A method and system for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback

Abstract

The present invention relates to the technical field of aluminum melting furnace equipment, and in particular to a method and system for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback. A hidden Markov prediction model is constructed; actual characteristic parameters of each smelting parameter adjustment module are collected, and the collected actual characteristic parameters are imported into a database; each actual characteristic parameter in the database is classified and processed to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period; the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period is imported into the hidden Markov prediction model for prediction, and the probability value of various abnormal smelting events occurring in the aluminum melting furnace equipment is obtained; if the probability value of the abnormal smelting event occurring in the aluminum melting furnace equipment is greater than the preset probability value, then a plan for adjusting the aluminum melting parameters is obtained, which can effectively prevent abnormal conditions during the smelting process and significantly improve production efficiency and aluminum liquid quality.

Description

Technical Field

[0001] The present invention relates to the technical field of aluminum melting furnace equipment, and in particular to a method and system for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback. Background Art

[0002] The traditional aluminum melting furnace equipment smelting parameter adjustment mainly relies on manual experience and simple control loops, which has many disadvantages. For example, manual experience depends on the proficiency of the operator, and it is difficult to ensure the stability and consistency of the smelting process; the simple control loop can only adjust a single parameter, which is difficult to deal with the complex mutual influence and coupling relationship in the smelting process, and it is easy to cause the smelting quality to decline or even safety accidents. In addition, the occurrence of various abnormal situations in the smelting process is often random and complex, and traditional adjustment methods are difficult to predict and prevent in advance, resulting in low smelting efficiency and waste of resources.

[0003] With the rapid development of artificial intelligence technology, especially the maturity of technologies such as hidden Markov models and knowledge graphs, new ideas and approaches have been provided for intelligent smelting parameter adjustment. Hidden Markov models can effectively model and predict time series data, and can use historical smelting data to build prediction models to predict abnormal situations that may occur in the smelting process in advance, thereby achieving early warning and prevention. Knowledge graphs can associate and organize various parameters, processes, faults and other information in the smelting process to form a structured knowledge base, providing a decision-making basis for intelligent smelting parameter adjustment.

[0004] However, most of the existing smelting parameter adjustment methods for aluminum melting furnace equipment lack intelligent feedback mechanisms, making it difficult to make dynamic adjustments based on real-time monitoring data. For example, some prediction models based on machine learning can only provide the probability of abnormal events, but cannot provide specific adjustment plans; some adjustment methods based on expert systems lack real-time data support and are difficult to adapt to dynamic changes in the smelting process. Therefore, there is an urgent need for a smelting parameter adjustment method that can combine historical data, real-time monitoring data, and expert knowledge to achieve precise adjustment of smelting parameters of aluminum melting furnace equipment, improve smelting efficiency and product quality, reduce energy consumption and pollution emissions, and ensure the safety and stability of the smelting process. Summary of the invention

[0005] The present invention overcomes the deficiencies of the prior art and provides a method and system for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is:

[0007] The first aspect of the present invention discloses a method for adjusting smelting parameters of an aluminum melting furnace based on intelligent feedback, comprising the following steps:

[0008] Obtaining historical characteristic parameters corresponding to each smelting parameter adjustment module of the aluminum melting furnace equipment within a preset time period before various abnormal smelting events occur, and constructing a hidden Markov prediction model according to the historical characteristic parameters;

[0009] During the operation of the aluminum melting furnace equipment, the actual characteristic parameters of each smelting parameter adjustment module are continuously collected at a number of preset time nodes, a database is constructed, and the collected actual characteristic parameters are imported into the database;

[0010] After the collection is completed, each actual characteristic parameter in the database is classified and processed to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period;

[0011] Importing the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period into the hidden Markov prediction model for prediction, and obtaining the probability values ​​of various smelting abnormal events occurring in the aluminum melting furnace equipment;

[0012] The probability values ​​of various abnormal smelting events occurring in the aluminum melting furnace equipment are compared with the preset probability values. If the probability values ​​of various abnormal smelting events occurring in the aluminum melting furnace equipment are not greater than the preset probability values, no adjustment is performed on the aluminum melting furnace equipment.

[0013] If the probability value of one or more abnormal smelting events occurring in the aluminum melting furnace equipment is greater than the preset probability value, the aluminum melting parameter adjustment measure plan is obtained from the knowledge graph constructed in advance, and the smelting parameters of the aluminum melting furnace equipment are adjusted according to the obtained aluminum melting parameter adjustment measure plan.

[0014] Preferably, the historical characteristic parameters corresponding to each smelting parameter adjustment module of the aluminum melting furnace equipment within a preset time period before various abnormal smelting events occur are obtained, and a hidden Markov prediction model is constructed according to the historical characteristic parameters, specifically:

[0015] Obtain a smelting log of an aluminum smelting furnace device, and obtain various abnormal smelting events that occurred in the aluminum smelting furnace device during historical smelting operations according to the smelting log;

[0016] Acquire, according to the smelting log, historical characteristic parameters corresponding to each smelting parameter adjustment module within a preset time period before various smelting abnormal events occur;

[0017] Aggregate and process the historical characteristic parameters corresponding to each smelting parameter adjustment module within a preset time period before various smelting abnormal events to obtain a pre-abnormal characteristic parameter set corresponding to various smelting abnormal events;

[0018] Construct a hidden Markov model, take various abnormal smelting events as observation nodes, and take the pre-abnormal characteristic parameter sets corresponding to various abnormal smelting events as state nodes;

[0019] Connect each observation node with the corresponding state node in a directed ring manner to generate a topological structure diagram between each observation node and each state node;

[0020] Calculate the Euclidean distance between each observation node and each state node according to the topological structure diagram, and determine the adjacency relationship between each observation node and each state node according to the corresponding Euclidean distance value;

[0021] An adjacency matrix is ​​constructed according to the adjacency relationship between each observation node and each state node, and the adjacency matrix is ​​imported into the hidden Markov model for training and learning until the model parameters meet the preset requirements, thereby obtaining a hidden Markov prediction model.

[0022] Preferably, each observation node is connected to the corresponding state node in a directed manner based on a directed ring, and a topological structure diagram between each observation node and each state node is generated, specifically:

[0023] Determine the order of observation nodes and state nodes based on the occurrence order of abnormal smelting events and the time series of parameter changes;

[0024] Traverse each observation node, establish directed connections between each observation node and state node according to the order of each observation node and state node, and obtain directed connection relationships between each observation node and state node;

[0025] Connect each observation node to its previous state node, and at the same time, connect each state node to its previous observation node to form a closed loop, and obtain a closed loop relationship between each observation node and state node;

[0026] According to the directed connection relationship and closed-loop relationship between each observation node and state node, the topological structure diagram between each observation node and each state node is drawn using the Python library;

[0027] The nodes in the topological structure diagram represent observation nodes and state nodes, and the directed edges represent the connection relationship between observation nodes and state nodes.

[0028] Preferably, the actual characteristic parameters in the database are classified and processed to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period, specifically:

[0029] Constructing a regression tree model, obtaining the number of each smelting parameter adjustment module, and determining a number of cutting nodes according to the number of each smelting parameter adjustment module;

[0030] Cutting the regression tree model based on the cutting node to cut out a plurality of branches in the regression tree model;

[0031] Acquire data characteristic information of characteristic parameters corresponding to each smelting parameter adjustment module, and assign a data characteristic label to each tree branch according to the data characteristic information of characteristic parameters corresponding to each smelting parameter adjustment module;

[0032] Obtain each actual feature parameter in the database, and calculate the cosine similarity between each actual feature parameter in the database and each data feature label in the regression tree model based on the cosine similarity algorithm;

[0033] If the cosine similarity between a certain actual feature parameter in the database and each data feature label in the regression tree model is not greater than a preset similarity threshold, the actual feature parameter in the database is marked as noise data and the noise data is filtered out;

[0034] If the cosine similarity between a certain actual feature parameter in the database and a certain data feature label in the regression tree model is greater than a preset similarity threshold, the actual feature parameter is assigned to the branch to which the data feature label belongs;

[0035] Repeat the above steps until all actual feature parameters in the database are assigned and processed, and obtain the actual feature parameter regression tree model.

[0036] The actual characteristic parameters in the database are classified and processed to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period, and the following steps are also included:

[0037] Calculating the Euclidean distance between each actual feature parameter on each branch and the corresponding branch center in the actual feature parameter regression tree model;

[0038] If the Euclidean distance between a certain actual feature parameter on a certain branch and the center of the corresponding branch is greater than a preset distance threshold, the actual feature parameter is screened out on the branch, and the actual feature parameter is simulated and distributed to the remaining branches;

[0039] Calculate the Euclidean distance between the actual characteristic parameters assigned to the remaining branches and the center of the remaining branches;

[0040] If the Euclidean distance between the actual characteristic parameter assigned to the remaining branches by the simulation and the center of the branches of the remaining branches is still greater than the preset distance, the actual characteristic parameter is marked as noise data and the noise data is filtered out;

[0041] If the Euclidean distance between the actual characteristic parameter assigned to the remaining branches by the simulation and the center of a branch of the remaining branches is not greater than the preset distance, it means that the actual characteristic parameter belongs to the remaining branches, and the actual characteristic parameter is reallocated to the remaining branches;

[0042] Repeat the above steps until the actual feature parameters on each branch in the actual feature parameter regression tree model are evaluated, and the final actual feature parameter regression tree model is obtained;

[0043] The final actual characteristic parameter regression tree model is pruned to obtain an actual characteristic parameter set of each smelting parameter adjustment module within a preset time period.

[0044] Preferably, if the probability value of one or more abnormal smelting events occurring in the aluminum melting furnace equipment is greater than a preset probability value, an aluminum melting parameter adjustment measure scheme is obtained in the knowledge graph constructed in advance, and the smelting parameters of the aluminum melting furnace equipment are adjusted according to the obtained aluminum melting parameter adjustment measure scheme, specifically:

[0045] Pre-prepare the aluminum melting parameter adjustment measures corresponding to various abnormal smelting events in the aluminum melting furnace equipment;

[0046] Construct a knowledge graph, import the extracted prefabricated aluminum melting parameter adjustment measures corresponding to various abnormal smelting events in the aluminum melting furnace equipment into the knowledge graph, and regularly update the knowledge graph;

[0047] If the probability value of one or more abnormal smelting events occurring in the aluminum melting furnace equipment is greater than a preset probability value, a search matching term is constructed according to the abnormal smelting events corresponding to the probability values ​​greater than the preset probability values;

[0048] Importing the search matching words into the knowledge graph for search matching, and obtaining relevant aluminum melting parameter adjustment measures;

[0049] The obtained aluminum melting parameter adjustment measure scheme is fed back to the control terminal of the aluminum melting furnace equipment, so as to adjust the smelting parameters of the aluminum melting furnace equipment based on the obtained aluminum melting parameter adjustment measure scheme.

[0050] A second aspect of the present invention provides an aluminum melting furnace device, which is applied to any one of the aluminum melting furnace device smelting parameter adjustment methods based on intelligent feedback, comprising:

[0051] The main body of the aluminum melting furnace is used to melt aluminum;

[0052] The furnace temperature adjustment module is used to control the temperature of the aluminum melting furnace to keep the melt temperature within a preset range to meet different smelting process requirements;

[0053] The furnace pressure regulating module is used to control and adjust the pressure in the aluminum melting furnace so that the pressure in the furnace remains within the preset range to ensure the safety and stability of the smelting process;

[0054] Stirring module, used to stir the melt to make the melt temperature more uniform and promote the full dissolution and reaction of alloy elements;

[0055] The data collection and storage module is responsible for collecting data collected by preset sensors installed inside the main body of the aluminum melting furnace, and storing the collected data in a database to provide a data basis for subsequent analysis and prediction; wherein the preset sensors include vibration sensors, temperature sensors and pressure sensors.

[0056] The third aspect of the present invention provides a smelting parameter adjustment system for aluminum melting furnace equipment based on intelligent feedback, the smelting parameter adjustment system for aluminum melting furnace equipment includes a memory and a processor, the memory stores a smelting parameter adjustment method program for aluminum melting furnace equipment, when the smelting parameter adjustment method program for aluminum melting furnace equipment is executed by the processor, any one of the steps of the smelting parameter adjustment method for aluminum melting furnace equipment is implemented.

[0057] The present invention solves the technical defects existing in the background technology, and has the following beneficial effects: constructing a hidden Markov prediction model; collecting actual characteristic parameters of each smelting parameter adjustment module, and importing the collected actual characteristic parameters into a database; classifying and processing each actual characteristic parameter in the database, and obtaining the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period; importing the actual characteristic parameter set of each smelting parameter adjustment module within the preset time period into the hidden Markov prediction model for prediction, and obtaining the probability value of various smelting abnormal events occurring in the aluminum melting furnace equipment; if the probability value of one or more smelting abnormal events occurring in the aluminum melting furnace equipment is greater than the preset probability value, obtaining the aluminum melting parameter adjustment measure plan in the knowledge graph constructed in advance, and the aluminum melting furnace equipment smelting parameter adjustment method based on intelligent feedback can not only effectively prevent abnormal conditions in the smelting process, but also significantly improve production efficiency and aluminum liquid quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0059] Figure 1 This is an overall method flow chart of the smelting parameter adjustment method of the aluminum melting furnace equipment;

[0060] Figure 2 This is a partial method flow chart of the smelting parameter adjustment method of the aluminum melting furnace equipment;

[0061] Figure 3 This is a schematic diagram of the equipment structure of this aluminum melting furnace equipment. DETAILED DESCRIPTION

[0062] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0063] 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. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0064] The first aspect of the present invention discloses a method for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback, such as Figure 1 As shown, the following steps are included:

[0065] S102: Obtaining historical characteristic parameters corresponding to each smelting parameter adjustment module of the aluminum melting furnace equipment within a preset time period before various abnormal smelting events occur, and constructing a hidden Markov prediction model according to the historical characteristic parameters;

[0066] S104: During the operation of the aluminum melting furnace equipment, the actual characteristic parameters of each smelting parameter adjustment module are continuously collected at a number of preset time nodes, a database is constructed, and the collected actual characteristic parameters are imported into the database;

[0067] S106: After the collection is completed, the actual characteristic parameters in the database are classified and processed to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period;

[0068] S108: importing the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period into the hidden Markov prediction model for prediction, and obtaining the probability value of various smelting abnormal events occurring in the aluminum melting furnace equipment;

[0069] S110: comparing the probability values ​​of various abnormal smelting events occurring in the aluminum melting furnace equipment with the preset probability values, and if the probability values ​​of various abnormal smelting events occurring in the aluminum melting furnace equipment are not greater than the preset probability values, no adjustment process is performed on the aluminum melting furnace equipment;

[0070] S112: If the probability value of one or more abnormal smelting events occurring in the aluminum melting furnace equipment is greater than a preset probability value, an aluminum melting parameter adjustment measure plan is obtained from a knowledge graph constructed in advance, and the smelting parameters of the aluminum melting furnace equipment are adjusted according to the obtained aluminum melting parameter adjustment measure plan.

[0071] Among them, the characteristic parameters include furnace temperature, furnace pressure, flow meter readings, chemical composition (such as aluminum content, silicon content, iron content), power supply current and voltage, vibration sensor readings, noise level, smelting duration, and temperature distribution in the furnace; the abnormal smelting events include aluminum liquid oxidation caused by excessive temperature, equipment overload caused by excessive pressure, product quality failure caused by composition deviation, equipment operation instability caused by current fluctuations, smelting interruption caused by equipment mechanical failure, and smelting parameter instability events.

[0072] It should be noted that, first, the historical characteristic parameters of each smelting parameter adjustment module of the aluminum melting furnace equipment within a period of time before the occurrence of smelting abnormality are collected. According to these historical characteristic parameters, a hidden Markov model is constructed. The hidden Markov model is a statistical model that is widely used in pattern recognition, natural language processing and other fields, and is particularly suitable for modeling of sequence data. During the operation of the aluminum melting furnace equipment, the actual characteristic parameters of each smelting parameter adjustment module are regularly collected, and these data are stored in the database, and these actual characteristic parameters are then used for subsequent analysis and prediction. The actual characteristic parameters in the database are classified and processed to form a set of actual characteristic parameters within a specific time range. These actual characteristic parameter sets are input into the pre-constructed hidden Markov model to predict the possibility of smelting abnormal events in the aluminum melting furnace equipment in the future. The predicted probability value of the occurrence of the smelting abnormal event will be compared with the preset threshold. If all predicted probability values ​​do not exceed the preset threshold, there is no need to adjust the smelting parameters. On the contrary, if the predicted probability of occurrence of certain smelting abnormal events exceeds the preset threshold, measures need to be taken to intervene. At this time, the system will search for the corresponding smelting parameter adjustment measures from the existing knowledge graph and adjust the smelting parameters accordingly. This smelting parameter adjustment method for aluminum melting furnace equipment based on intelligent feedback can not only effectively prevent abnormal conditions during the smelting process, but also significantly improve production efficiency and aluminum liquid quality.

[0073] Preferably, the historical characteristic parameters corresponding to each smelting parameter adjustment module of the aluminum melting furnace equipment within a preset time period before various abnormal smelting events occur are obtained, and a hidden Markov prediction model is constructed according to the historical characteristic parameters, such as Figure 2 As shown, specifically:

[0074] S202: Obtain a smelting log of the aluminum smelting furnace equipment, and obtain various abnormal smelting events that occurred in the aluminum smelting furnace equipment during the historical smelting work process according to the smelting log;

[0075] S204: acquiring, according to the smelting log, historical characteristic parameters corresponding to each smelting parameter adjustment module within a preset time period before various smelting abnormal events occur;

[0076] S206: Aggregating and processing the historical characteristic parameters corresponding to each smelting parameter adjustment module within a preset time period before various smelting abnormal events to obtain a pre-abnormal characteristic parameter set corresponding to various smelting abnormal events;

[0077] S208: constructing a hidden Markov model, taking various abnormal smelting events as observation nodes, and taking pre-abnormal feature parameter sets corresponding to various abnormal smelting events as state nodes;

[0078] S210: Directly connect each observation node with the corresponding state node based on a directed ring method to generate a topological structure diagram between each observation node and each state node;

[0079] S212: Calculate the Euclidean distance between each observation node and each state node according to the topological structure diagram, and determine the adjacency relationship between each observation node and each state node according to the corresponding Euclidean distance value;

[0080] S214: constructing an adjacency matrix according to the adjacency relationship between each observation node and each state node, and importing the adjacency matrix into the hidden Markov model for training and learning until the model parameters meet the preset requirements, thereby obtaining a hidden Markov prediction model.

[0081] It should be noted that all smelting logs recorded by the aluminum melting furnace equipment during the historical smelting process are collected, including logs of normal operations and abnormal events. All smelting abnormal events that have occurred are extracted from the smelting logs, and the historical characteristic parameters of each smelting parameter adjustment module in the preset time period before each smelting abnormal event occurs are obtained. The historical characteristic parameters of each smelting parameter adjustment module in the preset time period before various smelting abnormal events are aggregated and processed to form a "pre-abnormal characteristic parameter set" corresponding to each abnormal event. Various smelting abnormal events are used as observation nodes, and the "pre-abnormal characteristic parameter set" corresponding to these abnormal events is used as a state node. The observation nodes and state nodes are connected in a directed ring manner to generate a topological structure diagram. According to the topological structure diagram, the Euclidean distance value between each observation node and the state node is calculated, and the adjacency relationship between each observation node and the state node is determined according to these distance values. An adjacency matrix is ​​constructed according to the adjacency relationship, and the adjacency matrix is ​​imported into the hidden Markov model for training and learning until the model parameters meet the preset requirements, and finally a hidden Markov prediction model is obtained. By learning and analyzing historical data, the model can more accurately predict abnormal events that may occur during the smelting process, provide early warnings, and reduce the risk of accidents. The entire process relies on advanced data analysis and machine learning technologies to achieve automated and intelligent management of the smelting process, reduce manual intervention, and improve the system's response speed and accuracy.

[0082] Preferably, each observation node is connected to the corresponding state node in a directed manner based on a directed ring, and a topological structure diagram between each observation node and each state node is generated, specifically:

[0083] Determine the order of observation nodes and state nodes based on the occurrence order of abnormal smelting events and the time series of parameter changes;

[0084] Traverse each observation node, establish directed connections between each observation node and state node according to the order of each observation node and state node, and obtain directed connection relationships between each observation node and state node;

[0085] Connect each observation node to its previous state node, and at the same time, connect each state node to its previous observation node to form a closed loop, and obtain a closed loop relationship between each observation node and state node;

[0086] According to the directed connection relationship and closed-loop relationship between each observation node and state node, the topological structure diagram between each observation node and each state node is drawn using the Python library;

[0087] The nodes in the topological structure diagram represent observation nodes and state nodes, and the directed edges represent the connection relationship between observation nodes and state nodes.

[0088] It should be noted that the order of observation nodes and state nodes is determined based on the order of occurrence of abnormal smelting events and the time series of parameter changes. This step ensures that the arrangement of nodes conforms to the time logic of the actual smelting process. Traverse each observation node and establish directed connections between them according to the order of each observation node and state node. This step forms a directed connection relationship between observation nodes and state nodes. Connect each observation node to its previous state node, and connect each state node to its previous observation node to form a closed loop. This step ensures the closed-loop characteristics of the model and reflects the cyclic nature of the smelting process. Use Python libraries (such as NetworkX, etc.) to draw a topological structure diagram between each observation node and each state node. The nodes in the topological structure diagram represent observation nodes and state nodes, and the directed edges represent the connection relationship between nodes. The directed ring topological structure diagram constructed by the above steps can not only clearly show the node relationship in the smelting process, but also enhance the interpretability and accuracy of the model, thereby better supporting abnormal prediction and real-time monitoring in the smelting process.

[0089] Preferably, the actual characteristic parameters in the database are classified and processed to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period, specifically:

[0090] Constructing a regression tree model, obtaining the number of each smelting parameter adjustment module, and determining a number of cutting nodes according to the number of each smelting parameter adjustment module;

[0091] Cutting the regression tree model based on the cutting node to cut out a plurality of branches in the regression tree model;

[0092] Acquire data characteristic information of characteristic parameters corresponding to each smelting parameter adjustment module, and assign a data characteristic label to each tree branch according to the data characteristic information of characteristic parameters corresponding to each smelting parameter adjustment module;

[0093] Obtain each actual feature parameter in the database, and calculate the cosine similarity between each actual feature parameter in the database and each data feature label in the regression tree model based on the cosine similarity algorithm;

[0094] If the cosine similarity between a certain actual feature parameter in the database and each data feature label in the regression tree model is not greater than a preset similarity threshold, the actual feature parameter in the database is marked as noise data and the noise data is filtered out;

[0095] If the cosine similarity between a certain actual feature parameter in the database and a certain data feature label in the regression tree model is greater than a preset similarity threshold, the actual feature parameter is assigned to the branch to which the data feature label belongs;

[0096] Repeat the above steps until all actual feature parameters in the database are assigned and processed, and obtain the actual feature parameter regression tree model.

[0097] It should be noted that, by constructing a regression tree model, and determining the number of each smelting parameter adjustment module, several cutting nodes are determined according to the number of smelting parameter adjustment modules. Based on these cutting nodes, the regression tree model is cut, so as to cut out several branches in the regression tree model. The data feature information of the characteristic parameters corresponding to each smelting parameter adjustment module is obtained, and a data feature label is assigned to each branch according to these data feature information. The introduction of the data feature label enables the characteristic parameters to be classified into different branches according to their similarity, so as to more accurately extract and identify the characteristic parameters of different smelting parameter adjustment modules. The actual characteristic parameters in the database are obtained, and the cosine similarity algorithm is used to calculate the cosine similarity between each actual characteristic parameter in the database and each data feature label in the regression tree model. If the cosine similarity between a certain actual characteristic parameter in the database and each data feature label in the regression tree model is not greater than the preset similarity threshold, the actual characteristic parameter is calibrated as noise data and filtered out. The filtering of noise data enhances the robustness of the model, reduces the impact of abnormal data on the model, and improves the stability and reliability of the model. If the cosine similarity between an actual feature parameter in the database and a data feature label in the regression tree model is greater than the preset similarity threshold, the actual feature parameter is assigned to the branch to which the data feature label belongs. Repeat the above steps until all actual feature parameters in the database are assigned and processed, and the actual feature parameter regression tree model is obtained.

[0098] In summary, by building a regression tree model and using the cosine similarity algorithm, the actual feature parameters can be effectively classified and noise data can be removed, which helps to improve the quality and reliability of the data and make subsequent analysis more accurate. The complex data processing process can be simplified into several clear steps, which improves the efficiency and automation of data processing. Accurate feature parameter classification and denoising can help improve the accuracy of subsequent prediction models, thereby better predicting abnormal events in the smelting process and improving production efficiency and safety.

[0099] The actual characteristic parameters in the database are classified and processed to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period, and the following steps are also included:

[0100] Calculating the Euclidean distance between each actual feature parameter on each branch and the corresponding branch center in the actual feature parameter regression tree model;

[0101] If the Euclidean distance between a certain actual feature parameter on a certain branch and the center of the corresponding branch is greater than a preset distance threshold, the actual feature parameter is screened out on the branch, and the actual feature parameter is simulated and distributed to the remaining branches;

[0102] Calculate the Euclidean distance between the actual characteristic parameters assigned to the remaining branches and the center of the remaining branches;

[0103] If the Euclidean distance between the actual characteristic parameter assigned to the remaining branches by the simulation and the center of the branches of the remaining branches is still greater than the preset distance, the actual characteristic parameter is marked as noise data and the noise data is filtered out;

[0104] If the Euclidean distance between the actual characteristic parameter assigned to the remaining branches by the simulation and the center of a branch of the remaining branches is not greater than the preset distance, it means that the actual characteristic parameter belongs to the remaining branches, and the actual characteristic parameter is reallocated to the remaining branches;

[0105] Repeat the above steps until the actual feature parameters on each branch in the actual feature parameter regression tree model are evaluated, and the final actual feature parameter regression tree model is obtained;

[0106] The final actual characteristic parameter regression tree model is pruned to obtain an actual characteristic parameter set of each smelting parameter adjustment module within a preset time period.

[0107] It should be noted that in the actual feature parameter regression tree model, the Euclidean distance between each actual feature parameter on each branch and the corresponding branch center is calculated. If the Euclidean distance between a certain actual feature parameter on a branch and the corresponding branch center is greater than the preset distance threshold, the actual feature parameter is screened out on the branch. At the same time, the actual feature parameter is simulated and assigned to the remaining branches. The Euclidean distance between the actual feature parameters simulated and assigned to the remaining branches and the branches of the remaining branches is calculated. If the Euclidean distance between the actual feature parameters simulated and assigned to the remaining branches and the branches of the remaining branches is still greater than the preset distance, the actual feature parameter is calibrated as noise data and screened out. If the Euclidean distance between the actual feature parameters simulated and assigned to the remaining branches and the branches of the remaining branches is not greater than the preset distance, it means that the actual feature parameter belongs to the remaining branches and is reallocated to the remaining branches. Repeat the above steps until the actual feature parameters on each branch in the actual feature parameter regression tree model are evaluated, and the final actual feature parameter regression tree model is obtained. It can be seen that by calculating the Euclidean distance and screening out data with too large a distance, removing noise data and inappropriate feature parameters, and further optimizing the classification and categorization of feature parameters, the model can be made more accurate and reliable. The final model can not only more accurately reflect the actual feature parameters of each smelting parameter adjustment module within the preset time period, but also improve the accuracy of subsequent predictions and anomaly detection, ensuring the stability and efficiency of the smelting process.

[0108] Preferably, if the probability value of one or more abnormal smelting events occurring in the aluminum melting furnace equipment is greater than a preset probability value, an aluminum melting parameter adjustment measure scheme is obtained in the knowledge graph constructed in advance, and the smelting parameters of the aluminum melting furnace equipment are adjusted according to the obtained aluminum melting parameter adjustment measure scheme, specifically:

[0109] Pre-prepare the aluminum melting parameter adjustment measures corresponding to various abnormal smelting events in the aluminum melting furnace equipment;

[0110] Construct a knowledge graph, import the extracted prefabricated aluminum melting parameter adjustment measures corresponding to various abnormal smelting events in the aluminum melting furnace equipment into the knowledge graph, and regularly update the knowledge graph;

[0111] If the probability value of one or more abnormal smelting events occurring in the aluminum melting furnace equipment is greater than a preset probability value, a search matching term is constructed according to the abnormal smelting events corresponding to the probability values ​​greater than the preset probability values;

[0112] Importing the search matching words into the knowledge graph for search matching, and obtaining relevant aluminum melting parameter adjustment measures;

[0113] The obtained aluminum melting parameter adjustment measure scheme is fed back to the control terminal of the aluminum melting furnace equipment, so as to adjust the smelting parameters of the aluminum melting furnace equipment based on the obtained aluminum melting parameter adjustment measure scheme.

[0114] It should be noted that the relevant technical personnel pre-prepare the aluminum melting parameter adjustment measures corresponding to various abnormal smelting events in the aluminum melting furnace equipment, and these measures include specific adjustment measures for different abnormal events. For example, when it is predicted that the aluminum melting furnace will have an over-high temperature event that leads to aluminum liquid oxidation, the aluminum melting parameter adjustment measures such as reducing the heating power, increasing the cooling intensity of the cooling system, and adjusting the smelting cycle can be adopted to suppress the problem of excessive temperature of the aluminum melting furnace equipment, thereby avoiding aluminum liquid oxidation. When it is predicted that the aluminum melting furnace will have an over-high pressure event that leads to equipment overload, the aluminum melting parameter adjustment measures such as reducing the feed rate, increasing the exhaust valve opening to release excess gas pressure, and reducing the smelting load can be adopted to suppress the problem of excessive pressure of the aluminum melting furnace equipment, thereby avoiding equipment overload and potential safety risks.

[0115] Build a knowledge graph and import the extracted prefabricated aluminum melting parameter adjustment measures into it. The knowledge graph contains different abnormal events and their corresponding adjustment measures. The knowledge graph is updated regularly to ensure that it contains the latest adjustment measures and best practices. When the probability value of one or more abnormal smelting events in the aluminum melting furnace equipment is predicted to be greater than the preset probability value, a search matching term is constructed according to the abnormal smelting events corresponding to these probability values. The search matching term is imported into the knowledge graph for search and matching to obtain the relevant aluminum melting parameter adjustment measures. The obtained aluminum melting parameter adjustment measures are fed back to the control terminal of the aluminum melting furnace equipment, and the smelting parameters of the aluminum melting furnace equipment are adjusted based on these plans. By building a knowledge graph and using the prediction results to quickly obtain and apply the adjustment measures, not only the speed and accuracy of abnormal processing are improved, but also the robustness and safety of the system are enhanced, and the quality of aluminum liquid production is effectively improved.

[0116] Among them, the specific method steps of "constructing search matching words according to the abnormal melting events corresponding to these probability values" are as follows: first, a detailed analysis is performed on the abnormal melting events whose predicted probability values ​​are greater than the preset probability values. For each abnormal event, its key feature information is extracted. These key features can be parameters directly related to the operating status of the aluminum melting furnace equipment (such as abnormal performance in temperature, pressure, flow, etc.), the type name of the abnormal event (such as slagging in the furnace, aluminum liquid composition deviation, etc.), and the possible location or link of the abnormal event. Then, these key feature information are combined according to certain logic and rules. For example, if the two abnormal events are the temperature in the furnace is too high and the aluminum liquid composition deviation, the constructed search matching words are "furnace temperature is too high / aluminum liquid composition deviation / aluminum melting furnace adjustment measures". If more abnormal events are involved, their key features are also integrated in this way to ensure that the search matching words can accurately reflect the combination of these abnormal events, so as to perform accurate search matching in the knowledge graph, thereby obtaining the corresponding aluminum melting parameter adjustment measures.

[0117] In addition, during the actual operation of the aluminum melting furnace equipment, the smelting parameter adjustment method may further include the following steps:

[0118] Obtain characteristic working condition image information corresponding to various abnormal working conditions of aluminum melting furnace equipment during the smelting process through the big data network, and obtain the associated smelting parameter adjustment module corresponding to various abnormal working conditions;

[0119] Constructing an information library, compressing and bundling characteristic working condition image information corresponding to various abnormal working conditions and associated smelting parameter adjustment modules respectively to obtain a plurality of data packets; importing each data packet into the information library;

[0120] During the operation of the aluminum melting furnace equipment, real-time working condition image information of the aluminum liquid in the aluminum melting furnace equipment is obtained during a preset working time period, and the real-time working condition image information is imported into the information database;

[0121] Calculating a mean square error value between the real-time working condition image information and the characteristic working condition image information in each data packet; comparing the mean square error value between the real-time working condition image information and the characteristic working condition image information in each data packet with a preset error value threshold;

[0122] If there is a case where at least one mean square error value is greater than a preset error value threshold, extract the data packet corresponding to the mean square error value greater than the preset error value threshold, obtain the associated smelting parameter adjustment module in the extracted data packet, and define the associated smelting parameter adjustment module in the extracted data packet as a suspicious smelting parameter adjustment module;

[0123] Acquire the real-time smelting parameter of the suspicious smelting parameter adjustment module, determine whether the real-time smelting parameter is within a preset smelting parameter range, and if not, adjust the corresponding real-time smelting parameter to within the preset smelting parameter range;

[0124] Among them, the smelting parameter adjustment module includes a stirring module, a furnace temperature adjustment module and a furnace pressure adjustment module.

[0125] Among them, "correlated smelting parameter adjustment module" refers to the smelting parameter adjustment module directly related to the abnormal working condition when the aluminum melting furnace equipment has a certain abnormal working condition, such as the stirring module, furnace temperature adjustment module and furnace pressure adjustment module. The parameter changes of these modules may cause or aggravate the abnormal working condition. Therefore, in the process of abnormal detection and parameter adjustment, it is necessary to focus on the real-time parameters of these modules so as to take corresponding adjustment measures in time.

[0126] It should be noted that the characteristic working condition image information corresponding to various abnormal working conditions of the aluminum melting furnace equipment during the smelting process is obtained through the big data network. At the same time, the associated smelting parameter adjustment modules (such as stirring module, furnace temperature adjustment module, furnace pressure adjustment module) corresponding to these abnormal working conditions are obtained. The above-mentioned characteristic working condition image information and the associated smelting parameter adjustment module are compressed and bundled respectively to form several data packets. These data packets are imported into the information library.

[0127] During the operation of the aluminum melting furnace equipment, the real-time working condition image information of the aluminum liquid in the aluminum melting furnace equipment is obtained and imported into the information library. The mean square error value between the real-time working condition image information and the characteristic working condition image information in each data packet in the information library is calculated. These mean square error values ​​are compared with the preset error value threshold. If there is at least one mean square error value greater than the preset error value threshold, the data packet with the mean square error value greater than the preset error value threshold is extracted. The associated smelting parameter adjustment module in these data packets is obtained and defined as a suspicious smelting parameter adjustment module. The real-time smelting parameters of the suspicious smelting parameter adjustment module are obtained to determine whether these real-time smelting parameters are within the preset smelting parameter range. If not within the preset range, the corresponding real-time smelting parameters are adjusted to within the preset smelting parameter range. By comparing the real-time working condition image information with the historical characteristic working condition image information, the working state of the aluminum melting furnace equipment can be monitored in real time, abnormal conditions can be discovered in time, and the early warning capability can be improved. Once the suspicious smelting parameter adjustment module is identified, the relevant parameters can be adjusted in a targeted manner to ensure the stability and safety of the smelting process. Through timely parameter adjustment and abnormal handling, the downtime and resource waste in the smelting process can be reduced, the production efficiency and product quality can be improved, and the abnormal operating conditions of the aluminum melting furnace equipment can be effectively responded to and handled, thereby improving production efficiency and safety.

[0128] In addition, during the actual operation of the aluminum melting furnace equipment, the smelting parameter adjustment method may further include the following steps:

[0129] After the aluminum melting parameter adjustment measure is adjusted, the parameter values ​​of the adjusted smelting parameters are obtained at several time nodes, and an actual dynamic curve diagram of the adjusted smelting parameters is constructed according to the parameter values ​​of the adjusted smelting parameters obtained at several time nodes;

[0130] Obtaining a preset dynamic curve graph of the adjusted smelting parameters, and calculating the degree of overlap between the actual dynamic curve graph and the preset dynamic curve graph based on a Euclidean distance algorithm;

[0131] If the overlap is not greater than the preset overlap, the adjusted smelting parameter corresponding to the overlap being not greater than the preset overlap is calibrated as an adjustment invalid smelting parameter;

[0132] Acquire a signal receiver / transmitter corresponding to the adjustment invalid smelting parameter, and define it as a suspicious signal receiver / transmitter;

[0133] Obtaining, through a signal detector, whether the signal information fed back by the suspicious signal receiver / transmitter can be received within a preset time period;

[0134] If not, the suspected signal receiver / transmitter is defined as a faulty signal receiver / transmitter;

[0135] If yes, then obtaining the number of interruptions of the signal information fed back by the suspicious signal receiver / transmitter received within a preset time period;

[0136] If the number of interruptions is greater than a preset number, the suspicious signal receiver / transmitter is defined as a faulty signal receiver / transmitter.

[0137] It should be noted that, first, by adjusting the smelting parameters and collecting the data of these parameters at multiple time points, an actual dynamic curve graph is constructed. This step helps to monitor the changes in the smelting process in real time. Next, the Euclidean distance algorithm is used to compare the similarity or difference between the actual dynamic curve graph and the preset ideal curve graph. If the overlap between the two is low, it means that the current smelting parameters still deviate from the ideal smelting conditions. If it is found that the overlap between the actual dynamic curve graph and the preset dynamic curve graph is lower than the preset standard, it is considered that these adjusted smelting parameters are invalid. Further, the system will identify the signal receiver / transmitter corresponding to these invalid parameters as a suspicious object. Within a preset time period, a signal detector is used to check whether the suspicious signal receiver / transmitter is working properly. If its feedback signal cannot be received, or even if it is received but the number of signal interruptions exceeds the preset threshold, it can be determined that the signal receiver / transmitter is faulty. Through the above steps, the system can accurately locate those faulty signal receivers / transmitters that affect the stability of the smelting process, so as to take corresponding measures to repair or replace them. By monitoring the smelting parameters in real time and comparing them with the ideal state, the parameters can be adjusted in time to maintain the stability of the smelting process. Once a problem is found, the system can quickly locate the specific signal receiver / transmitter that may cause the problem, improving the efficiency of problem solving. It can not only ensure the efficient and stable operation of the smelting process, but also quickly locate the cause and take effective measures when problems occur, thereby improving the overall performance of the aluminum melting furnace equipment.

[0138] The second aspect of the present invention provides an aluminum melting furnace device, which is applied to any one of the aluminum melting furnace device smelting parameter adjustment methods based on intelligent feedback, such as Figure 3 As shown, including:

[0139] The main body of the aluminum melting furnace 2101 is used for melting aluminum materials;

[0140] The furnace temperature regulating module 2102 is used to control the temperature of the aluminum melting furnace so that the melt temperature is kept within a preset range to meet different smelting process requirements;

[0141] The furnace pressure regulating module 2103 is used to control and adjust the pressure in the aluminum melting furnace so that the pressure in the furnace is kept within a preset range to ensure the safety and stability of the smelting process;

[0142] A stirring module 2104 is used to stir the melt to make the melt temperature more uniform and promote the full dissolution and reaction of alloy elements;

[0143] The data collection and storage module is responsible for collecting data collected by preset sensors installed inside the main body of the aluminum melting furnace, and storing the collected data in a database to provide a data basis for subsequent analysis and prediction; wherein the preset sensors include vibration sensors, temperature sensors and pressure sensors.

[0144] The third aspect of the present invention provides a smelting parameter adjustment system for aluminum melting furnace equipment based on intelligent feedback, the smelting parameter adjustment system for aluminum melting furnace equipment includes a memory and a processor, the memory stores a smelting parameter adjustment method program for aluminum melting furnace equipment, when the smelting parameter adjustment method program for aluminum melting furnace equipment is executed by the processor, any one of the steps of the smelting parameter adjustment method for aluminum melting furnace equipment is implemented.

[0145] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback, characterized in that: The following steps are involved: Obtaining historical characteristic parameters corresponding to each smelting parameter adjustment module of the aluminum melting furnace equipment within a preset time period before various abnormal smelting events occur, and constructing a hidden Markov prediction model according to the historical characteristic parameters; During the operation of the aluminum melting furnace equipment, the actual characteristic parameters of each smelting parameter adjustment module are continuously collected at a number of preset time nodes, a database is constructed, and the collected actual characteristic parameters are imported into the database; After the collection is completed, each actual characteristic parameter in the database is classified and processed to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period; Importing the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period into the hidden Markov prediction model for prediction, and obtaining the probability values ​​of various smelting abnormal events occurring in the aluminum melting furnace equipment; The probability values ​​of various abnormal smelting events occurring in the aluminum melting furnace equipment are compared with the preset probability values. If the probability values ​​of various abnormal smelting events occurring in the aluminum melting furnace equipment are not greater than the preset probability values, no adjustment is performed on the aluminum melting furnace equipment. If the probability value of one or more abnormal smelting events occurring in the aluminum melting furnace equipment is greater than a preset probability value, the aluminum melting parameter adjustment measure plan is obtained from the knowledge graph constructed in advance, and the smelting parameters of the aluminum melting furnace equipment are adjusted according to the obtained aluminum melting parameter adjustment measure plan.

2. The method for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback according to claim 1 is characterized in that: Obtain the historical characteristic parameters corresponding to each smelting parameter adjustment module of the aluminum melting furnace equipment within a preset time period before various abnormal smelting events occur, and construct a hidden Markov prediction model based on the historical characteristic parameters, specifically: Obtain a smelting log of an aluminum smelting furnace device, and obtain various abnormal smelting events that occurred in the aluminum smelting furnace device during historical smelting operations according to the smelting log; Acquire, according to the smelting log, historical characteristic parameters corresponding to each smelting parameter adjustment module within a preset time period before various smelting abnormal events occur; Aggregate and process the historical characteristic parameters corresponding to each smelting parameter adjustment module within a preset time period before various smelting abnormal events to obtain a pre-abnormal characteristic parameter set corresponding to various smelting abnormal events; Construct a hidden Markov model, take various abnormal smelting events as observation nodes, and take the pre-abnormal characteristic parameter sets corresponding to various abnormal smelting events as state nodes; Connect each observation node with the corresponding state node in a directed ring manner to generate a topological structure diagram between each observation node and each state node; Calculate the Euclidean distance between each observation node and each state node according to the topological structure diagram, and determine the adjacency relationship between each observation node and each state node according to the corresponding Euclidean distance value; An adjacency matrix is ​​constructed according to the adjacency relationship between each observation node and each state node, and the adjacency matrix is ​​imported into the hidden Markov model for training and learning until the model parameters meet the preset requirements, thereby obtaining a hidden Markov prediction model.

3. The method for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback according to claim 2 is characterized in that: Based on the directed ring method, each observation node is connected to the corresponding state node in a directed manner to generate a topological structure diagram between each observation node and each state node, specifically: Determine the order of observation nodes and state nodes based on the occurrence order of abnormal smelting events and the time series of parameter changes; Traverse each observation node, establish directed connections between each observation node and state node according to the order of each observation node and state node, and obtain directed connection relationships between each observation node and state node; Connect each observation node to its previous state node, and at the same time, connect each state node to its previous observation node to form a closed loop, and obtain a closed loop relationship between each observation node and state node; According to the directed connection relationship and closed-loop relationship between each observation node and state node, the topological structure diagram between each observation node and each state node is drawn using the Python library; The nodes in the topological structure diagram represent observation nodes and state nodes, and the directed edges represent the connection relationship between observation nodes and state nodes.

4. The method for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback according to claim 1 is characterized in that: The actual characteristic parameters in the database are classified and processed to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period, specifically: Constructing a regression tree model, obtaining the number of each smelting parameter adjustment module, and determining a number of cutting nodes according to the number of each smelting parameter adjustment module; Cutting the regression tree model based on the cutting node to cut out a plurality of branches in the regression tree model; Acquire data characteristic information of characteristic parameters corresponding to each smelting parameter adjustment module, and assign a data characteristic label to each tree branch according to the data characteristic information of characteristic parameters corresponding to each smelting parameter adjustment module; Obtain each actual feature parameter in the database, and calculate the cosine similarity between each actual feature parameter in the database and each data feature label in the regression tree model based on the cosine similarity algorithm; If the cosine similarity between a certain actual feature parameter in the database and each data feature label in the regression tree model is not greater than a preset similarity threshold, the actual feature parameter in the database is marked as noise data and the noise data is filtered out; If the cosine similarity between a certain actual feature parameter in the database and a certain data feature label in the regression tree model is greater than a preset similarity threshold, the actual feature parameter is assigned to the branch to which the data feature label belongs; Repeat the above steps until all actual feature parameters in the database are assigned and processed, and obtain the actual feature parameter regression tree model.

5. The method for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback according to claim 4 is characterized in that: Classifying the actual characteristic parameters in the database to obtain the actual characteristic parameter set of each smelting parameter adjustment module within a preset time period also includes the following steps: Calculating the Euclidean distance between each actual feature parameter on each branch and the corresponding branch center in the actual feature parameter regression tree model; If the Euclidean distance between a certain actual feature parameter on a certain branch and the corresponding branch center is greater than a preset distance threshold, the actual feature parameter is screened out on the branch, and the actual feature parameter is simulated and distributed to the remaining branches; Calculate the Euclidean distance between the actual characteristic parameters assigned to the remaining branches and the center of the remaining branches; If the Euclidean distance between the actual characteristic parameter assigned to the remaining branches by the simulation and the center of the branches of the remaining branches is still greater than the preset distance, the actual characteristic parameter is marked as noise data and the noise data is filtered out; If the Euclidean distance between the actual characteristic parameter assigned to the remaining branches by the simulation and the center of a branch of a certain remaining branch is not greater than the preset distance, it means that the actual characteristic parameter belongs to the remaining branch, and the actual characteristic parameter is reallocated to the remaining branch; Repeat the above steps until the actual feature parameters on each branch in the actual feature parameter regression tree model are evaluated, and the final actual feature parameter regression tree model is obtained; The final actual characteristic parameter regression tree model is pruned to obtain an actual characteristic parameter set of each smelting parameter adjustment module within a preset time period.

6. The method for adjusting smelting parameters of aluminum melting furnace equipment based on intelligent feedback according to claim 1 is characterized in that: If the probability value of one or more abnormal smelting events occurring in the aluminum melting furnace equipment is greater than the preset probability value, the aluminum melting parameter adjustment measure scheme is obtained from the knowledge graph constructed in advance, and the smelting parameters of the aluminum melting furnace equipment are adjusted according to the obtained aluminum melting parameter adjustment measure scheme, specifically: Pre-prepare the aluminum melting parameter adjustment measures corresponding to various abnormal smelting events in the aluminum melting furnace equipment; Construct a knowledge graph, import the extracted prefabricated aluminum melting parameter adjustment measures corresponding to various abnormal smelting events in the aluminum melting furnace equipment into the knowledge graph, and regularly update the knowledge graph; If the probability value of one or more abnormal smelting events occurring in the aluminum melting furnace equipment is greater than a preset probability value, a search matching term is constructed according to the abnormal smelting events corresponding to the probability values ​​greater than the preset probability values; Importing the search matching words into the knowledge graph for search matching, and obtaining relevant aluminum melting parameter adjustment measures; The obtained aluminum melting parameter adjustment measure scheme is fed back to the control terminal of the aluminum melting furnace equipment, so as to adjust the smelting parameters of the aluminum melting furnace equipment based on the obtained aluminum melting parameter adjustment measure scheme.

7. A smelting parameter adjustment system for aluminum melting furnace equipment based on intelligent feedback, characterized in that: The aluminum melting furnace equipment smelting parameter adjustment system includes a memory and a processor, wherein the memory stores a method program for adjusting the smelting parameters of the aluminum melting furnace equipment. When the method program for adjusting the smelting parameters of the aluminum melting furnace equipment is executed by the processor, the steps of the method for adjusting the smelting parameters of the aluminum melting furnace equipment as described in any one of claims 1 to 6 are implemented.

Images