Intelligent smelting parameter control system for raw magnesium

By designing an intelligent smelting parameter control system during the smelting process, and using industrial sensors and deep learning technology for real-time data acquisition and analysis, the problems of measurement errors and response delays during the smelting process are solved, and production efficiency and system compatibility are improved. Achieve more efficient, energy-saving and environmentally friendly production.

CN119045431AInactive Publication Date: 2024-11-29SHENYANG HENGXING WEIYE MACHINERY MANUFACTURING CO LTD

Patent Information

Application Number
CN202411159799.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-22
Publication Date
2024-11-29
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

There are measurement errors, response time delays, unstable performance, real-time data processing bottlenecks in large data environments, and the inability of the model to accurately capture complex process changes and system compatibility problems during the existing raw magnesium smelting process, which affects the degree of intelligence and production efficiency.

Method used

An intelligent smelting parameter control system including data acquisition, preliminary processing, deep analysis, predictive modeling, intelligent control and communication and integration modules was designed, and real-time data acquisition was collected using industrial sensors, edge computing, deep learning and simulation simulation technologies. , analyze and predict to realize adaptive adjustment and closed-loop control.

Benefits of technology

It improves data acquisition accuracy and processing capabilities, enhances model performance and decision-making, solves compatibility problems, and achieves a more efficient, energy-saving and environmentally friendly production process.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119045431A_ABST
    Figure CN119045431A_ABST
Patent Text Reader

Abstract

The invention provides an intelligent smelting parameter control system for raw magnesium. Comprising a data acquisition module used for data acquisition; the data primary processing module is used for carrying out primary processing and analysis on the acquired data in a place close to a data source; the deep analysis module is used for analyzing the overall trend change of the historical data, analyzing the periodic change and trend of the key parameters and detecting abnormal behaviors in the historical data; the prediction modeling module is used for obtaining a parameter prediction model; the intelligent control module is used for designing an intelligent control strategy, deploying a decision support system and a closed-loop control scheme to adjust smelting parameters in real time, and testing and optimizing the intelligent control strategy by using an analog simulation technology; and the communication and integration module is used for realizing communication and providing a human-computer interface. The intelligent degree of the original magnesium smelting process and the smelting production efficiency can be greatly improved, and more efficient, more energy-saving and more environment-friendly production is achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent metal smelting, and in particular to an intelligent smelting parameter control system for raw magnesium. Background Art

[0002] Primary magnesium is one of the metallurgical terms. It has low density, high specific strength, and can form high-strength alloys with metals such as aluminum, copper, and zinc. It is an important alloying element. The intelligent smelting of primary magnesium mainly involves several key aspects, including the establishment of digital workshops, the development of high-purity magnesium preparation demonstration equipment, the development and application of new magnesium reduction equipment, and the innovation of primary magnesium smelting technology.

[0003] In the intelligent smelting of primary magnesium, intelligent smelting parameter control is mainly achieved through real-time monitoring, analysis and optimization of key parameters in the smelting process. Reasonable parameter control is the key link to achieve efficient, safe, energy-saving and high-quality production, and is of great significance to improving the competitiveness of the entire production system. The existing parameter control technology as a whole has the following problems: 1) There are measurement errors, response time delays or unstable performance in extreme environments; 2) In a large data volume environment, real-time data processing and analysis may become a bottleneck; 3) The model may not be able to accurately capture complex process changes, affecting the prediction effect; 4) There may be compatibility issues in the integration between different systems and equipment. Therefore, it is increasingly important to continuously improve the intelligence level of the primary magnesium smelting process. Summary of the invention

[0004] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide an intelligent smelting parameter control system for raw magnesium, which can greatly improve the intelligence level of the raw magnesium smelting process and the smelting production efficiency, and achieve more efficient, energy-saving and environmentally friendly production.

[0005] To achieve the above object, the present invention provides the following solution: an intelligent smelting parameter control system for raw magnesium, comprising:

[0006] Data acquisition module, used to collect data during the smelting process using industrial sensors and data acquisition cards;

[0007] The data preliminary processing module is used to perform preliminary processing and analysis of collected data close to the data source using edge computing technology;

[0008] The deep analysis module is used to analyze the overall trend changes of historical data using data analysis tools, identify the periodic changes and trends of key parameters in historical data using time series analysis, and detect abnormal behaviors in historical data using intelligent algorithms;

[0009] The prediction modeling module is used to select and train models using all collected data to obtain parameter prediction models;

[0010] Intelligent control module, used to design intelligent control strategies for adaptive adjustment of smelting process, deploy decision support system and closed-loop control scheme to adjust predicted smelting parameters in real time, and test and optimize the intelligent control strategies using simulation technology;

[0011] The communication and integration module is used to realize the communication between all modules and between any module and the external system, and provides a human-machine interface to monitor and control the smelting process.

[0012] Optionally, the data acquisition module includes:

[0013] A sensor unit, used to identify the parameters that need to be monitored in the smelting process and to collect parameters of industrial sensors in a redundant sensor layout; the sensor unit includes a temperature sensor, a pressure sensor, a gas sensor, a humidity sensor, and a calibration device for calibrating the sensor;

[0014] The data acquisition and processing unit is used to transmit the collected data in real time to the edge computing device or data center for storage and processing using wireless networks and data transmission protocols, and use data analysis methods to detect sensor failures. If a failure occurs, an alarm and replacement program will be automatically triggered.

[0015] Optionally, the data preliminary processing module includes:

[0016] An edge computing unit is used to deploy edge computing equipment at key locations on the smelting site and collect key parameters as well as clean and filter the key parameters; wherein the edge computing device is provided with a lightweight data analysis algorithm for calculating the mean and variance of key parameters and a temporary storage subunit for temporary data storage;

[0017] The preliminary processing unit is used to compress the processed key parameters. When the key parameters change or an abnormal event occurs, the data is transmitted to the central server using the data transmission protocol and the temporary storage subunit is triggered to perform temporary data backup.

[0018] Optionally, the in-depth analysis module includes:

[0019] An intelligent analysis unit is used to collect and preprocess the collected data to obtain relevant data sets, and based on the relevant data sets, use an advanced data analysis platform to perform correlation analysis and cluster analysis to obtain the relationship between parameters and natural groupings and patterns in the relevant data sets, and then perform visualization based on the analysis results;

[0020] A time series analysis unit, used to collect time series data in the collected data and align the time series data with timestamps, and then use a time series decomposition method to decompose the aligned time series data into trend data, seasonal data and random noise data to perform trend and periodic change analysis of key parameters;

[0021] An anomaly detection unit is used to establish a normal baseline model using normal smelting process data, compare the real-time data of smelting with the normal baseline model and use an anomaly detection algorithm to detect abnormal fluctuations; wherein the statistical indicators of the normal baseline model include mean, variance, maximum and minimum values, and the anomaly detection algorithm includes a machine learning algorithm, a deep learning algorithm, and a rule and logic detection mechanism.

[0022] Optionally, the prediction modeling module includes:

[0023] A feature extraction unit, used for combining the relevant data set with historical process parameters, environmental conditions and equipment status to obtain a comprehensive data set, and performing feature extraction and feature creation in the comprehensive data set to enhance data capture capability;

[0024] A model training unit, used to divide the comprehensive data set into a training set, a validation set and a test set, and use grid search or random search to optimize hyperparameters, then use the training set to train the model, use the validation set to evaluate the model performance, and use the test set to verify the model capability, so as to obtain a parameter prediction model;

[0025] The model tuning unit is used to establish a feedback mechanism to identify defects in the parameter prediction model in a data-driven manner, and to continuously iteratively update the model structure and parameters based on the defects.

[0026] Optionally, the intelligent control module includes:

[0027] A control strategy design unit, used to design a control strategy based on deep learning technology and integrate the control strategy into the control panel through an embedded system or control software;

[0028] A parameter adjustment unit, used for designing a decision support system based on the control strategy, integrating the decision support system into the parameter prediction model to generate an optimal parameter adjustment suggestion, combining the optimal parameter adjustment suggestion with applied control theory to formulate an optimal control strategy, and adjusting the smelting parameters in a closed-loop control manner based on the optimal control strategy;

[0029] The simulation optimization unit is used to build a simulation platform and design a variety of simulation control strategies, apply the simulation control strategies in the simulation platform and evaluate the impact on the smelting process, obtain simulation output, and optimize the optimal control strategy based on the simulation output.

[0030] Optionally, the control strategy design unit includes:

[0031] A deep learning subunit, configured to select a deep learning model to perform complex data pattern learning and strategy generation so that the deep learning model has adaptive capabilities, combine the output of the parameter prediction model in the deep learning model, and predict the impact of future control actions on the current state through iterative calculation and optimization, thereby generating an optimized control strategy;

[0032] The control strategy deployment subunit is used to receive and process parameter adjustment commands issued by the control strategy in real time to achieve adaptive control of parameters.

[0033] Optionally, the parameter adjustment unit includes:

[0034] A decision support subunit is used to integrate the designed decision support system architecture into the parameter prediction model by using a data processing engine to predict future parameters and detect anomalies, and then to clarify the objective function to be optimized and use an optimization algorithm to analyze the output of the integrated parameter prediction model to generate the optimal parameter adjustment suggestion; wherein, the decision support system architecture includes a data layer for collecting and storing real-time smelting data, a processing layer for prediction and decision-making, and a decision layer for executing parameter adjustment commands;

[0035] The closed-loop control subunit is used to integrate IoT devices, data transmission networks, control algorithms and human-machine interfaces to obtain an integrated system. The integrated system forms a closed-loop control by predicting future parameter changes, automatically executing and optimizing the optimal control strategy, and triggering an automatic adjustment mechanism.

[0036] Optionally, the simulation optimization unit includes:

[0037] The platform building subunit is used to build a mathematical model according to the smelting process and physical characteristics, combine the mathematical model with the simulation software and the platform to obtain a simulation environment, and design a variety of simulation control strategies according to different production goals and constraints. In the simulation environment, various simulation control strategies are applied to evaluate the impact on the smelting process and complete the simulation test;

[0038] The optimization evaluation subunit is used to analyze the output sensitivity and risk based on simulation tests, and to make real-time adjustments to strategies based on the analysis results.

[0039] Optionally, the communication and integration module includes:

[0040] A communication unit, used to select a communication protocol and use a gateway or a protocol converter to implement protocol conversion, and select middleware to perform data buffering and asynchronous communication;

[0041] The interface unit is used to develop interfaces and interface documents for each module and to perform compatibility testing and compatibility adjustments on the interfaces;

[0042] Security unit, which is used to set up authentication mechanisms and use encryption technology on communication protocols to ensure communication security, and set up logging and monitoring mechanisms to detect system behavior in real time;

[0043] The human-machine interface unit includes the functions of role definition, user interface and interactive element design, permission management, user training, technical support, interface testing and system optimization.

[0044] The present invention discloses the following technical effects by providing an intelligent smelting parameter control system for raw magnesium:

[0045] 1. High data acquisition accuracy: Data is collected through various types of industrial sensors, and sensor calibration is performed using calibration devices to ensure the accuracy of the measured data. Regular recalibration is performed to maintain accuracy. By connecting multiple sensors using data acquisition cards, the collected data can be converted into a computer-readable data format. The use of redundant sensor layout can prevent data loss caused by single point failures.

[0046] 2. Strong data processing and analysis capabilities: The data processing process is divided into two modules: preliminary analysis and deep analysis; 1) Preliminary analysis: using edge computing technology to perform preliminary data processing and analysis close to the data source can significantly reduce latency and bandwidth usage, while quickly responding to on-site anomalies, improving the real-time and reliability of the system, not only improving data processing efficiency, but also enhancing the system's expansion capabilities and anti-interference performance. 2) Deep analysis module: using data analysis tools to comprehensively analyze and visualize historical data, help reveal hidden patterns and trends in the raw magnesium smelting process, and provide data support for optimizing production processes, improving efficiency and reducing costs. Time series analysis can effectively identify and predict the periodic changes and trends of key parameters, providing a solid foundation for predictive modeling and intelligent decision-making. Intelligent abnormal behavior detection can detect and handle abnormal events in real time during the raw magnesium smelting process, prevent potential problems, and ensure the continuity and stability of production.

[0047] 3. High model performance: Through continuous training and tuning of the model, the prediction results are made more accurate, which can effectively support the generation of control strategy decisions, improve production efficiency and prevent potential problems.

[0048] 4. High decision-making: 1) Use deep learning algorithms (such as deep reinforcement learning) and control theory to design intelligent control strategies to achieve adaptive adjustment of the process. 2) Deploy a decision support system to adjust smelting parameters (such as temperature and pressure) in real time based on the output of the prediction model, and continuously optimize the control strategy to achieve the optimal process effect. 3) Use simulation technology (such as numerical simulation and simulation software) to test and optimize different operating strategies to ensure that the implementation of the control strategy will not cause risks to actual production. 4) Through the organic combination of the above-mentioned deep data analysis, predictive modeling and intelligent control decision-making, the intelligent smelting parameter control system can make the best decision in real time in a dynamically changing production environment, thereby improving production efficiency, stability and safety.

[0049] 5. Strong compatibility: Standardized interfaces and protocols are adopted to enhance system interoperability, achieve efficient system integration, improve the overall automation level, and solve possible compatibility problems, thereby ensuring the system's fast, efficient and safe operation.

[0050] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] 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 will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0052] Figure 1 A system architecture diagram provided for an embodiment of the present invention;

[0053] Figure 2 A schematic diagram of the process architecture of data processing and analysis provided by an embodiment of the present invention;

[0054] Figure 3 A schematic diagram of a process architecture for generating an intelligent control strategy provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0057] like Figure 1-3 As shown, the present invention provides an intelligent smelting parameter control system for raw magnesium, comprising:

[0058] 1. Data acquisition module, used to collect data in the smelting process using industrial sensors and data acquisition cards; the data acquisition module includes:

[0059] 1.1 Sensor Unit

[0060] It is used to identify the parameters that need to be monitored in the smelting process, such as temperature, pressure, gas composition (such as oxygen, carbon monoxide, etc.), humidity, etc. It also collects parameters of industrial sensors in a redundant sensor layout.

[0061] The sensor unit comprises:

[0062] Temperature sensor: choose high temperature resistant thermocouple or infrared temperature sensor to adapt to the high temperature conditions in the smelting environment;

[0063] Pressure sensors, using piezoresistive, strain gauge or capacitive sensors, ensuring high accuracy and reliability;

[0064] Gas sensor: choose the appropriate gas composition sensor, such as zirconium oxide sensor to measure oxygen, electrochemical or infrared sensor to measure carbon monoxide;

[0065] Humidity sensor, choose capacitive humidity sensor when necessary

[0066] Calibrate the sensor's calibration device to ensure the accuracy of the measurement data, and recalibrate it regularly to maintain accuracy.

[0067] Install sensors at key locations on smelting equipment to ensure measurement accuracy and data representativeness. For example, temperature sensors should be installed at locations where temperature changes are greatest or most critical.

[0068] The redundant sensor layout approach, that is, by using multiple sensors of the same type, ensures that even if one of the sensors fails, the other sensors can still provide sufficient information to ensure the safe operation of the system, thereby improving the reliability of the system.

[0069] 1.2 Data acquisition and processing unit

[0070] It is used to transmit the collected data in real time to the edge computing device or data center for storage and processing using wireless networks and data transmission protocols, and use data analysis methods to detect sensor failures. If a failure occurs, it automatically triggers an alarm and replacement program.

[0071] Connect the data acquisition card to multiple sensors, convert them into computer-readable data format, and configure a suitable data sampling rate for each sensor to ensure the real-time and signal integrity of the data.

[0072] 2. Data preliminary processing module, such as Figure 2 As shown, it is used to use edge computing technology to perform preliminary processing and analysis of collected data close to the data source; the data preliminary processing module includes:

[0073] 2.1 Edge Computing Unit

[0074] It is used to deploy edge computing devices at key locations of the smelting site and collect key parameters as well as clean and filter them. Simple algorithms (such as mean filtering and median filtering) can be applied to smooth the data and reduce the impact of short-term fluctuations on the analysis.

[0075] The edge computing device is provided with a lightweight data analysis algorithm for calculating the mean and variance of key parameters and a temporary storage subunit for temporary data storage. Among them, lightweight data analysis algorithms, such as statistical calculations, real-time anomaly detection (using simple thresholds, rules or condition monitoring), etc. The data analysis algorithm is used to perform real-time statistical analysis and calculate indicators such as the mean and variance of key parameters to simplify subsequent data processing.

[0076] Edge computing devices usually select industrial computing devices (such as edge gateways, industrial PCs, and embedded computers) with sufficient computing power and environmental tolerance to meet on-site processing needs. Edge computing devices are deployed at key locations on the smelting site, close to sensor data collection points, to reduce data transmission distances and delays, and network topology and physical environment are considered to ensure the connection stability and physical protection of the equipment.

[0077] 2.2 Preliminary processing unit

[0078] It is used to compress the processed key parameters to reduce the data volume and bandwidth occupancy. When the key parameters change or an abnormal event occurs, the data transmission protocol is used to transmit the data to the central server and trigger the temporary storage subunit to perform temporary data backup to ensure that key information is not lost when the network is interrupted.

[0079] 3. In-depth analysis modules, such as Figure 2 As shown, it is used to use data analysis tools to analyze the overall trend changes of historical data, use time series analysis to identify the periodic changes and trends of key parameters in historical data, and use intelligent algorithms to detect abnormal behaviors in historical data; the deep analysis module includes:

[0080] 3.1 Intelligent Analysis Unit

[0081] It is used to aggregate and preprocess the collected data to obtain relevant data sets, and based on the relevant data sets, use an advanced data analysis platform to perform correlation analysis and cluster analysis to obtain the relationship between parameters and the natural groupings and patterns in the relevant data sets (that is, it helps to reveal hidden patterns and trends in the raw magnesium smelting process), and then perform visualization based on the analysis results.

[0082] Different types of graphs can be drawn to present data, including but not limited to:

[0083] Line chart: shows trends and changes in time series data;

[0084] Scatter plot: shows the relationship between two variables;

[0085] Heatmaps: Identify dense areas in a dataset;

[0086] Bar chart: Compare changes in different categories.

[0087] 3.2 Time Series Analysis Unit

[0088] It is used to collect time series data in the acquired data and align the time series data with timestamps, and then use the time series decomposition method to decompose the aligned time series data into trend data, seasonal data and random noise data to perform trend and periodic change analysis of key parameters.

[0089] By observing the seasonal component, the frequency and magnitude of cyclical changes can be identified. By analyzing the trend component, long-term upward or downward trends can be identified, and moving averages or exponential smoothing methods can be used to further smooth the data and highlight the trend. The autocorrelation function (ACF) and partial autocorrelation function (PACF) can be used to identify cyclical patterns and determine the length and strength of the cycle.

[0090] 3.3 Anomaly Detection Unit

[0091] It is used to establish a normal baseline model using normal smelting process data, compare the real-time data of smelting with the normal baseline model and perform abnormal fluctuation detection using anomaly detection algorithm.

[0092] The statistical indicators of the normal baseline model include mean, variance, maximum and minimum values. Machine learning algorithms (such as principal component analysis PCA) can be used to find normal patterns in the data. Monitor the data flow entering the system in real time, record the changes and trends of key parameters, compare real-time data with the baseline model, detect abnormal fluctuations, and use statistical process control methods such as control charts (such as Shewhart control charts), CUSUM or EWMA to monitor parameter stability.

[0093] Anomaly detection algorithms include:

[0094] Machine learning algorithms: such as Isolation Forest, One-Class SVM, and DBSCAN clustering, which automatically identify abnormal patterns;

[0095] Deep learning algorithms: such as autoencoders or variational autoencoders (VAEs) for feature learning and anomaly detection;

[0096] Rules and logic detection mechanisms: Use business rules and expert experience to define abnormal conditions.

[0097] For example, machine learning algorithms are used for anomaly detection. Unsupervised learning, such as K-means and DBSCAN, is used to find abnormal patterns in unlabeled data. Supervised learning, such as support vector machines (SVM) and random forests, can be trained using labeled data. The alarm threshold is then set reasonably based on the model output to distinguish between normal and abnormal states, and new data is analyzed in real time to quickly identify and alert possible anomalies.

[0098] 4. Prediction modeling module, used to select and train models using all smelting data to obtain parameter prediction models; the prediction modeling module includes:

[0099] 4.1 Feature Extraction Unit

[0100] for combining the relevant data set with historical process parameters, environmental conditions and equipment status to obtain a comprehensive data set, and performing feature extraction and feature creation in the comprehensive data set to enhance data capture capabilities;

[0101] 4.2 Model Training Unit

[0102] The comprehensive dataset is divided into a training set, a validation set, and a test set, usually in a ratio of 70:15:15, to ensure the generalization ability of the model, and a grid search or random search is used to optimize the hyperparameters to improve the model performance, and then the training set is used for model training, and the validation set is used to evaluate the model performance (using indicators such as mean square error, mean absolute error, R 2 etc., to measure the accuracy of the model), use the test set to verify the model capability and obtain the parameter prediction model;

[0103] 4.3 Model Tuning Unit

[0104] It is used to establish a feedback mechanism to identify defects in the parameter prediction model in a data-driven manner, and to iteratively update the model structure and parameters based on the defects to ensure that the model always provides high-precision predictions.

[0105] 5. Intelligent control module, such as Figure 3As shown, it is used to design intelligent control strategies to perform adaptive adjustments to the smelting process, deploy decision support systems and closed-loop control solutions to adjust smelting parameters in real time, and use simulation technology to test and optimize the intelligent control strategies. The intelligent control module includes:

[0106] 5.1 Control Strategy Design Unit

[0107] Used to design control strategies based on deep learning technology and integrate the control strategies into the control panel through embedded systems or control software; the control strategy design unit includes:

[0108] 5.11 Deep Learning Subunit

[0109] Used to select deep learning models (such as deep Q network DQN, convolutional neural network CNN, recurrent neural network RNN, etc.) for complex data pattern learning and strategy generation so that the deep learning model has adaptive capabilities and can dynamically adjust parameters and strategies based on historical performance. Train deep learning models to maximize long-term benefits, such as allowing the system to automatically adjust working parameters to optimize output through a reward mechanism. Here, policy gradient methods, Q learning, or deep reinforcement learning methods can be applied to control strategy learning.

[0110] The output of the parameter prediction model is combined in the deep learning model, and the impact of future control actions on the current state is predicted through iterative calculation and optimization to generate the optimized control strategy.

[0111] 5.12 Control Strategy Deployment Subunit

[0112] It is used to receive and process parameter adjustment commands issued by the control strategy in real time to achieve adaptive control of prediction parameters.

[0113] 5.2 Parameter Adjustment Unit

[0114] It is used to design a decision support system based on the control strategy, and integrate the decision support system into the parameter prediction model to generate optimal parameter adjustment suggestions, and then combine the optimal parameter adjustment suggestions with applied control theory to formulate an optimal control strategy, and adjust the smelting parameters in a closed-loop control manner based on the optimal control strategy. The parameter adjustment unit includes:

[0115] 5.21 Decision Support Subunit

[0116] It is used to use the data processing engine to integrate the designed decision support system architecture into the parameter prediction model to predict future parameters and detect anomalies, then clarify the objective function that needs to be optimized (including production efficiency, cost function and resource utilization, etc.) and use the optimization algorithm (such as linear programming, nonlinear optimization, heuristic algorithm) to analyze the output of the integrated parameter prediction model to generate the optimal parameter adjustment suggestion.

[0117] Among them, the objective function formula is, for example but not limited to the following

[0118] Production efficiency: Production efficiency = total output / total input;

[0119] Cost function: Total Cost = FC + VC (Q), fixed cost (FC) refers to the cost that does not change with the output, such as equipment purchase and rental costs. Variable cost (VC) refers to the cost directly related to the production quantity (Q), such as raw material cost and labor cost.

[0120] Resource utilization: Resource Utilization = T consumed resources / available resources.

[0121] The comprehensive objective function is obtained by multiplying each objective function needed by their respective weights and then adding them together.

[0122] Before designing the decision support system architecture, we must first clarify the problems and goals that the system needs to solve, including which parameters need to be adjusted, and the specific optimization goals such as increasing production, reducing energy consumption, and stabilizing quality. The decision support system architecture includes a data layer for collecting and storing real-time smelting data, a processing layer for prediction and decision-making, and a decision layer for executing parameter adjustment commands.

[0123] 5.22 Closed-loop control subunit

[0124] An integrated system is obtained by integrating Internet of Things devices, data transmission networks, control algorithms and human-machine interfaces. The integrated system forms a closed-loop control by predicting future parameter changes, automatically executing and optimizing the optimal control strategy, and triggering an automatic adjustment mechanism.

[0125] 5.3 Simulation Optimization Unit

[0126] It is used to build a simulation platform and design a variety of simulation control strategies, apply the simulation control strategies in the simulation platform and evaluate the impact on the smelting process, obtain simulation output, and optimize the optimal control strategy based on the simulation output. The simulation optimization unit includes:

[0127] 5.31 Platform Construction Subunit

[0128] It is used to construct a mathematical model (including heat transfer, mass transfer, chemical reaction kinetics, etc.) according to the smelting process and physical characteristics, combine the mathematical model with simulation software and platform to obtain a simulation environment, and design a variety of simulation control strategies according to different production goals and constraints. These strategies may cover aspects such as parameter adjustment and process path selection. In the simulation environment, various simulation control strategies are applied to evaluate the impact on the smelting process (including the comprehensive impact on production efficiency, energy consumption, quality stability and safety) to complete simulation testing;

[0129] 5.32 Optimizing the Evaluation Subunit

[0130] Used for output sensitivity and risk analysis based on simulation tests, and real-time adjustment of strategies based on the analysis results.

[0131] 6. Communication and integration module, used to realize the communication between modules and between modules and external systems, and provide human-machine interface to monitor and control the smelting process. The communication and integration module includes:

[0132] 6.1 Communication Unit

[0133] Used to select communication protocols (such as OPC UA, Modbus, PROFINET, etc., to achieve interconnection between the device layer and the control layer) and use gateways or protocol converters to achieve protocol conversion (that is, to achieve interoperability between different protocols), select middleware for data buffering and asynchronous communication, and improve reliability and throughput; middleware can also act as a central hub for data exchange, managing data production and consumption of different modules.

[0134] 6.2 Interface Unit

[0135] Used to develop clear and standardized interfaces (APIs) for each module to support service requests and data exchange between modules. The interfaces contain detailed API documentation that describes interface functions, parameter requirements, return values, etc. Perform compatibility testing of interfaces to verify compatibility between different modules and between modules and external systems (such as ERP and MES systems). Use automated testing tools to improve test efficiency and coverage. When compatibility issues are found during testing, adjust module design and interfaces to support other standards and systems and complete compatibility adjustments.

[0136] 6.3 Security Unit

[0137] Used to set up authentication mechanisms and use encryption technology (such as TLS / SSL) on communication protocols to ensure communication security, and set up logging and monitoring mechanisms to detect system behavior in real time;

[0138] 6.4 Human-machine interface unit

[0139] Role definition function: clarify the different needs of operators and managers. Operators need real-time monitoring and simple control functions, while managers need comprehensive analysis and decision support. You can also collect specific functional requirements such as real-time data monitoring, alarm prompts, historical data query, operation records, etc. through interviews and questionnaires.

[0140] Design user interface and interactive elements: 1) Design the information architecture of the interface so that the hierarchy is clear and important information is easy to obtain. Different functional modules such as real-time data monitoring, alarm management, historical data, etc. should have clear navigation paths. 2) Use simple and intuitive UI design, adopt modern design principles such as responsive design, and ensure that the interface can be displayed well on different devices. 3) Use charts, dashboards and color coding to display equipment status, production parameters and abnormal alarms to make the information more readable. 4) Use modern web front-end frameworks (such as React, Angular, Vue.js) to build dynamic and responsive user interfaces. 5) Implement front-end subscription to real-time data through WebSocket or Server-Sent Events to ensure instant data updates. 6) Provide easy-to-use interface controls such as buttons, sliders, drop-down menus, etc., so that operators and managers can quickly get started. 7) Ensure that control functions have confirmation and cancellation mechanisms to reduce the risk of misoperation.

[0141] Permission management function: Set permissions according to different user roles (such as operators, engineers, managers) to limit the access of different levels of personnel to data and control functions.

[0142] User training function: Provide detailed user manuals and training materials, including function descriptions, operation instructions, etc. Organize regular training courses to improve operators' proficiency in the system.

[0143] Technical support function: Establish technical support channels, such as hotlines and online customer service, to promptly resolve problems encountered by users during use.

[0144] Interface testing function: organize operators and managers to conduct interface testing, collect feedback and improve the system.

[0145] System optimization function: Iterate development and optimization based on user feedback and usage data to improve system performance and user experience.

[0146] Therefore, the present invention can greatly improve the intelligence level of the raw magnesium smelting process and the smelting production efficiency by providing an intelligent smelting parameter control system for raw magnesium, thereby achieving more efficient, energy-saving and environmentally friendly production.

[0147] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0148] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. An intelligent smelting parameter control system for raw magnesium, characterized in that: include: Data acquisition module, used to collect data during the smelting process using industrial sensors and data acquisition cards; The data preliminary processing module is used to perform preliminary processing and analysis of collected data close to the data source using edge computing technology; The deep analysis module is used to analyze the overall trend changes of historical data using data analysis tools, identify the periodic changes and trends of key parameters in historical data using time series analysis, and detect abnormal behaviors in historical data using intelligent algorithms; The prediction modeling module is used to select and train models using all collected data to obtain parameter prediction models; Intelligent control module, used to design intelligent control strategies for adaptive adjustment of smelting process, deploy decision support system and closed-loop control scheme to adjust predicted smelting parameters in real time, and test and optimize the intelligent control strategies using simulation technology; The communication and integration module is used to realize the communication between all modules and between any module and the external system, and provides a human-machine interface to monitor and control the smelting process.

2. The intelligent smelting parameter control system of raw magnesium according to claim 1 is characterized in that: The data acquisition module comprises: A sensor unit, used to identify the parameters that need to be monitored in the smelting process and to collect parameters of industrial sensors in a redundant sensor layout; the sensor unit includes a temperature sensor, a pressure sensor, a gas sensor, a humidity sensor, and a calibration device for calibrating the sensor; The data acquisition and processing unit is used to transmit the collected data in real time to the edge computing device or data center for storage and processing using wireless networks and data transmission protocols, and use data analysis methods to detect sensor failures. If a failure occurs, an alarm and replacement program will be automatically triggered.

3. The intelligent smelting parameter control system of raw magnesium according to claim 2 is characterized in that: The data preliminary processing module comprises: An edge computing unit is used to deploy edge computing equipment at key locations on the smelting site and collect key parameters as well as clean and filter the key parameters; wherein the edge computing device is provided with a lightweight data analysis algorithm for calculating the mean and variance of key parameters and a temporary storage subunit for temporary data storage; The preliminary processing unit is used to compress the processed key parameters. When the key parameters change or an abnormal event occurs, the data is transmitted to the central server using the data transmission protocol and the temporary storage subunit is triggered to perform temporary data backup.

4. The intelligent smelting parameter control system of raw magnesium according to claim 3 is characterized in that: The in-depth analysis module includes: An intelligent analysis unit is used to collect and preprocess the collected data to obtain relevant data sets, and based on the relevant data sets, use an advanced data analysis platform to perform correlation analysis and cluster analysis to obtain the relationship between parameters and natural groupings and patterns in the relevant data sets, and then perform visualization based on the analysis results; A time series analysis unit, used to collect time series data in the collected data and align the time series data with timestamps, and then use a time series decomposition method to decompose the aligned time series data into trend data, seasonal data and random noise data to perform trend and periodic change analysis of key parameters; An anomaly detection unit is used to establish a normal baseline model using normal smelting process data, compare the real-time data of smelting with the normal baseline model and use an anomaly detection algorithm to detect abnormal fluctuations; wherein the statistical indicators of the normal baseline model include mean, variance, maximum and minimum values, and the anomaly detection algorithm includes a machine learning algorithm, a deep learning algorithm, and a rule and logic detection mechanism.

5. The intelligent smelting parameter control system of raw magnesium according to claim 4 is characterized in that: The prediction modeling module includes: A feature extraction unit, used for combining the relevant data set with historical process parameters, environmental conditions and equipment status to obtain a comprehensive data set, and performing feature extraction and feature creation in the comprehensive data set to enhance data capture capability; A model training unit, used to divide the comprehensive data set into a training set, a validation set and a test set, and use grid search or random search to optimize hyperparameters, then use the training set to train the model, use the validation set to evaluate the model performance, and use the test set to verify the model capability, so as to obtain a parameter prediction model; The model tuning unit is used to establish a feedback mechanism to identify defects in the parameter prediction model in a data-driven manner, and to continuously iteratively update the model structure and parameters based on the defects.

6. The intelligent smelting parameter control system of raw magnesium according to claim 5, characterized in that: The intelligent control module comprises: A control strategy design unit, used to design a control strategy based on deep learning technology and integrate the control strategy into the control panel through an embedded system or control software; A parameter adjustment unit, used for designing a decision support system based on the control strategy, integrating the decision support system into the parameter prediction model to generate an optimal parameter adjustment suggestion, combining the optimal parameter adjustment suggestion with applied control theory to formulate an optimal control strategy, and adjusting the smelting parameters in a closed-loop control manner based on the optimal control strategy; The simulation optimization unit is used to build a simulation platform and design a variety of simulation control strategies, apply the simulation control strategies in the simulation platform and evaluate the impact on the smelting process, obtain simulation output, and optimize the optimal control strategy based on the simulation output.

7. The intelligent smelting parameter control system of raw magnesium according to claim 6, characterized in that: The control strategy design unit includes: A deep learning subunit, configured to select a deep learning model to perform complex data pattern learning and strategy generation so that the deep learning model has adaptive capabilities, combine the output of the parameter prediction model in the deep learning model, and predict the impact of future control actions on the current state through iterative calculation and optimization, thereby generating an optimized control strategy; The control strategy deployment subunit is used to receive and process parameter adjustment commands issued by the control strategy in real time to achieve adaptive control of parameters.

8. The intelligent smelting parameter control system of raw magnesium according to claim 7, characterized in that: The parameter adjustment unit comprises: A decision support subunit is used to integrate the designed decision support system architecture into the parameter prediction model by using a data processing engine to predict future parameters and detect anomalies, and then to clarify the objective function to be optimized and use an optimization algorithm to analyze the output of the integrated parameter prediction model to generate the optimal parameter adjustment suggestion; wherein, the decision support system architecture includes a data layer for collecting and storing real-time smelting data, a processing layer for prediction and decision-making, and a decision layer for executing parameter adjustment commands; The closed-loop control subunit is used to integrate IoT devices, data transmission networks, control algorithms and human-machine interfaces to obtain an integrated system. The integrated system forms a closed-loop control by predicting future parameter changes, automatically executing and optimizing the optimal control strategy, and triggering an automatic adjustment mechanism.

9. The intelligent smelting parameter control system of raw magnesium according to claim 8, characterized in that: The simulation optimization unit comprises: The platform building subunit is used to build a mathematical model according to the smelting process and physical characteristics, combine the mathematical model with the simulation software and the platform to obtain a simulation environment, and design a variety of simulation control strategies according to different production goals and constraints. In the simulation environment, various simulation control strategies are applied to evaluate the impact on the smelting process and complete the simulation test; The optimization evaluation subunit is used to analyze the output sensitivity and risk based on simulation tests, and to make real-time adjustments to strategies based on the analysis results.

10. The intelligent smelting parameter control system of raw magnesium according to claim 9, characterized in that: The communication and integration module includes: A communication unit, used to select a communication protocol and use a gateway or a protocol converter to implement protocol conversion, and select middleware to perform data buffering and asynchronous communication; The interface unit is used to develop interfaces and interface documents for each module and to perform compatibility testing and compatibility adjustments on the interfaces; Security unit, which is used to set up authentication mechanisms and use encryption technology on communication protocols to ensure communication security, and set up logging and monitoring mechanisms to detect system behavior in real time; The human-machine interface unit includes the functions of role definition, user interface and interactive element design, permission management, user training, technical support, interface testing and system optimization.

Citation Information

Patent Citations

  • Non-ferrous metal smelting process control man-machine-object co-fusion manufacturing platform architecture system

    CN113379372A

  • Intelligent accurate control method for silicon content of molten iron

    CN117631623A

  • Intelligent digital twin simulation system for steel production process

    CN117930786A

  • Modeling method for temperature soft measurement of blast-furnace tuyere convolute area

    CN117995289A

Cited By

  • Intelligent smelting factory informatization management method and system

    CN119358845A

  • Quality rating and recommending method for magnesium-based alloy

    CN119479952A

  • Pulse power supply system parameter selection method and system

    CN119675488A

  • Data-driven multi-furnace continuous smelting intelligent regulation and control system and method

    CN119689859A

  • Acid mist absorption tower intelligent control system based on big data

    CN119847027A