the influence of a set of control parameters on the correction of the entire casting. Next, the optimization space is constructed with the maximization of the result of the comprehensive evaluation function as the optimization goal. Each dimension of the optimization space corresponds to a control parameter of a segmentation area. The dependence relationship between different areas and parameters and the constraints of sensitive adjustment parameters together constitute the boundaries and constraints of the optimization space.

[0039] Subsequently, the obtained sensitive adjustment parameters and control coefficients of each casting segmentation area are used as optimization constraints and added to the constructed optimization space, which limits the value range and adjustment priority of the sensitive adjustment parameters, ensuring that the optimization process can focus on optimizing the sensitive adjustment parameters and reduce abnormal risks. After that, a global iterative optimization is performed in the optimization space. On the premise of meeting various constraints, a set of control parameter schemes is randomly generated, that is, the control parameters of each casting segmentation area are randomly selected, and the set of control parameters are substituted into the comprehensive evaluation function to calculate its evaluation value as the current optimal solution. A new set of control parameter schemes is randomly generated again, and the evaluation value is substituted into the comprehensive evaluation function to calculate its evaluation value; the evaluation value of the new scheme is compared with the evaluation value of the current optimal scheme. If the evaluation value of the new scheme is larger, it is replaced with the new current optimal scheme. Repeat the iterative optimization until the convergence requirements are met (such as the evaluation value is greater than the preset evaluation value threshold) or the preset number of iterations is reached. At this point, through continuous random exploration and comparison, the global maximum point of the comprehensive evaluation function is searched in the optimization space, and the overall optimal control parameter combination scheme is obtained as the hot forming correction control information.

[0040] Furthermore, the embodiment of the present application also includes: Obtain control parameters and parameter accuracy of correction equipment; Based on the control coefficient, setting matching priorities for the sensitive adjustment parameters of each segmented area; According to the matching priority, the sensitive adjustment parameters of each segmented area are sequentially used with the control parameters and parameter accuracy of the correction device to perform traversal matching, determine the matching relationship between the segmented area and the correction device, and arrange the position relationship of the correction device; According to the thermoforming correction control information, the control parameters of each segmented area are extracted to generate the control information of the segmented area, and the control information is sent to the corresponding correction device based on the matching relationship.

[0041] In a preferred embodiment, first, collect and organize the information of the currently available correction equipment, and determine the control parameters of the correction equipment, such as the temperature control range and accuracy, the pressure application range and accuracy, the time control accuracy, etc. These parameters are related to whether the equipment can meet the correction control requirements of different segmented areas. At the same time, determine the adjustment accuracy of each control parameter, that is, the minimum adjustment unit and error range of the correction equipment on the parameter, to determine the fine-grainedness and accuracy of the correction equipment's control over the segmented area. Then, using the obtained control coefficient, the importance of the sensitive adjustment parameters of each casting segmentation area is sorted and the priority is set. The larger the control coefficient, the greater the impact of the sensitive adjustment parameter on the correction quality, and a higher priority needs to be given when matching the correction equipment. For example, according to the size of the control coefficient, the sensitive adjustment parameters of each casting segmentation area are divided into different priority levels, such as high, medium, low, etc. The higher the priority level, the more it means that the parameter needs to be given priority when matching the equipment and requires more stringent accuracy requirements.

[0042] Then, according to the set matching priority, the sensitive adjustment parameters of each casting segmentation area are matched one by one with the control parameters and parameter accuracy of the correction equipment. The matching principle is to give priority to the equipment whose control parameter range covers the casting segmentation area parameter requirements and whose parameter accuracy meets the casting segmentation area control requirements. The specific matching process adopts the exhaustive method to traverse all possible equipment combinations, evaluate their matching degree with the casting segmentation area parameter requirements, and select the equipment combination with the highest matching degree as the final matching result. After determining the matching relationship between the segmentation area and the correction equipment, the position of the correction equipment is reasonably arranged according to the spatial position and adjacency relationship of the segmentation area on the casting to achieve efficient and accurate control of the segmentation area. After that, the control parameters corresponding to each segmentation area are extracted from the obtained hot forming correction control information, and the equipment control instructions for each segmentation area are generated according to the control parameters, and the specific control actions and parameter settings that the equipment needs to perform are clearly specified. Subsequently, the control instructions for each segmentation area are sent to the corresponding correction equipment using the established matching relationship between the segmentation area and the correction equipment. After the device receives the control instruction, it performs necessary decoding and analysis to extract key control parameters and execution steps, and converts them into specific execution actions such as heating, pressurizing, and keeping warm based on the hardware characteristics and control logic of the device, thereby achieving precise control of the segmented area.

[0043] Furthermore, the expression of the evaluation function of the control parameters of each segmented area is: ; Among them, K is the control parameter evaluation result, Q is the target control parameter, N is the evaluation control parameter, The influence value depends on the influence parameter x.

[0044] In a preferred embodiment, the control parameter evaluation function for each segmented region is provided as: ; Among them, K is the control parameter evaluation result, which is a dimensionless relative index used to measure the degree of deviation between the control parameters obtained in the optimization process and the optimal control parameters predicted by the model; Q is the target control parameter, that is, the control parameter predicted by the parameter control analysis model; N is the evaluation control parameter, that is, the control parameter obtained in the current iteration during the optimization process. The influence value of the dependent influencing parameter x indicates the influence degree of other related parameters on the evaluation control parameter.

[0045] Therefore, the physical meaning of the control parameter evaluation function of each segmented area is: The K value represents the relative deviation between the current control parameter N and the control parameter Q predicted by the model during the optimization process, while taking into account the influence of other related parameters f(x). When N and Q deviate greatly or When the impact is large, the K value will become smaller, indicating that the optimization effect of the current control parameters is not ideal and there is room for improvement. In the optimization process, the K value is used as the objective function of the optimization, and the goal is to maximize the K value, that is, to minimize the deviation between the current control parameters and the optimal control parameters. At the same time, the K value of each casting segmentation area is set as a constraint condition, requiring it to be no less than a certain threshold to ensure that the control parameters of each casting segmentation area are as close as possible to the control parameters predicted by the model.

[0046] Through the evaluation function of the control parameters of each segmented area, the advantages and disadvantages of the control parameters of each segmented area can be accurately evaluated, so as to quickly and accurately find the optimal control parameter combination, realize the intelligent and automated hot forming correction control, and improve the correction efficiency and quality.

[0047] In summary, the hot forming correction control system for non-standard mineral castings provided in the embodiments of the present application has the following technical effects: Obtain the design requirements of non-standard mineral castings, including design drawings, description parameters, and material requirements. Based on the description parameters, the design drawings are segmented according to the identification difference points. The segmentation nodes are the points with the smallest identification difference, and the segmentation parts are the areas with identification difference. The casting segmentation results are obtained. By finding the points with the smallest difference and the differentiated areas, the reasonable division of castings is achieved, providing support for regional correction control. According to the material requirements, a sample data set is constructed through the material cumulative thickness, component shape, control temperature, correction force, correction time, and multi-dimensional experimental data. Through the sample data set, learning is carried out to obtain the parameter control analysis model, laying the foundation for determining the control parameters of each casting segmentation area. Based on the casting segmentation results, according to the casting shape and description parameters, through the parameter control analysis model, multi-dimensional parameter analysis is performed to determine the control parameters of each casting segmentation area. The control parameters of each casting segmentation area include control temperature, correction force, and correction time. According to the characteristics of each casting segmentation area, the corresponding control temperature, correction force, and correction time are determined to achieve precise control of each area. Based on the control parameters of each casting segmentation area, an abnormal case set is obtained, and the abnormal case set is used to calculate the abnormal probability of each casting segmentation area, which is converted into a control coefficient according to the abnormal probability, and the abnormal parameter is traced to determine the abnormal control parameters to improve the robustness of the correction process. According to the control coefficient, the control parameters of each casting segmentation area, and the abnormal control parameters, a global optimization is performed in combination with the casting segmentation area to obtain the hot forming correction control information, which is sent to the controller for control. While ensuring the correction effect of each area, the correction quality of the entire casting is taken into account, thereby improving the hot forming correction effect of non-standard mineral castings.

[0048] Any steps of the system described above may be stored as computer instructions or programs in an unlimited computer memory, and may be called and recognized by an unlimited computer processor to implement any system in the embodiments of the present application, without any unnecessary restrictions.

[0049] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these modifications and variations.

Claims

1. A hot forming correction control system for non-standard mineral castings, characterized in that: The system comprises: A casting segmentation module, which is used to obtain non-standard mineral casting design requirements, including design drawings, description parameters, and material requirements, and segment the design drawings according to identification difference points based on the description parameters. The segmentation nodes are points with the smallest identification difference, and the segmentation parts are areas with identification difference, so as to obtain casting segmentation results; A model acquisition module, wherein the model acquisition module is used to construct a sample data set according to the material requirements through the material cumulative thickness, component shape, control temperature, correction force, correction time, and multi-dimensional experimental data, and learn through the sample data set to obtain a parameter control analysis model; A control parameter determination module, which is used to perform multi-dimensional parameter analysis based on the casting segmentation result, according to the casting shape and description parameters, through the parameter control analysis model, to determine the control parameters of each casting segmentation area, wherein the control parameters of each casting segmentation area include control temperature, correction force, and correction time; An abnormality tracing module, the abnormality tracing module is used to obtain an abnormal case set based on the control parameters of each casting segmentation area, use the abnormal case set to calculate the abnormal probability of each casting segmentation area, convert it into a control coefficient according to the abnormal probability, and trace the abnormal parameters to determine the abnormal control parameters; The correction control module is used to perform global optimization based on the control coefficient, the control parameters of each casting segmentation area, and the abnormal control parameters in combination with the casting segmentation area to obtain hot forming correction control information and send it to the controller for control.

2. The system according to claim 1, characterized in that The design drawing is segmented according to the identification difference points based on the description parameters, the segmentation nodes are points with the smallest identification difference, and the segmentation parts are areas with identification difference, to obtain the casting segmentation results, including: Locate and extract shape change points and size change points according to the design drawing and the description parameters; Clustering is performed according to the shape change points and the size change points to obtain a plurality of change clustering areas; Based on the multiple change clustering areas, the transition area between the multiple change clustering areas without identifiable difference points is obtained for midline segmentation to obtain the casting segmentation result.

3. The system according to claim 1, characterized in that According to the material requirements, a sample data set is constructed through the material cumulative thickness, component shape, control temperature, correction force, correction time, and multi-dimensional experimental data. The parameter control analysis model is obtained through learning through the sample data set, including: According to the material requirements, the material composition is determined, and multiple groups of experimental samples are shaped according to the material casting thickness and component shape; In this way, the multiple groups of experimental samples are controlled by controlling the temperature, correction strength, and correction time, and the experimental data are recorded to construct the multi-dimensional experimental data, which includes the thickness and shape parameters of the multiple groups of experimental samples and the corresponding thermal imaging correction effects under different control temperatures, correction strengths, and correction times; A correction effect screening threshold is set, the multi-dimensional experimental data is screened, the sample data set is obtained, deep learning is performed on the sample data set, the parameter control relationship between the casting thickness, the component shape and the control temperature, the correction force, and the correction time is obtained, and the parameter control analysis model is obtained.

4. The system according to claim 1, characterized in that Based on the control parameters of each casting segmentation area, an abnormal case set is obtained, the abnormal probability of each casting segmentation area is calculated using the abnormal case set, the abnormal probability is converted into a control coefficient, and abnormal parameter tracing is performed to determine the abnormal control parameter, including: Based on the design requirements of the non-standard mineral castings and the control parameters of each casting segmentation area, a local abnormal case set is obtained, where the abnormal case set is a non-standard mineral casting case with overcorrection or undercorrection; Perform abnormal parameter tracing on the abnormal casting features of the abnormal case set, determine the abnormal control parameters, and use the abnormal control parameters as sensitive adjustment parameters; The probability of similar anomalies in the abnormal case set is calculated, converted into a control coefficient according to the abnormal probability, and the abnormal case set is used as the control weight of the sensitive adjustment parameter.

5. The system according to claim 4, characterized in that According to the control coefficient, the control parameters of each casting segmentation area, and the abnormal control parameters, global optimization is performed in combination with the casting segmentation area to obtain hot forming correction control information, and the information is sent to the controller for control, including: According to the casting segmentation area, a partition topology structure is constructed, and a mapping relationship between the control parameters of each casting segmentation area and the corresponding topological node is established; Based on the partition topology, the influence relationship between the parameters of adjacent partition areas is analyzed through abnormal case data to determine the dependent influence relationship between the control parameters of each casting partition area; According to the dependent influence relationship between the control parameters of each casting segmentation area, an optimization space is established, and the sensitive adjustment parameters and their control coefficients are added to the optimization space as constraints to perform global optimization and obtain the hot forming correction control information; According to the thermoforming correction control information, the correction equipment of each divided area is allocated, and the generated equipment control execution is sent to the correction equipment controller.

6. The system according to claim 5, characterized in that According to the dependent influence relationship between the control parameters of each casting segmentation area, an optimization space is established, and the sensitive adjustment parameters and their control coefficients are added to the optimization space as constraints to perform global optimization, including: Based on the dependence and influence relationship between the control parameters of each segmented area of ​​the casting, construct an evaluation function of the control parameters of each segmented area, merge the evaluation functions of the control parameters of all segmented areas to obtain a comprehensive evaluation function, and construct the optimization space based on the comprehensive evaluation function and taking the maximum target result as the optimization goal; The sensitive adjustment parameters and control coefficients of each segmented area are added to the optimization space as optimization constraints to perform global iterative optimization, including randomly extracting control parameter schemes based on the control parameters of each segmented area and satisfying the constraints, evaluating using the comprehensive evaluation function to obtain the current optimal solution, then randomly extracting control parameter schemes and comparing the evaluation values ​​with the current optimal solution, replacing the current optimal solution with the one with the larger evaluation value, and repeating until convergence requirement or number of iterations are reached.

7. The system according to claim 5, characterized in that Allocating correction equipment to each segmented area according to the thermoforming correction control information includes: Obtain control parameters and parameter accuracy of correction equipment; Based on the control coefficient, setting matching priorities for the sensitive adjustment parameters of each segmented area; According to the matching priority, the sensitive adjustment parameters of each segmented area are sequentially used with the control parameters and parameter accuracy of the correction device to perform traversal matching, determine the matching relationship between the segmented area and the correction device, and arrange the position relationship of the correction device; According to the thermoforming correction control information, the control parameters of each segmented area are extracted to generate the control information of the segmented area, and the control information is sent to the corresponding correction device based on the matching relationship.

8. The system according to claim 6, characterized in that The expression of the evaluation function of the control parameters of each segmented area is: ； Among them, K is the control parameter evaluation result, Q is the target control parameter, N is the evaluation control parameter, The influence value depends on the influence parameter x.

Images