A multi-process collaborative aluminum processing parameter tuning method and system

Abstract

The present application relates to the technical field of process parameter optimization, and discloses a method and system for tuning process parameters of aluminum processing with multi-process collaboration, including: constructing a process map by analyzing the correlation between multiple processes of aluminum processing; dynamically updating the priority of each process according to the process map through a preset process competition mechanism to obtain a priority order; analyzing the influence weight of the process parameters corresponding to each process based on the priority order, layering the process parameters to obtain multi-layer optimization level process parameters; combining the process map and the multi-layer optimization level process parameters, layering the process parameters of each process through a preset parameter optimization model to obtain optimized process parameters, so as to achieve tuning of the aluminum processing process parameters; the present application can improve the pertinence and efficiency of multi-process process parameter optimization, and improve the quality and stability of aluminum processing products through precise adjustment of process parameters.

Description

Technical Field

[0001] The present invention relates to the technical field of process parameter optimization, and in particular to a method and system for tuning process parameters of aluminum processing in a coordinated manner across multiple processes. Background Art

[0002] During the aluminum processing process, optimizing process parameters can improve product quality and production efficiency, and enhance the company's market competitiveness. At the same time, using optimized process parameters for aluminum processing can reduce the consumption of raw materials and energy waste, reduce the impact on the environment, and meet the requirements of sustainable development.

[0003] The existing technology has the following problems: lack of analysis of the correlation between multiple processes in aluminum processing, and failure to fully consider the causal relationship between processes, which makes it impossible to start from the perspective of the overall process flow when optimizing process parameters, easily leading to local optimization and ignoring the overall effect; using fixed process priorities, which cannot be dynamically updated according to actual production conditions, resulting in insufficient attention to key processes, affecting product quality and production efficiency; using a single process parameter optimization method, ignoring the different degrees of influence of different parameters on product quality and production efficiency, resulting in a lack of pertinence and low efficiency in the optimization process; in order to solve at least one of the above problems, the present application proposes a multi-process collaborative aluminum processing process parameter tuning method and system. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the main purpose of the present invention is to provide a method and system for optimizing aluminum processing parameters in a multi-process collaborative manner, which can effectively solve the problems in the background technology. The specific technical solutions of the present invention are as follows:

[0005] A multi-process collaborative aluminum processing parameter tuning method, comprising:

[0006] By analyzing the correlation between multiple processes of aluminum processing, a process map is constructed;

[0007] According to the process map, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order;

[0008] Based on the priority order, the influence weight of the process parameters corresponding to each process step is analyzed, the process parameters are layered, and multi-layer optimization level process parameters are obtained;

[0009] Combining the process map and the multi-layer optimization level process parameters, the process parameters of each process are optimized in layers through a preset parameter optimization model to obtain optimized process parameters, so as to achieve the tuning of aluminum processing process parameters.

[0010] Specifically, the process map is constructed by analyzing the correlation between multiple processes of aluminum processing, including:

[0011] Based on the pre-acquired aluminum processing data, the feature set corresponding to each process is extracted;

[0012] For each process, multiple features in the feature set are fused to obtain a feature vector;

[0013] The causal relationship between processes is analyzed according to the feature vector, and a process map is constructed, wherein the processes are used as nodes in the process map, and the causal relationship between processes is used as the weight of the connecting edges between the nodes.

[0014] Specifically, the causal relationship between processes is analyzed based on the feature vectors to construct a process map, wherein the processes are used as nodes in the process map, and the causal relationship between processes is used as the weight of the edges connecting the nodes, including:

[0015] Analyze the causal relationship between each process based on the characteristic vector and obtain the first causal direction;

[0016] Calculate the causal relationship strength between processes through the preset causal strength calculation model;

[0017] Taking the processes as nodes and the causal relationship strength between processes as edge weights, directed edges from cause nodes to effect nodes are established between process nodes to obtain a process graph.

[0018] Specifically, according to the process map, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order, including:

[0019] According to the process map, calculate the contribution weight of each process to the aluminum processing;

[0020] Based on the contribution weight, each process is assigned corresponding voting rights, and the processes vote to obtain voting results;

[0021] According to the voting results, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order.

[0022] Specifically, according to the voting results, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order, including:

[0023] According to the voting results, the priority of each process is calculated and sorted from high to low to obtain the first priority order;

[0024] Calculate the priority adjustment of each process through a sliding window within a preset time period;

[0025] Adjusting the first priority order according to the priority adjustment amount to obtain a second priority order;

[0026] By analyzing the process parameter adjustment directions in the first priority order and the second priority order, process competition is performed between processes with inconsistent parameter adjustment directions to obtain the priority order.

[0027] Specifically, based on the priority order, the influence weight of the process parameters corresponding to each process step is analyzed, and the process parameters are layered to obtain multi-layer optimization level process parameters, including:

[0028] Based on the priority order and the production data of each process, the influence weight of each process parameter is calculated through the preset parameter impact analysis model;

[0029] According to the influence weights, the process parameters of each process step are layered using a preset parameter layering standard to obtain multi-layer optimized process parameters.

[0030] Specifically, combining the process map and the multi-level optimization level process parameters, the process parameters of each process are optimized in layers through a preset parameter optimization model to obtain optimized process parameters to achieve the tuning of aluminum processing process parameters, including:

[0031] Combining the process map and multi-level optimization level process parameters, the parameter feature set of each optimization level process parameter is extracted;

[0032] According to the parameter feature set, each optimization level process parameter is optimized respectively by using a preset parameter optimization model to obtain a first optimized process parameter;

[0033] By analyzing the logical association of process parameters of different optimization levels, the first optimized process parameters are optimized to obtain optimized process parameters, so as to achieve the optimization of aluminum processing process parameters.

[0034] Specifically, according to the parameter feature set, each optimization level process parameter is optimized respectively through a preset parameter optimization model to obtain a first optimized process parameter, wherein the optimization level includes a core layer, an important layer, and a general layer, and the parameter optimization model includes a first parameter optimization model, a second parameter optimization model, and a third parameter optimization model, including:

[0035] For the core layer process parameters, according to the corresponding core layer process parameter feature set, a preset first parameter optimization model is used to perform real-time optimization to obtain first optimized process parameters of the core layer process parameters;

[0036] For the important layer process parameters, according to the corresponding important layer process parameter feature set, a preset second parameter optimization model is used to dynamically adjust the important layer process parameters to obtain the first optimized process parameters of the important layer process parameters;

[0037] For the general layer process parameters, batch optimization is performed through a preset third parameter optimization model according to the corresponding general layer process parameter feature set to obtain first optimized process parameters of the general layer process parameters.

[0038] Specifically, the first optimized process parameters are optimized by analyzing the logical association of process parameters of different optimization levels to obtain optimized process parameters to achieve the tuning of aluminum processing process parameters, including:

[0039] During the process of process parameter optimization, conflicting process parameters are identified and a conflict parameter set is obtained;

[0040] According to the conflict parameter set, the conflict priority of each parameter in the conflict parameter set is calculated by analyzing the priority of each parameter corresponding to the process, the parameter influence weight and the logical association between the parameters;

[0041] Through the preset parameter optimization competition mechanism, the parameters with high conflicting priorities are optimized first, until the optimization of each conflicting parameter is completed, and the optimized process parameters are obtained to achieve the tuning of the aluminum processing process parameters.

[0042] A multi-process collaborative aluminum processing parameter tuning system, used to implement the multi-process collaborative aluminum processing parameter tuning method, comprising:

[0043] The process map construction module constructs a process map by analyzing the correlation between multiple processes of aluminum processing;

[0044] The process priority analysis module dynamically updates the priority of each process based on the process map and a preset process competition mechanism to obtain a priority order;

[0045] A process parameter stratification module analyzes the influence weight of the process parameters corresponding to each process step based on the priority order, stratifies the process parameters, and obtains multi-layer optimized process parameters;

[0046] The process parameter optimization module combines the process map and the multi-level optimization level process parameters, and optimizes the process parameters of each process in layers through a preset parameter optimization model to obtain optimized process parameters to achieve the tuning of aluminum processing process parameters.

[0047] Compared with the prior art, this application has the following beneficial effects:

[0048] This application constructs a process map by analyzing the correlation between multiple processes in aluminum processing, dynamically updates the priority of each process in combination with the process competition mechanism, and optimizes the process parameters in layers, so that the optimization process can be flexibly adjusted according to actual production conditions. Different optimization strategies are adopted for parameters at different levels to improve the targetedness and efficiency of optimization. Through precise adjustment of process parameters, product defects caused by unreasonable parameters can be reduced and the quality and stability of production products can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a workflow diagram of a multi-process collaborative aluminum processing parameter tuning method in Example 1 of the present invention;

[0050] Figure 2 This is a schematic diagram of the process map constructed in Example 1 of the present invention;

[0051] Figure 3 Schematic diagram of hierarchical optimization of process parameters in Example 1 of the present invention;

[0052] Figure 4 This is a structural diagram of a multi-process collaborative aluminum processing parameter tuning system in Example 2 of the present invention. DETAILED DESCRIPTION

[0053] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0054] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0055] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0056] Example 1

[0057] This embodiment provides a method for optimizing aluminum processing parameters in a multi-process collaborative manner. Figure 1 As shown, the multi-process collaborative aluminum processing parameter tuning method includes:

[0058] S101. Construct a process map by analyzing the correlation between multiple processes of aluminum processing;

[0059] S102. According to the process map, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order;

[0060] S103, based on the priority order, analyzing the influence weight of the process parameters corresponding to each process step, stratifying the process parameters, and obtaining multi-layer optimized process parameters;

[0061] S104. Combining the process map and the multi-level optimization level process parameters, the process parameters of each process are optimized in layers through a preset parameter optimization model to obtain optimized process parameters, so as to achieve the optimization of the aluminum processing process parameters.

[0062] This embodiment analyzes the correlation between multiple processes in the aluminum processing process, constructs a process map based on the corresponding correlation, votes between each process in the process map, calculates the priority of the process, and optimizes the process parameters in each process in a hierarchical manner by analyzing the influence weights of the process parameters. Compared with the traditional method of directly optimizing process parameters with a single optimization model, this application can reasonably allocate production resources, give priority to optimizing important parameters in key processes, reduce unnecessary waste of production time, and improve overall production efficiency. At the same time, during the aluminum processing process, the process priority and process parameters are dynamically adjusted, and the process changes in the aluminum processing process are quickly responded to to adapt to the actual processing process.

[0063] In this embodiment, multiple processes in the aluminum processing process are analyzed, and a process map is constructed based on the correlation between each process. The aluminum processing process has multiple processes, including smelting, casting, rolling, heat treatment, etc. The logical relationship, sequence, data transfer and other related information between the processes are analyzed. For example, the output of the smelting process is the input of the casting process, and the process parameters of the rolling process will affect the effect of the subsequent heat treatment process, etc. The sequence, data transfer direction and degree of mutual influence between the processes are obtained. According to the analyzed correlation results, the processes are used as nodes, and the correlation between the processes are used as directed edges to construct a corresponding process map. By constructing the process map, the relationship between the processes in the aluminum processing process can be intuitively seen, thereby analyzing the process priority.

[0064] Specifically, according to the constructed process map, the priority of each process is analyzed, and the priority of each process is adjusted in real time during the aluminum processing process, which can ensure that during the aluminum processing process, priority is given to the processes that have a greater impact on the quality of the final product and production efficiency, and respond to changes in the aluminum processing process; for each process, the degree of influence of the process parameters on product quality and production efficiency is analyzed, and the influence weight of each process parameter in each process is obtained. The process parameters are stratified according to the influence weight and divided into different optimization levels. When tuning the parameters, the process parameters that have a greater impact on the results can be optimized first, thereby improving the tuning efficiency.

[0065] Specifically, by combining the correlation between processes in the process map and the multi-level optimization level process parameters, the process parameters of each process are optimized separately through the preset parameter optimization model. Different optimization methods are used for process parameters with different optimization levels. This can avoid the waste of computing resources caused by using a single parameter optimization method, achieve targeted optimization of process parameters, improve parameter optimization efficiency, and thus improve product quality and production efficiency.

[0066] This application constructs a process map by analyzing the correlation between multiple processes in aluminum processing, dynamically updates the priority of each process in combination with the process competition mechanism, and optimizes the process parameters in layers, so that the optimization process can be flexibly adjusted according to actual production conditions. Different optimization strategies are adopted for parameters at different levels to improve the targetedness and efficiency of optimization. Through precise adjustment of process parameters, product defects caused by unreasonable parameters can be reduced and the quality and stability of production products can be improved.

[0067] Furthermore, the process map is constructed by analyzing the correlation between multiple processes of aluminum processing, including:

[0068] S201, extracting a feature set corresponding to each process based on pre-acquired aluminum processing data;

[0069] S202: For each process, multiple features in the feature set are fused to obtain a feature vector;

[0070] S203. Analyze the causal relationship between processes based on the feature vectors and construct a process map, wherein the processes are used as nodes in the process map, and the causal relationship between processes is used as the weight of the connecting edges between the nodes.

[0071] In this embodiment, the differences in raw materials, equipment, process parameters, output products, etc. between different processes are analyzed from a large amount of aluminum processing data, and a feature set corresponding to each process is extracted. During the aluminum processing process, processing data of the aluminum processing process is collected using sensors, production management systems, etc., including but not limited to raw material composition data, equipment operation data, process parameter data, product quality inspection data, etc.; wherein the raw material composition data includes the alloy element content in the aluminum ingot, the equipment operation data includes the melting furnace temperature, the rolling mill speed, etc., the process parameter data includes the casting time, the heat treatment holding time, etc., and the product quality inspection data includes the hardness and dimensional deviation of the finished product; the collected data is preprocessed by data cleaning and standardization, and the features corresponding to each process are extracted based on the preprocessed processing data. For example, for the smelting process, features such as the smelting temperature range, smelting time, and raw material ratio are extracted; for the rolling process, features such as the rolling pass number, rolling speed, and reduction ratio are extracted to obtain a feature set corresponding to each process; through feature extraction, key information is screened from the aluminum processing data, data redundancy is reduced, and the correlation between each process is analyzed.

[0072] Specifically, the feature sets corresponding to each extracted process are fused to obtain the feature vector corresponding to each process, which can integrate the scattered multi-angle feature information to obtain comprehensive feature information. Feature fusion methods include weighted summation method, series connection method, etc. In this implementation, each eigenvalue is connected in sequence to form a feature vector through the series connection method; through feature fusion, the feature vector corresponding to each process is obtained, and multiple scattered features are integrated into a unified feature vector to obtain comprehensive feature information.

[0073] like Figure 2 As shown in the figure, the processes are taken as nodes, and the degree of causal relationship between processes is used as the weight of the connecting edge between nodes to construct a process map; the degree of causal relationship between each process is calculated according to the characteristic vector, and the characteristic vector of each process is analyzed by the machine learning algorithm to obtain the corresponding causal relationship strength as the weight of the connecting edge between the corresponding nodes. The larger the weight, the tighter the causal relationship between the two processes, and the greater the impact of the change of one process on the other process; by constructing the process map, the collaborative relationship between the various processes in the aluminum processing process can be grasped as a whole, providing a corresponding basis for the optimization of process parameters.

[0074] Furthermore, the causal relationship between the processes is analyzed based on the feature vectors to construct a process map, wherein the processes are used as nodes in the process map, and the causal relationship between the processes is used as the weight of the edges connecting the nodes, including:

[0075] S301, analyzing the causal relationship between each process according to the characteristic vector to obtain a first causal direction;

[0076] S302. Calculate the causal relationship strength between the processes using a preset causal strength calculation model;

[0077] S303. Take the processes as nodes and the causal relationship strength between processes as edge weights, establish directed edges from cause nodes to effect nodes between process nodes, and obtain a process graph.

[0078] In this embodiment, the causal relationship between each process is analyzed based on the characteristic vector of each process, and it is determined which process is the cause node and which process is the effect node, so as to obtain the first causal direction; specific causal relationship analysis methods include Granger causality test method, causal discovery algorithm based on machine learning, etc. This embodiment adopts a causal Bayesian network model, and trains the causal Bayesian network model through a large amount of historical data to obtain a pre-trained causal Bayesian network model. The probabilistic dependency relationship between the processes is analyzed according to the pre-trained causal Bayesian network model, and the corresponding causal structure is output to obtain the first causal direction; by analyzing the causal logic between the processes, it is possible to avoid misunderstanding the process relationship when constructing the map, ensure that the map can truly reflect the actual action relationship between the processes in the aluminum processing process, and provide corresponding logical guidance for process optimization and process parameter adjustment.

[0079] Specifically, according to the calculated first causal direction, the influence degree between different processes is analyzed by a preset causal strength calculation model, and the causal correlation strength between the corresponding processes is calculated; the causal strength calculation model includes a regression analysis model, a deep learning model, etc. This embodiment adopts a convolutional neural network model, and uses a large amount of historical feature vector data to train the convolutional neural network model to obtain a pre-trained convolutional neural network model. The feature vector of the corresponding process is input into the pre-trained convolutional neural network model, and the model outputs the causal correlation strength between every two processes with a causal relationship; for example, for the smelting process and the casting process, the feature vectors of the two processes are input, and the model outputs a numerical value representing the causal correlation strength between the two processes; by calculating the causal correlation strength between the processes, the causal relationship between the processes is quantified, the importance of the influence between the processes is understood, and data support is provided for the optimization of process parameters.

[0080] Specifically, after determining the causal direction and causal correlation strength between each process, the process is used as a node and the causal correlation strength is used as the weight of the edge. A directed edge from the cause node to the effect node is established between the process nodes to construct a process map. By constructing a process map, the causal relationship and influence degree between each process in the aluminum processing process are quantified, thereby identifying the key processes in multiple aluminum processing processes, providing a basis for optimizing process parameters, improving production efficiency and product quality. For example, the map can quickly find the upstream process that has the greatest impact on a certain process, or which downstream processes a certain process will have an important impact on.

[0081] Furthermore, according to the process map, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order, including:

[0082] S401. Calculate the contribution weight of each process to the aluminum processing process according to the process map;

[0083] S402: Based on the contribution weight, a corresponding voting right is assigned to each process, and the processes vote to obtain a voting result;

[0084] S403: According to the voting results, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order.

[0085] In this embodiment, the importance of each process to the entire aluminum processing process is analyzed according to the process map, and the contribution weight of the corresponding process to the aluminum processing process is calculated; first, in combination with the characteristics of the aluminum processing technology, evaluation indicators for evaluating the contribution of the process are set, including the degree of influence on the key quality indicators of the product, the proportion of process time in the total production cycle, the proportion of resource consumption, the complexity of the process, etc., among which the key quality indicators of the product include strength, hardness, dimensional accuracy, etc.; the contribution weight analysis method includes the hierarchical analysis method, the Delphi method and other methods, and the hierarchical analysis method is used in this embodiment to assign corresponding weights to each evaluation indicator; a weighted calculation is performed based on the data of each process on the evaluation indicator and the weight of the corresponding evaluation indicator to obtain the contribution weight of each process to the aluminum processing process; by calculating the contribution weight of each process, the process contribution is comprehensively evaluated from multiple dimensions to avoid the one-sidedness of the single indicator evaluation and obtain an accurate process importance evaluation result.

[0086] Specifically, voting rights are allocated to each process based on the calculated contribution weight. The larger the contribution weight of the process, the higher the corresponding voting rights, which means that the process has greater say in determining the process priority. The processes vote with each other to obtain the corresponding voting results. Corresponding voting rights are allocated to each process based on the contribution weight. For example, the contribution weight of process A is 73, and the voting rights allocated according to the rules are 73×10=730 votes. In the voting process, each process allocates voting rights to other processes based on factors such as its own relationship with other processes and its understanding of the overall production. For example, process B believes that process C has a greater impact on its aluminum processing production and casts more votes for process C. The total number of votes obtained by each process is counted to obtain the voting results. Voting between processes can comprehensively reflect the relationship and influence between processes, comprehensively reflect the importance of processes in the entire production system, and avoid misjudgment of priorities from a single perspective.

[0087] Specifically, the priority of each process is analyzed based on the voting results. At the same time, during the aluminum processing process, the priority of each process is dynamically updated through the preset process competition mechanism to obtain a process priority sequence that meets the current aluminum processing production status. By dynamically updating the process priority, it is possible to quickly respond to changes in the production process, adjust the process priority in time, and ensure smooth production.

[0088] Furthermore, based on the voting results, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order, including:

[0089] S501. Calculate the priority of each process based on the voting results, and sort them from high to low to obtain a first priority order;

[0090] S502. Calculate the priority adjustment amount of each process through a sliding window within a preset time period;

[0091] S503: Adjust the first priority order according to the priority adjustment amount to obtain a second priority order;

[0092] S504 , by analyzing the process parameter adjustment directions in the first priority order and the second priority order, process competition is performed between processes with inconsistent parameter adjustment directions to obtain a priority order.

[0093] In this embodiment, the priority of each process is calculated based on the voting results to determine the preliminary priority order among the processes; the total number of votes obtained by each process is divided by the total number of votes for all processes to calculate the corresponding priority score, and each process is sorted from high to low according to its priority score to obtain the first priority order. By calculating the priority score, the priorities among the processes are sorted to obtain a determined order of process importance, providing a framework for priority adjustment.

[0094] Specifically, within a preset time period, the corresponding priority adjustment amount is calculated through a sliding window, and the priority adjustment amount of each process in different time periods is analyzed; the size of the sliding window and the sliding step are set according to the actual calculation accuracy requirements and the priority update process, wherein the size of the sliding window is the number of time periods contained in the window, and the sliding step is the number of time periods each time the window moves; for example, the sliding window size is set to 3 time periods, the sliding step is 1 time period, and the preset time period is 1 hour, then the window will cover the 1st to 3rd hours, 2nd to 4th hours, 3rd to 5th hours and other time periods in turn; within each sliding window, the priority score of each process in the time period is recalculated, and the corresponding priority adjustment amount is obtained by comparing the changes in the priority score. By calculating the priority adjustment amount, the priority of the corresponding process can be updated in time.

[0095] Specifically, according to the calculated priority adjustment amount, the first priority order is adjusted to obtain a second priority order that can better reflect the current importance of the process, the priority adjustment amount of each process is combined with the original priority score in the first priority order, the comprehensive priority score of each process is recalculated, and all processes are reordered from high to low according to the comprehensive priority score to obtain the second priority order; by adjusting the priority order through the priority adjustment amount, the allocation of production resources can be optimized, ensuring that resources flow preferentially to the processes that currently have a greater impact on production, reducing resource waste, and improving overall production efficiency.

[0096] Specifically, the process parameter adjustment directions in the first priority sequence and the second priority sequence are analyzed, and the processes with inconsistent process parameter adjustment directions are adjusted. Processes with different parameter adjustment directions compete with each other to obtain an updated priority sequence, thereby avoiding production chaos caused by conflicting parameter adjustment directions; the process parameter adjustment directions of each process in the first priority sequence and the second priority sequence are compared to find process combinations with inconsistent adjustment directions; scoring criteria are set according to factors such as the contribution weight of the process, the priority adjustment amount, and the degree of influence on product quality, and the process combinations with inconsistent adjustment directions are scored, and the final priority of each process is determined based on the scoring results; for example, the parameter adjustment directions of processes B and C are inconsistent. After scoring, process B scores 85 points and process C scores 78 points, then the final priority of process B is higher than that of process C; by making processes with inconsistent process parameter adjustment directions compete with each other, the process conflict problem caused by inconsistent process parameter adjustment directions can be effectively solved, chaos and waste of resources in the production process can be avoided, and the orderly progress of the production process can be guaranteed.

[0097] Furthermore, based on the priority order, the influence weight of the process parameters corresponding to each process step is analyzed, and the process parameters are layered to obtain multi-layer optimization level process parameters, including:

[0098] S601. Based on the priority order and the production data of each process, the influence weight of each process parameter is calculated using a preset parameter impact analysis model;

[0099] S602: Based on the influence weights, the process parameters of each process step are layered using a preset parameter layering standard to obtain multi-layer optimized process parameters.

[0100] In this embodiment, according to the priority order and production data of each process, the relationship between the process parameters and the aluminum processing production results is analyzed through a preset parameter impact analysis model, and the influence weight of each process parameter on the aluminum processing production process and product quality is quantified; the parameter impact analysis model can be a multiple linear regression model, a random forest model, a neural network model, etc. The parameter impact analysis model in this embodiment is a random forest model. The random forest model is trained with a large amount of production data to obtain a pre-trained random forest model. The production data of each process is input into the pre-trained random forest model, and the model outputs the influence weight of each process parameter in each process; by calculating the influence weight of each process parameter, the key process parameters can be optimized first, and the optimization resources can be concentrated on the parameters that have a greater impact on production, thereby improving the optimization efficiency.

[0101] Specifically, according to the calculated influence weight of each process parameter, the corresponding process parameters are divided into different optimization levels through the preset parameter stratification standard; according to the actual production needs of aluminum processing, the corresponding parameter stratification standard is formulated. In this embodiment, the process parameters are sorted from high to low according to the influence weight, and the first 30% of the parameters are divided into the core layer optimization level, the middle 40% of the parameters are divided into the important layer optimization level, and the last 30% of the parameters are divided into the general layer optimization level; by stratifying the process parameters, the process parameters can be optimized according to the optimization level, thereby improving the efficiency and effect of the optimization work, adopting different optimization strategies for parameters of different optimization levels, and reasonably allocating optimization resources.

[0102] Furthermore, by combining the process map and the multi-level optimization level process parameters, the process parameters of each process are optimized in layers through a preset parameter optimization model to obtain optimized process parameters, thereby achieving the tuning of aluminum processing process parameters, including:

[0103] S701, combining the process map and the multi-level optimization level process parameters, extracting the parameter feature set of each optimization level process parameter;

[0104] S702: Optimize each optimization level process parameter separately using a preset parameter optimization model according to the parameter feature set to obtain first optimized process parameters;

[0105] S703 , optimizing the first optimized process parameters by analyzing the logical associations of process parameters at different optimization levels to obtain optimized process parameters, so as to achieve the optimization of aluminum processing process parameters.

[0106] In this embodiment, based on the process map and the multi-layer optimization level process parameters, the parameter feature set of each optimization level process parameter is extracted; the characteristic dimensions used to describe the process parameters are determined, including the parameter value range, change trend, the type of impact on product quality indicators, the correlation strength with other process parameters, the fluctuation frequency in the production process, etc.; for each optimization level process parameter, relevant data is collected from sources such as production records, quality inspection reports, and equipment operation data; the values ​​of the relevant data of each process parameter in each characteristic dimension are integrated to form the feature set of the parameter, and the parameter feature set corresponding to each optimization level process parameter is obtained; by extracting the parameter feature set, the complex parameter characteristics are converted into an analyzable data set, which facilitates the optimization of the parameters.

[0107] Specifically, such as Figure 3 According to the calculated parameter feature set, the preset parameter optimization model is used to optimize the process parameters of different optimization levels respectively. Through hierarchical optimization, different optimization methods are adopted for process parameters of different optimization levels. Taking into account the importance and characteristics of parameters of different levels, the optimization degree of parameters with high influence weight is higher. Through hierarchical optimization, appropriate optimization strategies can be adopted for parameters of different importance, and resources can be concentrated on optimizing the high-optimization level parameters that have the greatest impact on production, thereby improving optimization efficiency and effect and avoiding excessive consumption of resources on secondary parameters.

[0108] Furthermore, the logical relationships between process parameters of different optimization levels are integrated to further optimize the first optimized process parameters to ensure coordination between the various parameters, avoid the decline in overall process performance due to local optimization, and obtain optimized process parameters that are more in line with actual production needs; by considering the logical relationships between parameters, production problems caused by isolated optimization parameters are avoided, ensuring that the optimized process parameters can work together in actual production, and improving the stability and reliability of the overall process.

[0109] Furthermore, according to the parameter feature set, each optimization level process parameter is optimized respectively by a preset parameter optimization model to obtain a first optimized process parameter, wherein the optimization level includes a core layer, an important layer, and a general layer, and the parameter optimization model includes a first parameter optimization model, a second parameter optimization model, and a third parameter optimization model, including:

[0110] S801: For core layer process parameters, perform real-time optimization using a preset first parameter optimization model according to a corresponding core layer process parameter feature set to obtain first optimized process parameters of the core layer process parameters;

[0111] S802: dynamically adjust the important layer process parameters according to the corresponding important layer process parameter feature set through a preset second parameter optimization model to obtain first optimized process parameters of the important layer process parameters;

[0112] S803 : For general layer process parameters, perform batch optimization using a preset third parameter optimization model according to the corresponding general layer process parameter feature set to obtain first optimized process parameters of the general layer process parameters.

[0113] In this embodiment, different optimization methods are adopted for process parameters of different optimization levels. For the core layer process parameters, considering that the core layer process parameters play a decisive role in product quality and production efficiency, the core layer process parameters are optimized in real time. According to the real-time data and parameter characteristics in the production process, the parameters are quickly adjusted using the preset first parameter optimization model to ensure that the production is always in the optimal state and minimize the adverse effects caused by parameter changes; for example, the temperature in the smelting process, the cooling rate in the casting process and other parameters, small fluctuations in these parameters will lead to a significant decline in product quality or a significant reduction in production efficiency; high-precision sensors are installed on the production equipment to collect data on the core layer process parameters in real time, and calculate the core layer process parameters. The real-time feature set is input into a preset first parameter optimization model. The first parameter optimization model can be a machine learning model, a deep learning model, etc. This embodiment adopts a neural network model, and optimizes the neural network model through a large amount of historical data to obtain a pre-trained neural network model. The real-time feature set is input into the pre-trained neural network model, and the model quickly calculates the optimal value of the core layer process parameter in the current state, and obtains the first optimized process parameter of the corresponding core layer process parameter; by optimizing the core layer process parameters in real time, it is possible to respond to dynamic changes in the production process in a timely manner, avoid product quality defects and production interruptions caused by abnormal core parameters, and ensure the stability of product quality and the continuity of production.

[0114] For the important layer process parameters, taking into account that the important layer process parameters have a greater impact on product quality and production efficiency, but compared with the core layer parameters, their impact is slightly weaker and the changes are relatively less frequent, the important layer process parameters are dynamically adjusted through the preset second parameter optimization model to obtain the first optimized process parameters of the important layer process parameters; set the conditions for triggering the adjustment of the important layer process parameters, such as the production time interval, specifically including an adjustment every 50 batches of products, changes in the equipment operation status, specifically including the cumulative operation of the equipment reaching a certain number of hours, replacement of raw material batches, etc. At the same time, set the parameter change threshold. When the actual value of the important layer process parameter deviates from the target value by more than a certain range, , triggering adjustment; according to the parameter feature set of the important layer process parameters, parameter optimization is performed through a preset second parameter optimization model. The second parameter optimization model of this embodiment is a dynamic adjustment model based on a genetic algorithm. The model calculates the optimization adjustment plan of the important layer process parameters through a simulation evolution process based on the input data and the set production goals; for example, the model analysis finds that the rolling speed of the rolling process is related to the surface quality of the product. In order to improve the surface quality, it is calculated that the rolling speed should be appropriately reduced and the reduction rate should be adjusted. By dynamically adjusting the important layer process parameters, the parameter adjustment frequency can be reasonably controlled and the adjustment cost can be reduced under the premise of ensuring production quality and efficiency.

[0115] For general layer process parameters, considering that general layer process parameters have little impact on product quality and production efficiency, the cost of real-time or frequent optimization of each parameter individually is high and the benefits are limited. Through batch optimization strategy, production data within a certain period of time is accumulated, and the general layer process parameters are centrally analyzed and optimized; within a period of time, according to the parameter characteristic data of the general layer process parameters, through a preset third parameter optimization model, the third parameter optimization model of this embodiment is a batch optimization model based on a simulated annealing algorithm, the model calculates each group of parameters, and searches for the optimal parameter combination under the premise of meeting the basic production requirements. The basic production requirements specifically include qualified product quality, safe operation of production equipment, etc., to obtain the first optimized process parameters of the general layer process parameters; through batch optimization of the general layer process parameters, the optimization efficiency of the general layer process parameters is improved, the time and resource waste caused by optimizing each parameter individually is reduced, and the optimization cost is reduced.

[0116] Furthermore, the first optimized process parameters are optimized by analyzing the logical association of process parameters of different optimization levels to obtain optimized process parameters to achieve the tuning of aluminum processing process parameters, including:

[0117] S901. During the process of process parameter optimization, conflicting process parameters are identified to obtain a conflict parameter set.

[0118] S902. Calculate the conflict priority of each parameter in the conflict parameter set by analyzing the priority of each parameter corresponding to the process, the parameter impact weight, and the logical association between the parameters based on the conflict parameter set;

[0119] S903. Through a preset parameter optimization competition mechanism, the parameters with high conflicting priorities are optimized first, until the optimization of each conflicting parameter is completed, and the optimized process parameters are obtained to achieve the tuning of the aluminum processing process parameters.

[0120] In this embodiment, during the process parameter optimization process, the changes in each process parameter and the corresponding production data are collected in real time, and the adjustment process of each parameter, the values ​​before and after the adjustment, and the changes in the production data after the adjustment are recorded; through the preset conflict judgment rules, the process parameters of the optimization conflict are identified to obtain a set of conflicting parameters; for example, in the rolling process, increasing the rolling speed can improve production efficiency, but will lead to an increase in the surface roughness of the product, and the surface roughness is related to another parameter that controls the surface quality. At this time, the rolling speed and the parameter that controls the surface quality become conflicting parameters; by identifying the conflicting parameters, potential problems in the process of process parameter optimization can be discovered in a timely manner, production chaos and waste of resources caused by parameter conflicts can be avoided, and the correctness of the optimization direction can be ensured.

[0121] Specifically, according to the different degrees of influence and importance of different conflicting parameters on the production process, the conflict priority is calculated for each parameter in the conflicting parameter set; through the hierarchical analysis method, the corresponding weights are calculated for the three factors of the priority of the parameter corresponding to the process, the parameter influence weight and the logical association between the parameters; for each parameter in the conflicting parameter set, the score of the parameter on the three factors is calculated respectively, and the weighted calculation is performed in combination with the corresponding weights to obtain the conflict priority score of the parameter; the conflicting parameters are sorted according to the conflict priority score, and the conflict parameters with high conflict priority scores have corresponding conflict priorities; by performing priority calculation on the conflicting parameters, the processing order of the conflicting parameters can be determined to avoid blindly processing conflicts, so that the optimization resources can be prioritized on the conflicting parameters that have a greater impact on production, thereby improving the pertinence and effectiveness of the optimization.

[0122] Specifically, the conflicting parameters are optimized according to the conflict priority and the preset parameter optimization competition mechanism. For example, the parameter with the highest conflict priority is selected for optimization each time. During the optimization process, if the new parameter value causes the conflict of other parameters to intensify, the adjustment is rolled back and a better solution is found again. A limit is set on the number of optimization iterations to avoid falling into an infinite loop. The parameter with the highest conflict priority is selected from the conflicting parameter set, and the parameter is optimized through the preset conflicting parameter optimization model. The conflicting parameter optimization model of this embodiment is a particle swarm optimization model. Each conflicting parameter is optimized in turn until all conflicting parameters are optimized to obtain the optimized process parameters. By optimizing the conflicting parameters, the coordinated optimization of the conflicting parameters is achieved, so that the final optimized process parameters reach the optimal state as a whole, effectively improving the quality and efficiency of aluminum processing production.

[0123] Example 2

[0124] In this embodiment, if Figure 4 , provides a multi-process collaborative aluminum processing parameter tuning system, which is used to implement the multi-process collaborative aluminum processing parameter tuning method, including:

[0125] The process map construction module constructs a process map by analyzing the correlation between multiple processes of aluminum processing;

[0126] The process priority analysis module dynamically updates the priority of each process based on the process map and a preset process competition mechanism to obtain a priority order;

[0127] A process parameter stratification module analyzes the influence weight of the process parameters corresponding to each process step based on the priority order, stratifies the process parameters, and obtains multi-layer optimized process parameters;

[0128] The process parameter optimization module combines the process map and the multi-level optimization level process parameters, and optimizes the process parameters of each process in layers through a preset parameter optimization model to obtain optimized process parameters to achieve the tuning of aluminum processing process parameters.

[0129] In this embodiment, the process map construction module analyzes the correlation between the various processes of aluminum processing to construct a process map, extracts the feature set of each process, merges the features into a feature vector, and based on the feature vector, determines the causal relationship between the processes through a causal analysis algorithm, and constructs a process map that presents the logical relationship and influence of the processes with the processes as nodes and the causal relationship strength as the edge weight; the process priority analysis module analyzes the priority of each process according to the process map, calculates the contribution weight of each process to the aluminum processing process, allocates voting rights to the processes, determines the process priority through the voting results, and calculates the process priority adjustment amount using a sliding window within a preset time period. Through a preset process competition mechanism, the process priority is dynamically updated to ensure that key processes receive priority resource support, thereby improving production efficiency and flexibility.

[0130] Specifically, the process parameter stratification module stratifies the process parameters by analyzing the influence weight of the process parameters corresponding to each process, calculates the influence weight of each process parameter through the preset parameter impact analysis model, and divides the process parameters into multiple optimization levels such as core layer, important layer and general layer according to the preset parameter stratification standard. By stratifying the process parameters, optimization resources can be reasonably allocated, and differentiated optimization strategies can be adopted for parameters at different levels to improve the pertinence and efficiency of process parameter optimization; the process parameter optimization module combines the process relationship information of the process map construction module and the parameter stratification results of the process parameter stratification module, and optimizes the process parameters in layers through the preset parameter optimization model, and identifies the conflicting process parameters in the optimization process. Through the preset parameter optimization competition mechanism, the parameters with high conflict priority are optimized first, the conflict problem between process parameters is resolved, and the parameters are coordinated with each other. Finally, the overall optimal optimized process parameters are obtained, and the comprehensive adjustment of aluminum processing process parameters is achieved, thereby improving product quality and production efficiency.

[0131] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-process collaborative aluminum processing parameter tuning method, characterized in that: include: By analyzing the correlation between multiple processes of aluminum processing, a process map is constructed; According to the process map, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order; Based on the priority order, the influence weight of the process parameters corresponding to each process step is analyzed, the process parameters are layered, and multi-layer optimization level process parameters are obtained; Combining the process map and the multi-layer optimization level process parameters, the process parameters of each process are optimized in layers through a preset parameter optimization model to obtain optimized process parameters, so as to achieve the tuning of aluminum processing process parameters.

2. The method for optimizing aluminum processing parameters by multi-process collaboration according to claim 1, characterized in that: The process map is constructed by analyzing the correlation between multiple processes of aluminum processing, including: Based on the pre-acquired aluminum processing data, the feature set corresponding to each process is extracted; For each process, multiple features in the feature set are fused to obtain a feature vector; The causal relationship between processes is analyzed according to the feature vector, and a process map is constructed, wherein the processes are used as nodes in the process map, and the causal relationship between processes is used as the weight of the connecting edges between the nodes.

3. The method for optimizing aluminum processing parameters by multi-process collaboration according to claim 2, characterized in that: Analyzing the causal relationship between processes based on the feature vectors to construct a process map, wherein the processes are used as nodes in the process map and the causal relationship between processes is used as the weight of the edges connecting the nodes, including: Analyze the causal relationship between each process based on the characteristic vector and obtain the first causal direction; Calculate the causal relationship strength between processes through the preset causal strength calculation model; Taking the processes as nodes and the causal relationship strength between processes as edge weights, directed edges from cause nodes to effect nodes are established between process nodes to obtain a process graph.

4. The method for optimizing aluminum processing parameters by multi-process collaboration according to claim 1, characterized in that: According to the process map, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order, including: According to the process map, calculate the contribution weight of each process to the aluminum processing process; Based on the contribution weight, each process is assigned corresponding voting rights, and the processes vote to obtain voting results; According to the voting results, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order.

5. The method for optimizing aluminum processing parameters with multi-process collaboration according to claim 4, characterized in that: According to the voting results, the priority of each process is dynamically updated through a preset process competition mechanism to obtain a priority order, including: According to the voting results, the priority of each process is calculated and sorted from high to low to obtain the first priority order; Calculate the priority adjustment of each process through a sliding window within a preset time period; Adjusting the first priority order according to the priority adjustment amount to obtain a second priority order; By analyzing the process parameter adjustment directions in the first priority order and the second priority order, process competition is performed between processes with inconsistent parameter adjustment directions to obtain the priority order.

6. The method for optimizing aluminum processing parameters by coordinating multiple processes according to claim 1, characterized in that: Based on the priority order, the influence weight of the process parameters corresponding to each process step is analyzed, and the process parameters are layered to obtain multi-layer optimization level process parameters, including: Based on the priority order and the production data of each process, the influence weight of each process parameter is calculated through the preset parameter impact analysis model; According to the influence weights, the process parameters of each process step are layered using a preset parameter layering standard to obtain multi-layer optimized process parameters.

7. The method for optimizing aluminum processing parameters by multi-process collaboration according to claim 1, characterized in that: Combining the process map and the multi-level optimization level process parameters, the process parameters of each process are optimized in layers through a preset parameter optimization model to obtain optimized process parameters to achieve the tuning of aluminum processing process parameters, including: Combining the process map and multi-level optimization level process parameters, the parameter feature set of each optimization level process parameter is extracted; According to the parameter feature set, each optimization level process parameter is optimized respectively by using a preset parameter optimization model to obtain a first optimized process parameter; By analyzing the logical association of process parameters of different optimization levels, the first optimized process parameters are optimized to obtain optimized process parameters, so as to achieve the optimization of aluminum processing process parameters.

8. The method for optimizing aluminum processing parameters with multi-process collaboration according to claim 7, characterized in that: According to the parameter feature set, each optimization level process parameter is optimized respectively by a preset parameter optimization model to obtain a first optimized process parameter, wherein the optimization level includes a core layer, an important layer, and a general layer, and the parameter optimization model includes a first parameter optimization model, a second parameter optimization model, and a third parameter optimization model, including: For the core layer process parameters, according to the corresponding core layer process parameter feature set, a preset first parameter optimization model is used to perform real-time optimization to obtain first optimized process parameters of the core layer process parameters; For the important layer process parameters, according to the corresponding important layer process parameter feature set, a preset second parameter optimization model is used to dynamically adjust the important layer process parameters to obtain the first optimized process parameters of the important layer process parameters; For the general layer process parameters, batch optimization is performed through a preset third parameter optimization model according to the corresponding general layer process parameter feature set to obtain first optimized process parameters of the general layer process parameters.

9. The method for optimizing aluminum processing parameters by multi-process collaboration according to claim 7, characterized in that: The step of optimizing the first optimized process parameters by analyzing the logical association of process parameters at different optimization levels to obtain optimized process parameters to achieve the optimization of aluminum processing process parameters includes: During the process of process parameter optimization, conflicting process parameters are identified and a conflict parameter set is obtained; According to the conflict parameter set, the conflict priority of each parameter in the conflict parameter set is calculated by analyzing the priority of each parameter corresponding to the process, the parameter influence weight and the logical association between the parameters; Through the preset parameter optimization competition mechanism, the parameters with high conflicting priorities are optimized first, until the optimization of each conflicting parameter is completed, and the optimized process parameters are obtained to achieve the tuning of the aluminum processing process parameters.

10. A multi-process collaborative aluminum processing parameter tuning system, characterized in that: A method for optimizing aluminum processing parameters for achieving multi-process collaboration as described in any one of claims 1 to 9, comprising: The process map construction module constructs a process map by analyzing the correlation between multiple processes of aluminum processing; The process priority analysis module dynamically updates the priority of each process based on the process map and a preset process competition mechanism to obtain a priority order; A process parameter stratification module analyzes the influence weight of the process parameters corresponding to each process step based on the priority order, stratifies the process parameters, and obtains multi-layer optimized process parameters; The process parameter optimization module combines the process map and the multi-level optimization level process parameters, and optimizes the process parameters of each process in layers through a preset parameter optimization model to obtain optimized process parameters to achieve the tuning of aluminum processing process parameters.

Images