Smelting furnace group parameter tuning method, device, equipment, storage medium and product

Abstract

The invention discloses a smelting furnace group parameter tuning method, device and equipment, a storage medium and a product, and relates to the technical field of smelting furnace group tuning, and the smelting furnace group parameter tuning method comprises the steps of obtaining multi-dimensional data of a smelting furnace group, determining the working condition corresponding to each smelting furnace according to a preset working condition prediction model, and determining the working condition of each smelting furnace according to the working condition prediction model. Further determining self-adaptive tuning parameters of each smelting furnace; and inputting the self-adaptive tuning parameters of each smelting furnace into a preset multi-target particle swarm optimization model to carry out group optimization on the smelting furnace group to obtain an optimal parameter combination of the smelting furnace group. According to the method, the multi-dimensional data of the smelting furnace group is obtained, the working condition corresponding to each smelting furnace is determined by using the preset working condition prediction model, the adaptive tuning parameters are determined, the parameters are input into the multi-target particle swarm optimization model for group optimization, and the optimal parameter combination of the smelting furnace group is output. And efficient cooperative control and overall performance improvement of the smelting furnace group are realized.

Description

Technical Field

[0001] This application relates to the technical field of smelting furnace group optimization, and particularly to a method, device, equipment, storage medium and product for optimizing parameters of a smelting furnace group. Background Art

[0002] In the operation management of a smelting furnace group, how to optimize the working parameters of each smelting furnace in real time to improve production efficiency and product quality has always been a key issue of concern in the industrial community. Traditional optimization methods often rely on fixed empirical parameters, and this single mode is difficult to adapt to complex working conditions changes and production requirements, resulting in low production efficiency and energy waste. To address this challenge, people often adopt data-driven methods for optimization, by collecting operation data and running parameters of the smelting furnace, analyzing and predicting changes in working conditions, and then adjusting the working parameters of the smelting furnace.

[0003] However, due to different working condition characteristics and spatio-temporal dependencies of each furnace in the smelting furnace group, the optimization of a single furnace cannot fully utilize the synergy effect of the entire smelting furnace group. Therefore, how to optimize the parameters of the smelting furnace group is an urgent problem to be solved. Summary of the Invention

[0004] The main purpose of this application is to provide a method, device, equipment, storage medium and product for optimizing parameters of a smelting furnace group, aiming to solve the technical problem that the parameters of the smelting furnace group cannot be optimized.

[0005] To achieve the above object, this application proposes a method for optimizing parameters of a smelting furnace group, and the method includes: Obtain multi-dimensional data of the smelting furnace group, confirm the working conditions corresponding to each smelting furnace according to a preset working condition prediction model, and then determine the adaptive optimization parameters of each smelting furnace; Input the adaptive optimization parameters of each smelting furnace into a preset multi-objective particle swarm optimization model for group optimization of the smelting furnace group to obtain the optimal parameter combination of the smelting furnace group.

[0006] In one embodiment, the step of inputting the adaptive optimization parameters of each smelting furnace into a preset multi-objective particle swarm optimization model for group optimization of the smelting furnace group to obtain the optimal parameter combination of the smelting furnace group includes: Initialize the multi-objective particle swarm optimization model, where a preset penalty function is used as a process constraint condition; According to the input adaptive optimization parameters of each smelting furnace, iteratively update the particle velocity and position through the multi-objective particle swarm optimization model, record the historical optimal position and the global optimal position, and output the optimal parameter combination of the smelting furnace group.

[0007] In one embodiment, the steps of obtaining multi-dimensional data of the smelting furnace group, confirming the working conditions corresponding to each smelting furnace according to a preset working condition prediction model, and then determining the adaptive tuning parameters of each smelting furnace include: Obtain multi-dimensional operation data of the smelting furnace group, and preprocess to obtain multi-dimensional data of the smelting furnace; Based on the random forest algorithm, extract key features from the multi-dimensional data of the smelting furnace to determine the corresponding key features of the smelting furnace; Input the key features of the smelting furnace into a preset working condition prediction model to predict the working condition law result of the molten aluminum; According to the working condition law result of the molten aluminum, calculate the time interval between the corresponding smelting furnace and the preset working condition, and dynamically output the adaptive tuning parameters as a constraint condition.

[0008] In one embodiment, the steps of inputting the key features of the smelting furnace into a preset working condition prediction model to predict the working condition law result of the molten aluminum include: Use time series conversion technology to convert the key features of the smelting furnace into the input format of the time series neural network; Adjust the structural parameters and training parameters of the time series neural network through parameter optimization methods; Verify the model performance based on multi-dimensional evaluation indicators, predict the on-off state of the molten aluminum in the future period, and output the working condition law result of the molten aluminum.

[0009] In one embodiment, the steps of calculating the time interval between the corresponding smelting furnace and the preset working condition according to the working condition law result of the molten aluminum, and dynamically outputting the adaptive tuning parameters as a constraint condition include: If the time interval is less than the first preset threshold, increase the opening of the gas valve and the control value of the burner air motor; If the time interval is between the first preset threshold and the second preset threshold, keep the current parameters; If the time interval is greater than the second preset threshold, decrease the opening of the gas valve and the control value of the burner air motor.

[0010] In one embodiment, the method further includes the collaborative control steps of the smelting furnace group: According to the optimal parameter combination of the smelting furnace group, dynamically allocate the flow path of the molten aluminum between the smelting furnace and the holding furnace; Real-time monitor the molten aluminum capacity status of each holding furnace, and automatically adjust the flow distribution of the aluminum scrap furnace when the capacity is detected to be insufficient; Based on the closed-loop feedback mechanism, optimize the production control parameters between the smelting furnace and the holding furnace groups.

[0011] In addition, to achieve the above object, the present application further provides an apparatus for optimizing parameters of a smelting furnace group, the apparatus for optimizing parameters of a smelting furnace group comprising: An acquisition module, configured to acquire multi-dimensional data of the smelting furnace group, confirm the working conditions corresponding to each smelting furnace according to a preset working condition prediction model, and further determine the adaptive optimization parameters of each smelting furnace; An optimization module, configured to input the adaptive optimization parameters of each smelting furnace into a preset multi-objective particle swarm optimization model to perform population optimization on the smelting furnace group, so as to obtain an optimal parameter combination of the smelting furnace group.

[0012] In addition, to achieve the above object, the present application further provides a device for optimizing parameters of a smelting furnace group, the device comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the method for optimizing parameters of a smelting furnace group as described above.

[0013] In addition, to achieve the above object, the present application further provides a storage medium, the storage medium being a computer-readable storage medium, and a computer program being stored on the storage medium, the computer program, when executed by a processor, implementing the steps of the method for optimizing parameters of a smelting furnace group as described above.

[0014] In addition, to achieve the above object, the present application further provides a computer program product, the computer program product comprising a computer program, the computer program, when executed by a processor, implementing the steps of the method for optimizing parameters of a smelting furnace group as described above.

[0015] One or more technical solutions proposed by the present application have at least the following technical effects: Compared with the traditional optimization methods in the related art, which often rely on fixed empirical parameters, the present application acquires multi-dimensional data of the smelting furnace group, confirms the working conditions corresponding to each smelting furnace according to a preset working condition prediction model, and further determines the adaptive optimization parameters of each smelting furnace; inputs the adaptive optimization parameters of each smelting furnace into a preset multi-objective particle swarm optimization model to perform population optimization on the smelting furnace group, so as to obtain an optimal parameter combination of the smelting furnace group. It can be understood that the method for optimizing parameters of a smelting furnace group acquires multi-dimensional data of the smelting furnace group, uses a preset working condition prediction model to confirm the working conditions corresponding to each smelting furnace, determines the adaptive optimization parameters, then inputs these parameters into a preset multi-objective particle swarm optimization model for population optimization, and finally outputs the optimal parameter combination of the smelting furnace group, so as to achieve efficient collaborative control of the smelting furnace group and improvement of the overall performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings here are incorporated into the specification and form a part of this specification, showing embodiments consistent with this application, and are used together with the specification to explain the principles of this application.

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] Figure 1 It is a schematic flowchart provided for the first embodiment of the method for optimizing the parameters of the melting furnace group in this application; Figure 2 It is a schematic diagram of the convergence curve of the optimization algorithm provided in the embodiments of the method for optimizing the parameters of the melting furnace group in this application; Figure 3 It is a schematic flowchart of the single-furnace optimization provided in the embodiments of the method for optimizing the parameters of the melting furnace group in this application; Figure 4 It is a schematic flowchart of the collaborative optimization provided in the embodiments of the method for optimizing the parameters of the melting furnace group in this application; Figure 5 It is a schematic diagram of the module structure of the device for optimizing the parameters of the melting furnace group in the embodiments of this application; Figure 6 It is a schematic diagram of the device structure of the hardware operating environment involved in the method for optimizing the parameters of the melting furnace group in the embodiments of this application.

[0019] The realization of the purpose, functional features, and advantages of this application will be further described in conjunction with the embodiments with reference to the accompanying drawings. Detailed Embodiments

[0020] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of this application and are not used to limit this application.

[0021] To better understand the technical solutions of this application, the following will be described in detail in conjunction with the accompanying drawings of the specification and specific embodiments.

[0022] The main solution of the embodiments of this application is: Obtain multi-dimensional data of the melting furnace group, confirm the working conditions corresponding to each melting furnace according to a preset working condition prediction model, and then determine the adaptive tuning parameters of each melting furnace; Input the adaptive tuning parameters of each melting furnace into a preset multi-objective particle swarm optimization model for group optimization of the melting furnace group to obtain the optimal parameter combination of the melting furnace group.

[0023] In this embodiment, the present application takes the melting furnace group parameter optimization device as the execution entity. For the convenience of description, it is hereinafter referred to as the "device" for specific description.

[0024] Since the prior art often relies on fixed empirical parameters, this single mode is difficult to adapt to complex working conditions and production requirements, resulting in low production efficiency and energy waste. To address this challenge, a data-driven approach is adopted for optimization. By collecting the operation data and running parameters of the melting furnace, analyzing and predicting the changes in working conditions, and then adjusting the working parameters of the melting furnace.

[0025] The present application provides a solution. By obtaining multi-dimensional data of the melting furnace group, using a preset working condition prediction model to confirm the working conditions corresponding to each melting furnace, and determining the adaptive optimization parameters. Then, these parameters are input into a preset multi-objective particle swarm optimization model for group optimization, and finally, the optimal parameter combination of the melting furnace group is output. The specific working principle is as follows: First, the key features of the melting furnace are extracted through the random forest algorithm, and then these features are input into the working condition prediction model after time series conversion and parameter optimization to further predict the working condition law of the molten aluminum. According to the prediction results, the time interval required for the melting furnace to reach the preset working condition is calculated, and the opening degree of the gas valve and the control value of the burner air motor are dynamically adjusted accordingly, so as to achieve the adaptive optimization of the melting furnace. Further, the multi-objective particle swarm optimization model is used to perform group optimization on these adaptive optimization parameters to optimize the optimal parameter combination. The optimization module not only optimizes the parameters of the melting furnace itself, but also dynamically allocates the flow path of the molten aluminum between the melting furnace and the holding furnace through a closed-loop feedback mechanism, monitors and adjusts the capacity state in real time, and finally realizes the efficient collaborative control and overall performance improvement of the melting furnace group.

[0026] Based on this, the embodiment of the present application provides a method for optimizing the parameters of a melting furnace group, referring to Figure 1 , Figure 1 is the flow chart of the first embodiment of the method for optimizing the parameters of the melting furnace group of the present application.

[0027] In this embodiment, the method for optimizing the parameters of the melting furnace group includes steps S10 to S20: Step S10, obtain multi-dimensional data of the melting furnace group, confirm the working conditions corresponding to each melting furnace according to a preset working condition prediction model, and then determine the adaptive optimization parameters of each melting furnace; It should be noted that the melting furnace group refers to a set of equipment composed of multiple melting furnaces and holding furnaces, which is used for operations such as melting, slag removal, and casting of metals.

[0028] Multi-dimensional data refers to multi-source heterogeneous data covering natural gas meter data, feeding information, molten aluminum transfer, and furnace group operation data, etc.

[0029] The operating condition prediction model is a prediction model based on the LSTM neural network, which is used to predict the switching state and operating condition change rules of aluminum liquid in the future period.

[0030] Adaptive tuning parameters are control parameters that are dynamically adjusted based on the operating condition prediction results, such as gas valve opening, burner air motor control value, etc.

[0031] It is understandable that a metal processing enterprise uses a melting furnace group to produce aluminum liquid, and the group includes but is not limited to a pre-treatment furnace, a melting furnace A, a melting furnace B, an aluminum chip furnace, a static furnace A, and a static furnace B. In order to optimize production efficiency and energy consumption, the enterprise adopts a melting furnace group parameter tuning method, and the specific implementation steps are as follows: Data collection and preprocessing: The multi-dimensional operation data of the smelting furnace group is collected through sensors, including instantaneous natural gas flow, aluminum liquid temperature, feeding time, etc. The collected data is cleaned, abnormal values ​​(such as abnormally high values ​​of instantaneous natural gas flow) are eliminated, and the Min-Max normalization method is used to standardize the data.

[0032] Feature extraction: Use the random forest algorithm to screen the preprocessed data, evaluate the importance of each feature, and retain the top N key features (such as aluminum liquid temperature and natural gas instantaneous flow rate).

[0033] Working condition prediction: Input key features into the working condition prediction model based on LSTM neural network, and convert time series data into model input format through sliding window technology. Use grid search to optimize model hyperparameters (such as 2 hidden layers and 64 hidden units) to predict the switching state of aluminum liquid in the next T minutes.

[0034] Adaptive tuning parameter calculation: Based on the prediction results, calculate the time interval from each smelting furnace to the "next aluminum liquid release state". For example, if the prediction interval is 10 minutes (less than the first preset threshold of 15 minutes), increase the gas valve opening and burner air motor control value; if the interval is 20 minutes (between 15-30 minutes), maintain the current parameters; if the interval is 40 minutes (greater than the second preset threshold of 30 minutes), reduce the gas valve opening to reduce energy consumption.

[0035] Swarm optimization: Input the adaptive tuning parameters of each melting furnace into the multi-objective particle swarm optimization model, initialize the model and set the penalty function to handle process constraints (such as temperature fluctuations and rest time). By iteratively updating the particle speed and position, the optimal parameter combination of the melting furnace group is finally output to achieve global optimization of the production rhythm between furnace groups.

[0036] Through the above method, the company successfully realized intelligent energy-saving control of the smelting furnace group, significantly reduced energy consumption and improved production efficiency.

[0037] In a feasible implementation manner, step S10 may include: Obtain multi-dimensional operation data of the smelting furnace group, and preprocess it to obtain multi-dimensional data of the smelting furnace; Based on the random forest algorithm, extract key features from the multi-dimensional data of the smelting furnace, and determine the corresponding key features of the smelting furnace; Input the key features of the smelting furnace into a preset working condition prediction model, and predict the working condition law result of the molten aluminum; According to the working condition law result of the molten aluminum, calculate the time interval between the corresponding smelting furnace and the preset working condition, and dynamically output adaptive tuning parameters as constraint conditions.

[0038] It should be noted that the random forest algorithm is an ensemble learning method that improves prediction accuracy and stability by constructing multiple decision trees and integrating their results.

[0039] The working condition prediction model is a prediction model based on the LSTM neural network, which is used to predict the switching state and working condition change law of the molten aluminum in the future time period.

[0040] The working condition law result of the molten aluminum is the state change law of the molten aluminum in the future time period obtained through the working condition prediction model, including the switching state, etc.

[0041] Exemplarily, a certain metal processing enterprise uses a smelting furnace group to produce molten aluminum. To optimize production efficiency and energy consumption, the enterprise implements the following parameter tuning method for the smelting furnace group, with particular emphasis on the intelligence of data processing, feature extraction, and working condition identification: Data collection and preprocessing: Collect multi-dimensional operation data of the smelting furnace group through sensors, covering instantaneous natural gas flow, molten aluminum temperature, feeding time, etc. The data is cleaned to remove outliers (such as extremely high instantaneous natural gas flow values), and the Min-Max normalization method is used for standardization to ensure data consistency and accuracy.

[0042] Key feature extraction: Use the random forest algorithm to screen features from the preprocessed data and evaluate the importance of each feature. For example, features such as molten aluminum temperature and instantaneous natural gas flow are identified as key features, providing accurate input for subsequent prediction.

[0043] Working condition prediction: Input the key features into the working condition prediction model based on the LSTM neural network, and convert the time series data format through the sliding window technique. Use grid search to optimize the model hyperparameters (such as the number of hidden layers is 2 and the number of hidden units is 64), predict the switching state of the molten aluminum in the next T minutes, and output the working condition law result of the molten aluminum.

[0044] Calculation of adaptive tuning parameters: According to the prediction results, calculate the time interval between each smelting furnace and the "next molten aluminum discharge open state". For example: If the time interval is less than 10 minutes, increase the opening of the gas valve and the control value of the burner air motor to quickly preheat the molten aluminum. If the time interval is between 10 - 30 minutes, maintain the current parameters for stable operation. If the time interval is greater than 30 minutes, reduce the opening of the gas valve to reduce energy consumption.

[0045] Intelligent optimization: Through the multi - objective particle swarm optimization model and combined with adaptive tuning parameters, global optimization of the production rhythm among furnace groups is achieved. The optimized parameter combination significantly improves production efficiency and energy utilization efficiency.

[0046] This method realizes rapid response and precise control of complex working conditions through intelligent data processing and prediction, effectively reducing energy consumption and enhancing the stability and reliability of the production process.

[0047] In a feasible implementation manner, the step of inputting the key features of the melting furnace into a preset working condition prediction model to predict the result of the molten aluminum working condition law includes: Using time - series conversion technology to convert the key features of the melting furnace into the input format of a time - series neural network. Adjust the structural parameters and training parameters of the time - series neural network through parameter optimization methods. Verify the model performance based on multi - dimensional evaluation indicators, predict the on - off state of the molten aluminum in the future period, and output the result of the molten aluminum working condition law.

[0048] It should be noted that time - series conversion technology is a technology that converts time - series data into an input format suitable for a neural network, usually generating input sequences and corresponding target values through a sliding window method.

[0049] Multi - dimensional evaluation indicators are multiple indicators used to comprehensively evaluate the model performance, including accuracy, recall rate, F1 - value, etc.

[0050] Exemplarily, a metal processing enterprise uses a group of melting furnaces for molten aluminum production. To optimize production efficiency and energy consumption, the enterprise implements the following parameter tuning method for the melting furnace group, especially emphasizing the intelligence of time - series conversion and model optimization: Time - series conversion: Using the sliding window technique, convert the key features of the melting furnace (such as molten aluminum temperature, instantaneous natural gas flow) into the input format of a time - series neural network. Set the time window to 120 time steps to generate input sequences and corresponding target values, making it into a supervised learning format.

[0051] Parameter Optimization: The structural and training parameters of the LSTM neural network are adjusted through the grid search method. The optimized hyperparameters include the number of hidden layers (2 layers), the number of hidden units (64), the learning rate (0.0001), and the number of iterations (100 times). The optimal configuration is found by traversing the predefined hyperparameter combinations.

[0052] Model Validation and Prediction: By evaluating the performance of each hyperparameter combination on the test set, the optimal hyperparameter combination is selected to complete the final prediction. The evaluation metric is the Macro Average Accuracy, and the recall rate, F1 value, etc. are also used to evaluate the prediction performance of the LSTM model, as shown in the following formula: where T is the time length, and Accuracy is the accuracy at the T - th step;

[0053] The prediction results of each furnace model are shown in Table 1.

[0054] Table 1: Prediction Results of Each Furnace Model

[0055] The optimized LSTM model is used to predict the switching state of the molten aluminum within the next T minutes, and the operating condition law results of the molten aluminum are output.

[0056] Intelligent Tuning: According to the prediction results, the operating parameters of the melting furnace are dynamically adjusted. For example, if it is predicted that the switching state of the molten aluminum will occur in 10 minutes, the opening degree of the gas valve and the control value of the burner air motor are increased to ensure that the temperature of the molten aluminum meets the requirements; if the predicted interval is long, the parameter values are decreased to reduce energy consumption.

[0057] It can be understood that through intelligent time - series conversion and parameter optimization, this method significantly improves the prediction accuracy and response speed of the model, realizes rapid adaptation and precise control of complex working conditions, effectively reduces energy consumption, and improves the stability and reliability of the production process.

[0058] In a feasible implementation manner, the step of calculating the time interval between the corresponding melting furnace and the preset working condition according to the molten aluminum working condition law result and dynamically outputting the adaptive tuning parameters as a constraint condition includes: If the time interval is less than the first preset threshold, the opening degree of the gas valve and the control value of the burner air motor are increased; If the time interval is between the first preset threshold and the second preset threshold, the current parameters are maintained; If the time interval is greater than the second preset threshold, the opening degree of the gas valve and the control value of the burner air motor are decreased.

[0059] It should be noted that the first preset threshold is a set time threshold used to determine whether the time interval is small, so as to decide whether to increase the opening degree of the gas valve and the control value of the burner air motor.

[0060] The second preset threshold is a set time threshold used to determine whether the time interval is large, so as to decide whether to decrease the opening degree of the gas valve and the control value of the burner air motor.

[0061] It can be understood that a metal processing enterprise uses a group of melting furnaces to produce molten aluminum. To optimize production efficiency and energy consumption, the enterprise has implemented the following parameter optimization method for the melting furnace group, with particular emphasis on the dynamic parameter adjustment mechanism based on the time interval: Data collection and preprocessing: Collect multi-dimensional operation data of the melting furnace group through sensors, including instantaneous natural gas flow, molten aluminum temperature, feeding time, etc. The data is cleaned and normalized to ensure data consistency and accuracy.

[0062] Key feature extraction: Use the random forest algorithm to screen features from the preprocessed data, evaluate the importance of each feature, and retain the top N key features (such as molten aluminum temperature, instantaneous natural gas flow).

[0063] Working condition prediction: Input the key features into the working condition prediction model based on the LSTM neural network. Convert the time series data into the model input format through the sliding window technique. Use grid search to optimize the model hyperparameters (such as the number of hidden layers is 2 and the number of hidden units is 64), and predict the on / off state of the molten aluminum within the next T minutes.

[0064] Dynamic parameter adjustment: According to the prediction results, calculate the time interval for each melting furnace from the current state to the "next molten aluminum discharge on state", and dynamically adjust the opening degree of the gas valve and the control value of the burner air motor based on the following logic: If the time interval is less than the first preset threshold (such as 10 minutes): Increase the opening degree of the gas valve and the control value of the burner air motor to quickly preheat the molten aluminum and ensure that the requirements are met when discharging the molten aluminum next time.

[0065] If the time interval is between the first preset threshold (10 minutes) and the second preset threshold (30 minutes): Keep the current parameters and operate stably.

[0066] If the time interval is greater than the second preset threshold (30 minutes): Decrease the opening degree of the gas valve and the control value of the burner air motor to reduce energy consumption.

[0067] Intelligent optimization: Through the multi-objective particle swarm optimization model, combined with the dynamically adjusted parameters, achieve the global optimization of the production rhythm among the furnace groups. The optimized parameter combination significantly improves production efficiency and energy utilization efficiency.

[0068] Exemplarily, referring toFigure 3 , a metal processing enterprise uses a group of smelting furnaces to produce aluminum liquid. To optimize the operating efficiency and energy consumption of a single furnace, the enterprise implemented the following single-furnace optimization method based on LSTM: Collect the historical operating data of Furnace A, including instantaneous natural gas flow rate, aluminum liquid temperature, feeding time, etc. The data sampling frequency is once per minute. Clean the collected data, remove outliers (such as extremely high instantaneous natural gas flow rate), and standardize the data using the Min-Max normalization method.

[0069] Use the random forest algorithm to perform feature screening on the preprocessed data, evaluate the importance of each feature, and retain the top N key features (such as aluminum liquid temperature, instantaneous natural gas flow rate).

[0070] Input the key features into a prediction model based on the LSTM neural network. Convert the time series data into the model input format through the sliding window technique. Set the time window to 120 time steps to generate the input sequence and the corresponding target values. Optimize the model hyperparameters (such as the number of hidden layers is 2, the number of hidden units is 64, and the learning rate is 0.0001) using grid search to predict the on / off state of the aluminum liquid within the next T minutes.

[0071] Among them, Figure 3 in represents the input information; represents the memory cell at the previous time step; represents the hidden state at the previous time step; σ and tanh represent activation functions; represents the updated memory cell; represents the generated hidden state; represents the forget gate; represents the input gate; represents the output gate; represents the candidate memory cell.

[0072] Verify the model performance based on multi-dimensional evaluation metrics (such as macro-average accuracy, recall rate, F1 value). For example, the macro-average accuracy of Furnace A reaches 92.87%. Use the optimized LSTM model to predict the on / off state of the aluminum liquid within the next T minutes and output the aluminum liquid condition rule results.

[0073] According to the prediction results, calculate the time interval of Furnace A from the "next aluminum liquid discharge on state". For example, if the predicted interval is 10 minutes, increase the gas valve opening and the burner air motor control value; if the predicted interval is 20 minutes, keep the current parameters; if the predicted interval is 40 minutes, reduce the gas valve opening to reduce energy consumption.

[0074] This method achieves rapid response and precise control of complex working conditions through dynamic parameter adjustment based on time intervals, effectively reduces energy consumption, and improves the stability and reliability of the production process.

[0075] Step S20, inputting the adaptive tuning parameters of each smelting furnace into a preset multi-objective particle swarm optimization model to perform group optimization of the smelting furnaces to obtain an optimal parameter combination of the smelting furnaces.

[0076] It should be noted that the multi-objective particle swarm optimization model is a model based on the particle swarm optimization algorithm, which is used to simultaneously optimize multiple objective functions (such as maximizing the aluminum liquid discharge efficiency, minimizing temperature fluctuations and standing waiting time), and handle process constraints through penalty functions.

[0077] Swarm optimization is to globally optimize the parameters of the smelting furnace group through a multi-objective particle swarm optimization model to achieve the coordination of production rhythm among furnace groups and improve the overall efficiency.

[0078] It is understandable that a metal processing company uses a melting furnace group to produce aluminum liquid, which includes melting furnace A, melting furnace B, aluminum chip furnace, static furnace A and static furnace B. In order to optimize production efficiency and energy consumption, the company implemented the following melting furnace group parameter tuning method: Data collection and preprocessing: The multi-dimensional operation data of the smelting furnace group is collected through sensors, including instantaneous natural gas flow, aluminum liquid temperature, feeding time, etc. The collected data is cleaned, abnormal values ​​(such as abnormally high values ​​of instantaneous natural gas flow) are eliminated, and the Min-Max normalization method is used to standardize the data.

[0079] Feature extraction: Use the random forest algorithm to screen the preprocessed data, evaluate the importance of each feature, and retain the top N key features (such as aluminum liquid temperature and natural gas instantaneous flow rate).

[0080] Working condition prediction: Input key features into the working condition prediction model based on LSTM neural network, and convert time series data into model input format through sliding window technology. Use grid search to optimize model hyperparameters (such as 2 hidden layers and 64 hidden units) to predict the switching state of aluminum liquid in the next T minutes.

[0081] Adaptive tuning parameter calculation: Based on the prediction results, calculate the time interval from each smelting furnace to the "next aluminum liquid release state". For example, if the prediction interval is 10 minutes (less than the first preset threshold of 15 minutes), increase the gas valve opening and burner air motor control value; if the interval is 20 minutes (between 15-30 minutes), maintain the current parameters; if the interval is 40 minutes (greater than the second preset threshold of 30 minutes), reduce the gas valve opening to reduce energy consumption.

[0082] Group optimization: Input the adaptive tuning parameters of each smelting furnace into the multi-objective particle swarm optimization model, initialize the model and set up a penalty function to handle process constraints (such as temperature fluctuations and standing time). By iteratively updating the particle velocity and position, finally output the optimal parameter combination of the smelting furnace group, and achieve the global optimization of the production rhythm among furnace groups.

[0083] Dynamic path allocation and feedback optimization: According to the optimized parameter combination, dynamically allocate the flow path of molten aluminum between the smelting furnace and the standing furnace. Real-time monitor the molten aluminum capacity status of each standing furnace. When it is detected that the capacity is insufficient, automatically adjust the flow path allocation of the aluminum scrap furnace. Through the closed-loop feedback mechanism, further optimize the production control parameters between the smelting furnace and the standing furnace groups.

[0084] Through the above methods, the enterprise has successfully achieved the intelligent energy-saving control of the smelting furnace group, significantly reducing energy consumption and improving production efficiency.

[0085] In a feasible implementation manner, step S20 may include: Initialize the multi-objective particle swarm optimization model, where a preset penalty function is used as the process constraint condition; According to the input adaptive tuning parameters of each smelting furnace, iteratively update the particle velocity and position through the multi-objective particle swarm optimization model, record the historical optimal position and the global optimal position, and output the optimal parameter combination of the smelting furnace group.

[0086] It should be noted that the penalty function is a function used to handle process constraint conditions. By imposing penalties on particles that violate the constraint conditions during the optimization process, reducing their fitness values, so as to ensure that the optimization results meet the process requirements.

[0087] Exemplarily, referring to Figure 2 , Figure 2 shows the convergence process of the particle swarm optimization algorithm in the collaborative optimization of furnace groups. Set the particle swarm size to 40 particles, the initial value of the inertia weight to 0.9, linearly decreasing to 0.4, the learning factor c1 = 2.0, c2 = 2.0, and the maximum number of iterations to 200 times. Use the preset penalty function as the process constraint condition to handle hard constraints such as temperature fluctuations and standing time.

[0088] Input the adaptive tuning parameters of each smelting furnace into the multi-objective particle swarm optimization model, and iteratively update the particle velocity and position. In each iteration, simulate the operation of the furnace group, calculate the molten aluminum discharge efficiency, temperature change, and standing waiting time corresponding to the current particle. If the constraint conditions are violated (such as insufficient standing time or over-limit temperature), reduce its fitness value through the penalty function. Record the historical optimal position and the global optimal position of each particle.

[0089] Draw the convergence curve of the particle swarm optimization algorithm to show the fitness change of the algorithm during the iteration process. The curve shows that the algorithm converges rapidly in the first 50 iterations and the fitness value gradually stabilizes, indicating that the model has found a relatively optimal combination of control parameters.

[0090] After 200 iterations, the algorithm converges and outputs the optimal parameter combination of the melting furnace group, including the opening degree of the gas valve, the selection of the flow direction of the aluminum scrap furnace, and the aluminum liquid discharge rate, etc.

[0091] Exemplarily, a metal processing enterprise uses a melting furnace group for aluminum liquid production. To optimize production efficiency and energy consumption, the enterprise implements the following parameter tuning method for the melting furnace group: Initialize the multi-objective particle swarm optimization model: Set the particle swarm size to 40 particles, the initial value of the inertia weight to 0.9, linearly decreasing to 0.4, the learning factor c1 = 2.0, c2 = 2.0, and the maximum number of iterations to 200 times. Use the preset penalty function as the process constraint condition to handle hard constraints such as temperature fluctuations and standing time.

[0092] Iterative optimization: Input the adaptive tuning parameters of each melting furnace into the multi-objective particle swarm optimization model, and update the particle velocity and position through iteration. In each iteration, simulate the operation of the furnace group, calculate the aluminum liquid discharge efficiency, temperature change, and standing waiting time corresponding to the current particle. If the constraint conditions are violated (such as insufficient standing time, temperature exceeding the limit), reduce its fitness value through the penalty function. Record the historical optimal position and the global optimal position of each particle.

[0093] Output the optimal parameter combination: After 200 iterations, the algorithm converges and outputs the optimal parameter combination of the melting furnace group, including the opening degree of the gas valve, the selection of the flow direction of the aluminum scrap furnace, and the aluminum liquid discharge rate, etc.

[0094] Through the above method, the enterprise has successfully achieved intelligent energy-saving control of the melting furnace group, significantly reducing energy consumption and improving production efficiency.

[0095] In a feasible implementation manner, the method further includes the collaborative control step of the melting furnace group: Dynamically allocate the flow path of aluminum liquid between the melting furnace and the standing furnace according to the optimal parameter combination of the melting furnace group; Real-time monitor the aluminum liquid capacity status of each standing furnace, and automatically adjust the flow direction distribution of the aluminum scrap furnace when it is detected that the capacity is insufficient; Optimize the production control parameters between the melting furnace and the standing furnace group based on the closed-loop feedback mechanism.

[0096] It should be noted that the closed-loop feedback mechanism is a mechanism that optimizes the system performance by real-time monitoring the system output and comparing it with the target output, thereby adjusting the system input.

[0097] In this embodiment, a metal processing enterprise uses a group of melting furnaces to produce molten aluminum. To optimize production efficiency and energy consumption, the enterprise implements the following method for optimizing the parameters of the melting furnace group, with particular emphasis on the real-time and intelligence of the closed-loop feedback mechanism: Dynamic path allocation: According to the optimal parameter combination of the melting furnace group, dynamically allocate the flow path of molten aluminum between the melting furnace and the holding furnace. For example, the molten aluminum from Melting Furnace A preferentially flows to Holding Furnace A, and the molten aluminum from Melting Furnace B preferentially flows to Holding Furnace B. The flow direction of the aluminum scrap furnace is dynamically adjusted according to the real-time capacity and production demand.

[0098] Real-time monitoring and automatic adjustment: Real-time monitor the molten aluminum capacity status of each holding furnace. When it is detected that the capacity of Holding Furnace A is insufficient (such as the remaining capacity is less than 20%), the system automatically adjusts the flow direction of the aluminum scrap furnace to preferentially flow to Holding Furnace A to avoid the risk of overflow and ensure production continuity.

[0099] Closed-loop feedback optimization: Based on the closed-loop feedback mechanism, the system continuously collects production data (such as molten aluminum temperature, discharge rate, holding time, etc.) and compares it with the target parameters. Through the multi-objective particle swarm optimization model, dynamically adjust the production control parameters of the melting furnace and the holding furnace, such as the opening degree of the gas valve, the control value of the burner air motor, etc., to ensure that the system operates in the optimal state.

[0100] It can be understood that this method realizes the real-time monitoring and intelligent adjustment of the production process through the closed-loop feedback mechanism, effectively avoids production bottlenecks, improves the stability and reliability of the production process, and meets the industry's requirements for efficient, energy-saving, and intelligent production.

[0101] This embodiment provides a method for optimizing the parameters of a melting furnace group. By obtaining multi-dimensional data of the melting furnace group, using a preset working condition prediction model to confirm the working conditions corresponding to each melting furnace, and determining the adaptive optimization parameters, and then inputting these parameters into a preset multi-objective particle swarm optimization model for population optimization, finally output the optimal parameter combination of the melting furnace group to achieve efficient collaborative control and overall performance improvement of the melting furnace group.

[0102] In a feasible embodiment, referring to Figure 3 , Figure 3 shows the specific implementation steps of optimizing the parameters of the melting furnace group, aiming to achieve the optimal parameter combination of the melting furnace group through multi-dimensional data acquisition, feature extraction, working condition prediction model, and multi-objective particle swarm optimization model, so as to improve production efficiency and product quality.

[0103] Data acquisition and preprocessing: Obtain multi-dimensional operation data of the melting furnace group, including natural gas usage information, feeding plan and record of molten aluminum quality inspection, furnace group operation data, etc. Preprocess the acquired data to ensure the quality and consistency of the data.

[0104] Feature extraction: Based on the random forest algorithm, key features of the preprocessed multi-dimensional data of the smelting furnace are extracted to determine the corresponding key features of the smelting furnace.

[0105] Operating condition prediction model: The extracted key features are input into a pre-constructed operating condition prediction model (such as Figure 3 the LSTM neural network model exemplified in it), and the operating condition law results of the molten aluminum are predicted. The time series conversion technology is used to convert the key features into the input format of the time series neural network, and the network structure parameters and training parameters are adjusted through the parameter optimization method. The model performance is verified based on multi-dimensional evaluation indicators, the switching state of the molten aluminum in the future period is predicted, and the operating condition law results of the molten aluminum are output.

[0106] Adaptive tuning parameter calculation: According to the operating condition law results of the molten aluminum, calculate the time interval between the corresponding smelting furnace and the preset operating condition, and dynamically output the adaptive tuning parameters as a constraint condition.

[0107] If the time interval is less than the first preset threshold, increase the opening of the gas valve and the control value of the burner air motor; if the time interval is between the first preset threshold and the second preset threshold, keep the current parameters; if the time interval is greater than the second preset threshold, decrease the opening of the gas valve and the control value of the burner air motor.

[0108] Multi-objective particle swarm optimization model: Initialize the multi-objective particle swarm optimization model, where the preset penalty function is used as the process constraint condition.

[0109] According to the input adaptive tuning parameters of each smelting furnace, the particle velocity and position are iteratively updated through the multi-objective particle swarm optimization model, the historical optimal position and the global optimal position are recorded, and the optimal parameter combination of the smelting furnace group is output.

[0110] Coordinated control of the smelting furnace group: According to the optimal parameter combination of the smelting furnace group, dynamically allocate the flow path of the molten aluminum between the smelting furnace and the holding furnace. Real-time monitor the molten aluminum capacity status of each holding furnace, and automatically adjust the flow distribution of the aluminum scrap furnace when it is detected that the capacity is insufficient. Based on the closed-loop feedback mechanism, optimize the production control parameters between the smelting furnace and the holding furnace group.

[0111] Production execution: According to the optimal parameter combination, execute production operations, including opening the burner, adjusting the opening of the gas valve air motor, valve opening, etc. During the production process, real-time monitor the smoke exhaust detection signal, molten aluminum quality detection, furnace group operation data, natural gas usage data, etc. to ensure the stability of the production process and product quality.

[0112] It can be understood that this application can effectively optimize the parameters of the smelting furnace group, improve production efficiency and product quality, and at the same time reduce energy consumption and production costs.

[0113] For example, in order to help understand the implementation process of the method for optimizing the parameters of the smelting furnace group obtained by combining this embodiment with the above-mentioned embodiment 1, please refer to Figure 4 , Figure 4 A brief flow chart of a method for tuning parameters of a melting furnace group is provided, specifically: The multi-dimensional operation data of the smelting furnace group is collected through sensors, including instantaneous natural gas flow, aluminum liquid temperature, feeding time, etc. The collected data is cleaned, abnormal values ​​(such as abnormally high values ​​of instantaneous natural gas flow) are eliminated, and the Min-Max normalization method is used to standardize the data.

[0114] The random forest algorithm is used to screen the preprocessed data, evaluate the importance of each feature, and retain the top N key features (such as aluminum liquid temperature and natural gas instantaneous flow rate).

[0115] The key features are input into the working condition prediction model based on LSTM neural network, and the time series data is converted into the model input format through sliding window technology. The model hyperparameters (such as the number of hidden layers is 2 and the number of hidden units is 64) are optimized by grid search to predict the switching state of aluminum liquid in the next T minutes.

[0116] According to the prediction results, the time interval from each smelting furnace to the "next aluminum liquid release open state" is calculated, and the gas valve opening and burner air motor control value are dynamically adjusted; The adaptive tuning parameters of each melting furnace are input into the multi-objective particle swarm optimization model, the model is initialized and the penalty function is set to handle process constraints (such as temperature fluctuations and standing time). By iteratively updating the particle speed and position, the optimal parameter combination of the melting furnace group is finally output to achieve global optimization of the production rhythm between furnace groups.

[0117] According to the optimized parameter combination, the flow path of aluminum liquid between the smelting furnace and the static furnace is dynamically allocated. The aluminum liquid capacity status of each static furnace is monitored in real time. When insufficient capacity is detected, the flow distribution of the aluminum chip furnace is automatically adjusted. Through the closed-loop feedback mechanism, the production control parameters between the smelting furnace and the static furnace are further optimized.

[0118] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the parameter tuning method of the smelting furnace group of the present application. More simple transformations based on this technical concept are all within the protection scope of the present application.

[0119] This application also provides a smelting furnace group parameter tuning device, please refer to Figure 5 , the smelting furnace group parameter tuning device comprises: The acquisition module 10 is used to acquire multi-dimensional data of the smelting furnace group, confirm the working conditions corresponding to each smelting furnace according to a preset working condition prediction model, and then determine the adaptive tuning parameters of each smelting furnace. The optimization module 20 is used to input the adaptive tuning parameters of each smelting furnace into a preset multi-objective particle swarm optimization model for group optimization of the smelting furnace group, and obtain the optimal parameter combination of the smelting furnace group.

[0120] And / or, the optimization module 20 includes: The first initialization module is used to initialize the multi-objective particle swarm optimization model, where a preset penalty function is used as the process constraint condition. The first iteration module is used to iteratively update the particle velocity and position through the multi-objective particle swarm optimization model according to the input adaptive tuning parameters of each smelting furnace, record the historical optimal position and the global optimal position, and output the optimal parameter combination of the smelting furnace group.

[0121] And / or, the acquisition module 10 includes: The first acquisition module is used to acquire the multi-dimensional operation data of the smelting furnace group and preprocess it to obtain the multi-dimensional data of the smelting furnace. The first extraction module is used to extract key features from the multi-dimensional data of the smelting furnace based on the random forest algorithm and determine the corresponding key features of the smelting furnace. The first prediction module is used to input the key features of the smelting furnace into a preset working condition prediction model and predict the result of the aluminum liquid working condition law. The first calculation module is used to calculate the time interval between the corresponding smelting furnace and the preset working condition according to the result of the aluminum liquid working condition law, and dynamically output the adaptive tuning parameter as a constraint condition.

[0122] And / or, the first prediction module includes: The first conversion module is used to convert the key features of the smelting furnace into the input format of the time series neural network by using the time series conversion technology. The first adjustment module is used to adjust the structure parameters and training parameters of the time series neural network by using the parameter optimization method. The second prediction module is used to verify the model performance based on multi-dimensional evaluation indexes, predict the on-off state of the aluminum liquid in the future period, and output the result of the aluminum liquid working condition law.

[0123] And / or, the first calculation module includes: The first increase module is used to increase the opening degree of the gas valve and the control value of the burner air motor if the time interval is less than the first preset threshold. The first holding module is used to keep the current parameters if the time interval is between the first preset threshold and the second preset threshold. A first reduction module, configured to reduce the opening degree of the gas valve and the control value of the burner air motor if the time interval is greater than a second preset threshold.

[0124] And / or, the melting furnace group parameter tuning device includes: A first distribution module, configured to dynamically distribute the flow path of the molten aluminum between the melting furnace and the holding furnace according to the optimal parameter combination of the melting furnace group; A first monitoring module, configured to monitor the molten aluminum capacity status of each holding furnace in real time, and automatically adjust the flow distribution of the aluminum scrap furnace when it detects insufficient capacity; A first optimization module, configured to optimize the production control parameters between the melting furnace and the holding furnace groups based on a closed-loop feedback mechanism.

[0125] The melting furnace group parameter tuning device provided by the present application adopts the melting furnace group parameter tuning method in the above embodiment, and can solve the technical problem of being unable to tune the parameters of the melting furnace group. Compared with the prior art, the beneficial effects of the melting furnace group parameter tuning device provided by the present application are the same as those of the melting furnace group parameter tuning method provided by the above embodiment, and other technical features in the melting furnace group parameter tuning device are the same as the features disclosed in the above embodiment method, which will not be elaborated here.

[0126] The present application provides a melting furnace group parameter tuning device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the melting furnace group parameter tuning method in the first embodiment above.

[0127] Next, refer to Figure 6 , which shows a schematic structural diagram of a melting furnace group parameter tuning device suitable for implementing the embodiments of the present application. The melting furnace group parameter tuning device in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, tablet computers, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The shown melting furnace group parameter tuning device is only an example, and should not bring any limitations to the functions and usage scopes of the embodiments of the present application.

[0128] As Figure 6As shown, the melting furnace group parameter tuning device may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM: Read Only Memory) 1002 or the program loaded from the storage device 1003 into the random access memory (RAM: Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of the melting furnace group parameter tuning device are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. The input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the melting furnace group parameter tuning device to communicate with other devices wirelessly or wiredly to exchange data. Although the figure shows a melting furnace group parameter tuning device having various systems, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems may be implemented or had alternatively.

[0129] Specifically, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device, or installed from the storage device 1003, or installed from the ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are executed.

[0130] The melting furnace group parameter tuning device provided by the present application adopts the melting furnace group parameter tuning method in the above-mentioned embodiment, and can solve the technical problem that the parameters of the melting furnace group cannot be tuned. Compared with the prior art, the beneficial effects of the melting furnace group parameter tuning device provided by the present application are the same as those of the melting furnace group parameter tuning method provided by the above-mentioned embodiment, and the other technical features in the melting furnace group parameter tuning device are the same as the features disclosed in the method of the previous embodiment, and will not be elaborated here.

[0131] It should be understood that each part disclosed in this application can be implemented by hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0132] As described above, the above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0133] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the smelting furnace group parameter tuning method in the above embodiments.

[0134] The computer-readable storage medium provided by this application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination of the above. More specific examples of computer-readable storage media can include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or combined with an instruction execution system, device, or device. The program code contained on the computer-readable storage medium can be transmitted by any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0135] The above computer-readable storage medium can be included in the smelting furnace group parameter tuning device; or it can exist separately without being assembled into the smelting furnace group parameter tuning device.

[0136] The above computer-readable storage medium carries one or more programs, which, when executed by the smelting furnace group parameter tuning device, cause the smelting furnace group parameter tuning device to: obtain multi-dimensional data of the smelting furnace group, confirm the working conditions corresponding to each smelting furnace according to a preset working condition prediction model, and then determine the adaptive tuning parameters of each smelting furnace; input the adaptive tuning parameters of each smelting furnace into a preset multi-objective particle swarm optimization model for group optimization of the smelting furnace group to obtain an optimal parameter combination for the smelting furnace group.

[0137] Computer program code for performing the operations of the present application may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0138] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that, in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system for performing the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0139] The modules described in the embodiments of the present application may be implemented in software or in hardware. Wherein, the name of the module does not constitute a limitation to the unit itself in some cases.

[0140] The readable storage medium provided by this application is a computer-readable storage medium. The computer-readable storage medium stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned parameter tuning method for the smelting furnace group, and can solve the technical problem of being unable to perform parameter tuning on the smelting furnace group. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by this application are the same as those of the parameter tuning method for the smelting furnace group provided in the above embodiments, and will not be elaborated here.

[0141] This application also provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the steps of the parameter tuning method for the smelting furnace group as described above.

[0142] The computer program product provided by this application can solve the technical problem of being unable to perform parameter tuning on the smelting furnace group. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as those of the parameter tuning method for the smelting furnace group provided in the above embodiments, and will not be elaborated here.

[0143] The above are only some embodiments of this application, and do not limit the patent scope of this application. Any equivalent structural transformation made by using the content of the specification and drawings of this application under the technical concept of this application, or any direct / indirect application in other related technical fields, is included in the patent protection scope of this application.

Claims

1. A method for optimizing parameters of a smelting furnace group, characterized in that: The method includes: Obtain multi-dimensional data of the smelting furnace group, confirm the corresponding working conditions of each smelting furnace according to the preset working condition prediction model, and then determine the adaptive tuning parameters of each smelting furnace; The adaptive tuning parameters of each smelting furnace are input into a preset multi-objective particle swarm optimization model to perform group optimization of the smelting furnace group to obtain the optimal parameter combination of the smelting furnace group.

2. The method according to claim 1, characterized in that The step of inputting the adaptive tuning parameters of each smelting furnace into a preset multi-objective particle swarm optimization model to perform group optimization of the smelting furnace group to obtain the optimal parameter combination of the smelting furnace group includes: Initializing a multi-objective particle swarm optimization model, wherein a preset penalty function is used as a process constraint; According to the input adaptive tuning parameters of each smelting furnace, the particle speed and position are iteratively updated through the multi-objective particle swarm optimization model, the historical optimal position and the global optimal position are recorded, and the optimal parameter combination of the smelting furnace group is output.

3. The method according to claim 1, characterized in that The step of obtaining multi-dimensional data of the smelting furnace group, confirming the working condition corresponding to each smelting furnace according to a preset working condition prediction model, and then determining the adaptive tuning parameters of each smelting furnace includes: Acquire multi-dimensional operation data of the smelting furnace group, and pre-process to obtain multi-dimensional data of the smelting furnace; Extracting key features from the multi-dimensional data of the smelting furnace based on a random forest algorithm to determine corresponding key features of the smelting furnace; Inputting the key characteristics of the smelting furnace into a preset operating condition prediction model to predict the operating condition of the aluminum liquid; According to the results of the aluminum liquid working condition law, the time interval between the corresponding smelting furnace and the preset working condition is calculated, and the adaptive tuning parameters are dynamically output as constraint conditions.

4. The method according to claim 3, characterized in that The step of inputting the key characteristics of the smelting furnace into a preset working condition prediction model to predict the working condition law of the aluminum liquid includes: Using time series conversion technology to convert the key features of the smelting furnace into an input format of a time series neural network; Adjusting the structural parameters and training parameters of the temporal neural network by a parameter optimization method; The model performance is verified based on multi-dimensional evaluation indicators, the switching state of the aluminum liquid in the future period is predicted, and the results of the aluminum liquid working condition law are output.

5. The method according to claim 3, characterized in that The step of calculating the time interval between the corresponding smelting furnace and the preset working condition according to the aluminum liquid working condition law result and dynamically outputting the adaptive tuning parameters as a constraint condition comprises: If the time interval is less than the first preset threshold, increasing the gas valve opening and the burner air motor control value; If the time interval is between the first preset threshold and the second preset threshold, the current parameter is maintained; If the time interval is greater than the second preset threshold, the gas valve opening and the burner air motor control value are reduced.

6. The method according to claim 1, characterized in that The method further comprises a smelting furnace group coordinated control step: Dynamically allocating the flow path of the aluminum liquid between the smelting furnace and the static furnace according to the optimal parameter combination of the smelting furnace group; Monitor the aluminum liquid capacity of each static furnace in real time, and automatically adjust the flow distribution of the aluminum chip furnace when insufficient capacity is detected; Based on the closed-loop feedback mechanism, the production control parameters between the melting furnace and the static furnace are optimized.

7. A device for optimizing parameters of a smelting furnace group, characterized in that: The device comprises: An acquisition module is used to acquire multi-dimensional data of the smelting furnace group, confirm the working conditions corresponding to each smelting furnace according to a preset working condition prediction model, and then determine the adaptive tuning parameters of each smelting furnace; The optimization module is used to input the adaptive tuning parameters of each smelting furnace into a preset multi-objective particle swarm optimization model to perform group optimization of the smelting furnace group to obtain the optimal parameter combination of the smelting furnace group.

8. A smelting furnace group parameter tuning device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the method for tuning parameters of a smelting furnace group according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the method for optimizing parameters of a smelting furnace group according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the method for tuning parameters of a smelting furnace group according to any one of claims 1 to 6 are implemented.

Images