Intelligent sintering method, system and equipment based on multi-source data analysis and medium

Abstract

The invention belongs to the technical field of steel smelting sintering data processing, and particularly relates to an intelligent sintering method, system and equipment based on multi-source data analysis and a medium. According to the method, the multi-source data and the quality inspection data are collected, preprocessing of noise removal, abnormal value removal, missing value supplementation and feature extraction is carried out on the multi-source data, and the preprocessed data are temporarily protected; performing batching precision analysis, sintering process analysis, energy consumption control analysis, negative pressure ignition control analysis, air pressure and air quantity control analysis and machine speed four-machine linkage analysis on the preprocessed data; model dynamic parameters are adjusted through quality inspection data, an intelligent decision model is constructed through a multi-objective optimization algorithm, and an optimal production control strategy is generated; generating an optimal decision scheme by adopting rule reasoning, case reasoning and fuzzy reasoning; and evaluating and feeding back the decision according to the difference between the actual production effect and the expected target.

Description

Technical Field

[0001] This invention belongs to the field of sintering data processing technology, and particularly relates to an intelligent sintering method, system, equipment and medium based on multi-source data analysis. Background Technology

[0002] The sintering production process is lengthy and involves many stages. From a control perspective, the sintering process is complex, nonlinear, time-varying, and uncertain, making it a typical complex controlled object. How steel companies can ensure product quality at each stage while reducing steel production costs, thereby enhancing their competitiveness, has become a pressing issue. For a long time, sintering batching has been largely controlled manually by operators based on experience. However, with a deeper understanding of the properties of sintering raw materials, improvements in sintering equipment, the adoption of relevant technologies, and the introduction and implementation of automatic control concepts, the application of intelligent optimized ore blending systems has become an important means to pursue "high quality, high output, and low consumption" in sintering production, and a key focus for sintering plants both domestically and internationally to improve their technological levels.

[0003] The prior art disclosed in CN115952636A is an intelligent ore blending method for sintering processes based on digital twins, including the following steps: 1) Analyzing the requirements for digital twins of the sintering process and equipment, studying the digital replication of the geometric attributes of the sintering process and equipment and multi-temporal-scale modeling of the operating mechanism, establishing a digital twin model of the entire sintering production process using digital twin technology, and inputting the operating data of the real physical model, including process data and product data, into the digital twin model to realize the virtual-real interaction between the physical model and the twin model; on this basis, constructing a material formulation optimization module based on digital twins; 2) Using advanced sensing, data mining and processing, data modeling, and control optimization, establishing a sintering material formulation optimization module based on the digital twin model of the entire sintering production process, forming a visualized virtual sintering production system. The system achieves parallel operation with the physical production line on site, providing a twin system platform for simulation, monitoring and diagnosis, prediction and optimization, and testing and verification for safe and efficient operation of sintering production; 3) virtual and real data are generated through digital twin technology to expand the database; at the same time, data cleaning, data dimensionality reduction and data association are achieved by using data mining and database management technologies; a sintering process knowledge base is constructed; 4) a sintering process batching optimization module is constructed using the XGBoost model and an improved gray wolf optimization algorithm, and an improved gray wolf optimization algorithm is proposed for solving the model. The improvement includes two aspects: convergence factor adjustment based on the sigmoid function and individual update based on the differential mutation strategy; 5) the constructed batching optimization module is uploaded to the digital twin system to optimize the raw material ratio of the sintering process in real time and display it through the front end.

[0004] Therefore, the aforementioned inventions only focus on the ore blending stage, lacking effective control and optimization methods for other important stages of sintering production. This may lead to insufficient coordination in the entire production process, affecting the quality of the final product and production efficiency. Although digital twin models can perform simulation and prediction, deviations may exist between the model and the actual situation during actual production. Furthermore, updating and adjusting the model requires a certain amount of time, making it difficult to achieve real-time dynamic control of the production process. Optimization using virtual and real data generated by digital twin models does not fully utilize the large amount of high-frequency data and low-frequency quality inspection information generated in real time at the production site, potentially missing some key changes in production status and optimization opportunities. Constructing digital twin models requires high technical barriers and substantial computing resources, resulting in high system development and maintenance costs. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides an intelligent sintering method, system, equipment, and medium based on multi-source data analysis. This invention collects multi-source data and quality inspection data, performs preprocessing on the multi-source data (noise removal, outlier removal, missing value supplementation, and feature extraction), and temporarily protects the preprocessed data. The preprocessed data is then analyzed for batching accuracy, sintering process, energy consumption control, negative pressure ignition control, air pressure and air volume control, and four-machine speed linkage. The quality inspection data is used to adjust the dynamic parameters of the model. A multi-objective optimization algorithm is used to construct an intelligent decision-making model to generate the optimal production control strategy. Rule-based reasoning, case-based reasoning, and fuzzy reasoning are employed to generate the optimal decision scheme. The decision is evaluated and feedback is provided based on the difference between the actual production effect and the expected target.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides an intelligent sintering method based on multi-source data analysis, comprising the following steps: Step S1, multi-source data acquisition step, to acquire production data of each stage in sintering production, to periodically collect quality inspection data of sintered minerals, and to upload the collected data to the edge preprocessing terminal and multi-source data analysis server. Step S2, edge preprocessing step: receive the uploaded collected data, filter, clean and aggregate the uploaded data, and upload the preprocessed data to the multi-source data analysis server; Step S3, multi-source data analysis step, accepts pre-processed data and quality inspection data, stores the pre-processed data and quality inspection data, and analyzes and controls the batching accuracy, sintering process, energy consumption control, negative pressure ignition control, air pressure and air volume control, and four-machine linkage of machine speed based on the pre-processed data; Step S4: Intelligent decision-making and control steps. Based on the process mechanism and historical data, a multi-objective optimization algorithm is used to construct an intelligent decision-making model to generate the optimal control strategy; multiple inference algorithms are used to perform decision reasoning to generate the optimal decision scheme, and the control strategy is distributed to each actuator device. Step S5, Decision Evaluation and Feedback Step: The system compares the uploaded actual production data with the expected decision data, provides feedback to the decision model and reasoning mechanism based on the comparison results, and optimizes them through a reinforcement learning agent.

[0007] In a second aspect, the present invention provides an intelligent sintering system based on multi-source data analysis, comprising a data acquisition module, an edge preprocessing module, an actuator module, a data analysis module, an intelligent decision-making module, a decision evaluation and feedback module, and a control distribution module; The data acquisition module includes sensor sub-modules and quality inspection equipment sub-modules distributed throughout the various stages of sintering production; by collecting production data and product quality inspection data during sintering production, it provides multi-source data for sintering production analysis. The sensor submodule collects production data in real time during the sintering process by installing various sensors. The quality inspection equipment submodule obtains product quality inspection data by periodically conducting quality inspections on sintered products; The edge preprocessing module includes an edge computing submodule, a local storage submodule, and an edge controller submodule. It removes noise and outliers through preprocessing to improve data quality, shortens the analysis time of multi-source data analysis servers, and provides temporary storage to ensure that data is not lost in the event of network or server failure, and that data is retransmitted after the network or server is restored. The edge computing submodule uses filtering algorithms to remove noise from the uploaded multi-source data, and interpolation algorithms and statistical analysis methods to remove outliers and fill in missing values. After processing, it uses principal component analysis, correlation analysis, and wavelet transform to extract key feature information related to sintering quality, efficiency, and cost. The local storage submodule temporarily saves preprocessed data to the local storage device when the network is interrupted or the multi-source data analysis server fails. After the network interruption or multi-source data analysis server failure is restored, the data is retransmitted. It supports fast query and analysis of local data. The edge controller submodule receives control commands from the control module and then sends them to the actuator module. The actuator module sets the operating parameters of the sintering process and controls the smooth execution of the sintering production process through preset parameters; it also receives control commands from the edge controller submodule to precisely control the production equipment in real time. The data storage and analysis module includes a data storage submodule, a batching accuracy analysis submodule, a sintering process analysis submodule, an energy consumption control analysis submodule, a negative pressure ignition control analysis submodule, an air pressure and air volume control analysis submodule, and a machine speed and four-machine linkage analysis submodule. By analyzing the above six aspects of multi-source data, the production equipment is adjusted in real time based on the analysis results, and the model parameters in each analysis submodule are dynamically adjusted in real time based on quality inspection data. The data storage submodule is used to store data uploaded by edge preprocessing and quality inspection equipment. It adopts distributed file system and database technology for data storage, and performs long-term storage and management of the collected multi-source data; including relational database for storing structured data and non-relational database for storing unstructured data; The ingredient accuracy analysis submodule uses regression analysis and PID control algorithms to establish a correlation model between ingredient ratios and finished product chemical composition. It performs ingredient accuracy analysis on the actual ingredient quantity data from the uploaded multi-source data and the finished product chemical composition data from the quality inspection data. By comparing the deviation between the ingredient set value and the actual ingredient quantity in real time, it generates control commands and dynamically adjusts the model parameters based on the quality inspection data. The sintering process analysis submodule uses fuzzy logic control and adaptive control algorithms to construct a multi-factor coupled sintering process control model. It analyzes the sintering process of uploaded multi-source data and quality inspection data, generates control commands, and dynamically adjusts the model parameters based on the quality inspection data. The energy consumption control and analysis submodule uses a genetic algorithm and an energy consumption prediction model to perform energy consumption analysis on the uploaded multi-data, generate control commands, and dynamically adjust the parameters of the energy consumption prediction model based on the quality inspection data. The negative pressure ignition control and analysis submodule uses a PID control algorithm to establish a correlation model between negative pressure, ignition effect, and sintering quality. It performs negative pressure ignition analysis on real-time monitoring data from high-frequency negative pressure sensors, temperature and gas flow data from the igniter, and low-frequency quality inspection information of the sintered ore surface layer, generates control commands, and dynamically adjusts the parameters of the correlation model based on the uploaded quality inspection data. The wind pressure and air volume control analysis submodule uses a fuzzy adaptive control algorithm and mathematical model to analyze the uploaded high-frequency wind pressure and air volume sensor real-time data, sintering material layer permeability detection data and low-frequency sintering internal quality inspection information, generate control commands, and dynamically adjust the algorithm rules and parameters according to the quality inspection data. The four-machine linkage analysis submodule uses a neural network control algorithm to establish a collaborative control model between the sintering machine, the feeding machine, the crusher, and the cooler. It performs four-machine linkage analysis on the uploaded multi-source data and quality inspection data, generates control commands, and dynamically adjusts the weights and thresholds of the multi-objective optimization algorithm based on the quality inspection data. The intelligent decision-making module includes a production control decision-making submodule and a decision-making scheme submodule; The production control decision-making submodule adopts a multi-objective optimization algorithm, constructs an intelligent decision-making model based on process mechanism and historical data, and comprehensively considers the control instructions and production objectives of each algorithm to generate the optimal production control strategy. The decision-making scheme submodule uses rule-based reasoning, case-based reasoning, and fuzzy reasoning to analyze and judge the output of multi-source data and analysis models. Based on different production conditions and constraints, it flexibly adjusts decision-making strategies to generate the optimal decision-making scheme. The decision evaluation and feedback module provides feedback and optimization to the decision-making model and reasoning mechanism by comparing the differences between actual production results and expected goals. The control distribution module distributes control commands generated by the data analysis module and optimal production control commands and optimal decision-making schemes generated by the intelligent decision-making module to the edge controller submodule via the industrial optical network.

[0008] A third aspect of the present invention provides an electronic device including a memory 102, a processor 101, a display module 103, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the preceding intelligent sintering methods based on multi-source data analysis.

[0009] A fourth aspect of the present invention provides a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the preceding intelligent sintering methods based on multi-source data analysis.

[0010] The beneficial effects of this invention are as follows: By comprehensively collecting multi-source information from the entire sintering production process, including high-frequency production information and low-frequency quality inspection information, and employing advanced data processing and analysis technologies to extract key features, this invention analyzes preprocessed data and constructs multiple targeted closed-loop algorithms to achieve precise control and collaborative optimization of each stage of sintering production. Furthermore, by adopting a cloud-based data analysis structure, the invention fully leverages the powerful computing capabilities of cloud computing, combined with the real-time processing capabilities of edge computing and the data acquisition and execution capabilities of terminal devices, enabling efficient data flow and processing. This, in turn, improves the quality and efficiency of sintering production and reduces production costs. Attached Figure Description

[0011] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1A flowchart illustrating the method of this invention; Figure 2 This is a schematic diagram of the system structure of the present invention; Figure 3 This is a schematic diagram of the device structure of the present invention.

[0013] Among them, 101 is the processor, 102 is the memory, and 103 is the display module. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] Unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this application and in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0016] Example 1, such as Figure 1 The intelligent sintering method based on multi-source data analysis shown includes the following steps: Step S1, multi-source data acquisition step, involves acquiring production data from each stage of the sintering process, periodically collecting quality inspection data of the sintered ore, and uploading the collected data to the edge preprocessing terminal and the multi-source data analysis server; the specific steps are as follows: Step S11: Deploy high-precision, high-reliability, and long-life sensors at various stages of sintering production, including raw material warehouses, batching equipment, mixers, sintering machines, and coolers. For example, use electronic scales to measure weight at the batching point and install temperature and pressure sensors on the sintering machine; collect various production data in real time, such as temperature, pressure, flow rate, rotation speed, and composition. Step S12 involves periodically conducting quality inspections on the sintered ore products using quality inspection equipment, including chemical composition analysis, particle size distribution detection, and strength testing. The quality inspection equipment interacts with a multi-source data analysis server, feeding back the test results in real time to provide a basis for optimizing the closed-loop algorithm. Step S13: Industrial Ethernet or wireless communication is used to ensure stable data transmission. A multi-threaded concurrent acquisition mechanism is adopted to feed production data back to the edge preprocessing end in real time and quality inspection data back to the multi-source data analysis server in real time.

[0017] Step S2, edge preprocessing step, involves receiving the uploaded collected data, filtering, cleaning, and aggregating the uploaded data, and then uploading the preprocessed data to the multi-source data analysis server; the specific steps are as follows: Step S21: For the uploaded multi-source data, use a filtering algorithm to remove noise, and use an interpolation algorithm and statistical analysis methods to remove outliers and fill in missing values. Step S22: The data processed in step S21 is subjected to principal component analysis, correlation analysis, and wavelet transform to extract key feature information related to sintering quality, efficiency, and cost, thereby reducing the amount of data transmission. Step S23: The preprocessed data from steps S21 and S23 is uploaded to the multi-source data analysis server via a standardized protocol. In the event of a network interruption or a failure of the multi-source data analysis server, the preprocessed data is temporarily saved to a local storage device. Once the network interruption or the multi-source data analysis server failure is resolved, the data is retransmitted. The temporary storage device supports rapid querying and analysis of local data, providing real-time production information to on-site operators.

[0018] Step S3, multi-source data analysis step, involves receiving pre-processed data and quality inspection data, storing the pre-processed data and quality inspection data, and analyzing and controlling the batching accuracy, sintering process, energy consumption control, negative pressure ignition control, air pressure and air volume control, and four-machine linkage based on the pre-processed data; the specific steps are as follows: Step S31: A correlation model between the proportion of ingredients and the chemical composition of the finished product is established using regression analysis and PID control algorithm. The accuracy of the ingredient proportioning is analyzed by comparing the actual ingredient proportioning data in the multi-source data and the chemical composition data of the finished product in the quality inspection data. By comparing the deviation between the setpoint and the actual amount of ingredients in real time, control commands are generated and sent to the edge preprocessing end. The edge preprocessing end sends the control commands and dynamically adjusts the speed of the feeding equipment. The model parameters are dynamically adjusted according to the quality inspection data. Step S32: Using fuzzy logic control and adaptive control algorithms, a multi-factor coupled sintering process control model is constructed. The sintering process is analyzed based on the uploaded multi-source data and quality inspection data, and control commands are generated. The control commands are then sent to the edge preprocessing end. The edge preprocessing end sends out the control commands and dynamically adjusts the fan speed, sintering machine speed and fuel addition amount. The model parameters are dynamically adjusted based on the quality inspection data. Step S33: Use a genetic algorithm and energy consumption prediction model to perform energy consumption analysis on the uploaded multi-data, generate control commands, and send the control commands to the edge preprocessing end. The edge preprocessing end will then send the control commands to dynamically adjust the operating parameters of each device. The parameters of the energy consumption prediction model will be dynamically adjusted based on the uploaded quality inspection data. Step S34: Using a PID control algorithm, a correlation model is established between negative pressure, ignition effect, and sintering quality. Real-time monitoring data from high-frequency negative pressure sensors, igniter temperature and gas flow data, and low-frequency sintering quality inspection information from the sintered ore surface are analyzed for negative pressure ignition to generate control commands. These commands are then sent to the edge preprocessing end, which dynamically adjusts the flue gate opening, gas supply, and air ratio. The correlation model parameters are dynamically adjusted based on the uploaded quality inspection data. Step S35: Using a fuzzy adaptive control algorithm and mathematical model, analyze the uploaded high-frequency wind pressure and air volume sensor real-time data, sintering material layer permeability detection data, and low-frequency sinter internal quality inspection information to generate control commands. Send the control commands to the edge preprocessing end. The edge preprocessing end sends the control commands to dynamically adjust the fan speed and damper opening. The algorithm rules and parameters are dynamically adjusted according to the uploaded quality inspection data. Step S36: Using a neural network control algorithm, a collaborative control model is established among the sintering machine, the feeding machine, the crusher, and the cooler. The uploaded multi-source data and quality inspection data are analyzed to generate control commands, which are then sent to the edge preprocessing end. The edge preprocessing end then sends the control commands to dynamically adjust the operating speeds of the sintering machine, the feeding machine, the crusher, and the cooler. The weights and thresholds of the multi-objective optimization algorithm are dynamically adjusted based on the uploaded quality inspection data.

[0019] Step S4, Intelligent Decision-Making and Control Steps: Based on the process mechanism and historical data, a multi-objective optimization algorithm is used to construct an intelligent decision-making model to generate the optimal control strategy; multiple inference algorithms are used to perform decision reasoning to generate the optimal decision scheme, and the control strategy is then distributed to each actuator device; the specific steps are as follows: Step S41: Using a multi-objective optimization algorithm, the relationship between multiple factors such as quality, efficiency and cost is balanced. Based on the process mechanism and historical data, an intelligent decision-making model is constructed. The control instructions and production objectives of each algorithm are comprehensively considered to generate the optimal production control strategy. Step S42: Use rule-based reasoning, case-based reasoning, and fuzzy reasoning to analyze and judge the output of multi-source data and analysis models. Based on different production conditions and constraints, such as sintering temperature range constraints, energy consumption constraints, heating rate constraints, and quality constraints, flexibly adjust decision-making strategies to generate the optimal decision-making scheme and improve the adaptability and flexibility of the system. Step S43: The generated control strategy and decision scheme are sent to the edge preprocessing end, and then the edge preprocessing end sends the control strategy and decision scheme to the actuator. Step S5, the decision evaluation and feedback step, involves comparing the uploaded actual production data with the expected decision data, and providing feedback to the decision model and inference mechanism based on the comparison results, which are then optimized through a reinforcement learning agent; the specific steps are as follows: Step S51: Compare the uploaded actual production data with the data expected for decision-making, and provide feedback to the decision-making model and reasoning mechanism based on the comparison results; Step S52: To guide the system towards optimal production goals, a reasonable reward mechanism is designed. The reward function comprehensively considers multiple factors such as production quality, efficiency, and cost. For example, positive rewards are given when the quality indicators of sintered ore meet or exceed expected targets, production efficiency improves, and energy consumption decreases; conversely, negative rewards are given when quality problems occur, production is delayed, or energy consumption is too high. Reward signals are obtained, and the reinforcement learning agent continuously optimizes the control strategy based on these signals, constantly improving the accuracy and reliability of decision-making to ensure the system is always in an optimal operating state. Through interaction with the production environment, it learns which actions to take in different states to obtain the maximum cumulative reward. For example, in the four-machine linkage algorithm, the speed of the four machines is dynamically adjusted based on the processing progress and quality feedback of the sintered ore in each process to achieve the best production effect. As learning progresses, the reinforcement learning agent's strategy gradually converges to the optimal strategy, thereby achieving continuous optimization of the production process.

[0020] Example 2, as Figure 2 As shown, an intelligent sintering system based on multi-source data analysis includes a data acquisition module, an edge preprocessing module, an actuator module, a data storage and analysis module, an intelligent decision-making module, a decision evaluation and feedback module, and a control distribution module. The data acquisition module includes sensor sub-modules and quality inspection equipment sub-modules distributed throughout the various stages of sintering production; by collecting production data and product quality inspection data during sintering production, it provides multi-source data for sintering production analysis. The sensor submodule collects production data in real time during the sintering process by installing various sensors. The quality inspection equipment submodule obtains product quality inspection data by periodically conducting quality inspections on sintered products; The edge preprocessing module includes an edge computing submodule, a local storage submodule, and an edge controller submodule. It removes noise and outliers through preprocessing to improve data quality, shortens the analysis time of multi-source data analysis servers, and provides temporary storage to ensure that data is not lost in the event of network or server failure, and that data is retransmitted after the network or server is restored. The edge computing submodule uses filtering algorithms to remove noise from the uploaded multi-source data, and interpolation algorithms and statistical analysis methods to remove outliers and fill in missing values. After processing, it uses principal component analysis, correlation analysis, and wavelet transform to extract key feature information related to sintering quality, efficiency, and cost. The local storage submodule temporarily saves preprocessed data to the local storage device when the network is interrupted or the multi-source data analysis server fails. After the network interruption or multi-source data analysis server failure is restored, the data is retransmitted. It supports fast query and analysis of local data. The edge controller submodule receives control commands from the control module and sends them to the actuator module. It has a certain degree of autonomous decision-making ability and can perform emergency handling according to preset rules in the event of network latency or failure, so as to ensure the continuity and stability of the production process. The actuator module includes motors, valves, frequency converters, etc., and sets the operating parameters of the sintering process. It controls the smooth execution of the sintering production process through preset parameters. It also receives control commands from the edge controller submodule to precisely control the production equipment in real time. The data storage and analysis module includes sub-modules for data storage, batching accuracy analysis, sintering process analysis, energy consumption control analysis, negative pressure ignition control analysis, air pressure and air volume control analysis, and four-machine linkage analysis. By analyzing multi-source data in these six aspects, the module adjusts production equipment in real time based on the analysis results and dynamically adjusts model parameters in each analysis sub-module using quality inspection data. As the core computing and storage center of the system, it undertakes tasks such as centralized data processing and model training. It possesses powerful computing capabilities and storage capacity, enabling it to handle large-scale production data and complex algorithm calculations. The data storage submodule stores data uploaded from edge preprocessing and quality inspection equipment. It employs a distributed file system and database technology for long-term storage and management of the collected multi-source data. This includes a relational database for storing structured data, such as raw material composition and process parameters, and a non-relational database for storing unstructured data, such as equipment operation logs and image data. The ingredient accuracy analysis submodule uses regression analysis and PID control algorithms to establish a correlation model between ingredient ratios and finished product chemical composition. It performs ingredient accuracy analysis on the actual ingredient quantity data from the uploaded multi-source data and the finished product chemical composition data from the quality inspection data. By comparing the deviation between the ingredient set value and the actual ingredient quantity in real time, it generates control commands and dynamically adjusts the model parameters based on the quality inspection data. The sintering process analysis submodule uses fuzzy logic control and adaptive control algorithms to construct a multi-factor coupled sintering process control model. It analyzes the sintering process of uploaded multi-source data and quality inspection data, generates control commands, and dynamically adjusts the model parameters based on the quality inspection data. The energy consumption control and analysis submodule uses a genetic algorithm and an energy consumption prediction model to perform energy consumption analysis on the uploaded multi-data, generate control commands, and dynamically adjust the parameters of the energy consumption prediction model based on the quality inspection data. The negative pressure ignition control and analysis submodule uses a PID control algorithm to establish a correlation model between negative pressure, ignition effect, and sintering quality. It performs negative pressure ignition analysis on real-time monitoring data from high-frequency negative pressure sensors, temperature and gas flow data from the igniter, and low-frequency quality inspection information of the sintered ore surface layer, generates control commands, and dynamically adjusts the parameters of the correlation model based on the uploaded quality inspection data. The wind pressure and air volume control analysis submodule uses a fuzzy adaptive control algorithm and mathematical model to analyze the uploaded high-frequency wind pressure and air volume sensor real-time data, sintering material layer permeability detection data and low-frequency sintering internal quality inspection information, generate control commands, and dynamically adjust the algorithm rules and parameters according to the quality inspection data. The four-machine linkage analysis submodule uses a neural network control algorithm to establish a collaborative control model between the sintering machine, the feeding machine, the crusher, and the cooler. It performs four-machine linkage analysis on the uploaded multi-source data and quality inspection data, generates control commands, and dynamically adjusts the weights and thresholds of the multi-objective optimization algorithm based on the quality inspection data. The intelligent decision-making module includes a production control decision-making submodule and a decision scheme submodule. It integrates various machine learning and deep learning algorithm libraries for training and optimizing process mechanism models and closed-loop algorithms. By continuously learning from historical data and real-time feedback information, it improves the accuracy and adaptability of the model. The production control decision-making submodule employs a multi-objective optimization algorithm, constructs an intelligent decision-making model based on process mechanisms and historical data, comprehensively considers the control instructions and production objectives of each algorithm, and generates the optimal production control strategy. Based on the sintering process principle, it establishes mathematical models for each link, such as the relationship model between raw materials and chemical composition, and the relationship model between sintering process parameters and quality. It uses historical data to train and optimize the model to improve its accuracy and adaptability. The decision-making scheme submodule uses rule-based reasoning, case-based reasoning, and fuzzy reasoning to analyze and judge the output of multi-source data and analysis models. Based on different production conditions and constraints, it flexibly adjusts decision-making strategies to generate the optimal decision-making scheme. The decision evaluation and feedback module compares the actual production results with the expected goals, providing feedback and optimization to the decision-making model and reasoning mechanism. Based on the trained model and algorithm, it performs real-time monitoring and analysis of the production process, providing a scientific basis for production decisions. It can generate optimal production control strategies according to different production goals and constraints. The control distribution module distributes control commands generated by the data analysis module and optimal production control commands and optimal decision-making schemes generated by the intelligent decision-making module to the edge controller submodule via the industrial optical network.

[0021] To ensure the normal operation and stability of the system, regular unit debugging, integration testing, and optimization adjustments are performed. Unit debugging involves testing the functions of the data acquisition module, edge preprocessing module, actuator module, data analysis module, intelligent decision-making module, decision evaluation and feedback module, and control issuance module to ensure normal operation; checking the accuracy of sensor data acquisition and the correctness of actuator actions; integration testing involves performing full system integration testing to verify the smoothness of data transmission, processing, and control command issuance; simulating different production conditions to check the system response and control effect; and optimization adjustments involve optimizing system parameters and algorithms based on the debugging results to improve system performance and stability.

[0022] Furthermore, the system is monitored in real time to track the operating status of each device, data collection and transmission, observe changes in production indicators, and promptly identify anomalies; emergency response plans are developed to quickly respond to and resolve equipment failures, communication interruptions, and other issues; data is backed up regularly to prevent data loss; production data and feedback information are collected to continuously improve models and algorithms, and system configurations and strategies are adjusted in a timely manner as production processes change.

[0023] The intelligent sintering system constructed in this invention comprehensively covers the entire sintering production process from raw material procurement to finished product output, overcoming the limitations of traditional methods that focus only on a single stage. By mining the correlation features between data through information preprocessing and feature extraction modules, a solid data foundation is provided for the closed-loop algorithm. Multiple closed-loop algorithms are designed for data analysis and dynamically adjusted based on real-time data and feedback information, forming a complete data closed loop. This enables real-time monitoring, precise control, and continuous optimization of the production process. Through precise batching control, energy consumption optimization, and production process coordination, raw material waste and energy consumption are effectively reduced, significantly lowering production costs. Simultaneously, based on the continuous optimization of the multi-source information closed-loop algorithm, the quality stability of the sintered ore is improved, making key indicators such as the chemical composition, particle size distribution, and strength of the product more consistent with the requirements of blast furnace smelting.

[0024] Example 3, as Figure 3As shown, a computer device includes a processor 101, a memory 102, a display module 103, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the intelligent sintering method based on multi-source data analysis described in Embodiment 1.

[0025] Example 4: A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the intelligent sintering method based on multi-source data analysis described in Example 1.

[0026] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0027] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0028] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0029] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1The steps of the function specified in one or more boxes.

[0030] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0031] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A smart sintering method based on multi-source data analysis, characterized in that: Includes the following steps: Step S1, multi-source data acquisition step, to acquire production data of each stage in sintering production, to periodically collect quality inspection data of sintered minerals, and to upload the collected data to the edge preprocessing terminal and multi-source data analysis server. Step S2, edge preprocessing step: receive the uploaded collected data, filter, clean and aggregate the uploaded data, and upload the preprocessed data to the multi-source data analysis server; Step S3, multi-source data analysis step, accepts pre-processed data and quality inspection data, stores the pre-processed data and quality inspection data, and analyzes and controls the batching accuracy, sintering process, energy consumption control, negative pressure ignition control, air pressure and air volume control, and four-machine linkage of machine speed based on the pre-processed data; Step S4: Intelligent decision-making and control steps are issued. Based on the process mechanism and historical data, a multi-objective optimization algorithm is used to construct an intelligent decision-making model to generate the optimal control strategy. Multiple inference algorithms are used to generate the optimal decision scheme, and the control strategy is then distributed to each actuator device. Step S5, Decision Evaluation and Feedback Step: The system compares the uploaded actual production data with the expected decision data, provides feedback to the decision model and reasoning mechanism based on the comparison results, and optimizes them through a reinforcement learning agent.

2. The intelligent sintering method based on multi-source data analysis according to claim 1, characterized in that: The specific steps for S1 are as follows: Step S11: By deploying a sensor network, production data from each stage of sintering is collected to form multi-source data. Step S12: Conduct regular quality inspections on sintered ore products using quality inspection equipment; Step S13: A multi-threaded concurrent acquisition mechanism is adopted to feed production data back to the edge preprocessing end in real time and to feed quality inspection data back to the multi-source data analysis server in real time.

3. The intelligent sintering method based on multi-source data analysis according to claim 2, characterized in that: The specific steps for S2 are as follows: Step S21: For the uploaded multi-source data, use a filtering algorithm to remove noise, and use an interpolation algorithm and statistical analysis methods to remove outliers and fill in missing values. Step S22: The data processed in step S21 is subjected to principal component analysis, correlation analysis, and wavelet transform to extract key feature information related to sintering quality, efficiency, and cost. Step S23: The preprocessed data from steps S21 and S23 is uploaded to the multi-source data analysis server using a standardized protocol. In the event of a network interruption or a failure of the multi-source data analysis server, the preprocessed data is temporarily saved to a local storage device. Once the network interruption or the multi-source data analysis server failure is resolved, the data is re-transmitted. The temporary storage device supports fast querying and analysis of local data.

4. The intelligent sintering method based on multi-source data analysis according to claim 3, characterized in that: The specific steps for S3 are as follows: Step S31: A correlation model between the proportion of ingredients and the chemical composition of the finished product is established using regression analysis and PID control algorithm. The accuracy of the ingredient proportioning is analyzed by comparing the actual ingredient proportioning data in the multi-source data and the chemical composition data of the finished product in the quality inspection data. By comparing the deviation between the setpoint and the actual amount of ingredients in real time, control commands are generated and sent to the edge preprocessing end. The edge preprocessing end sends the control commands and dynamically adjusts the speed of the feeding equipment. The model parameters are dynamically adjusted according to the quality inspection data. Step S32: Using fuzzy logic control and adaptive control algorithms, a multi-factor coupled sintering process control model is constructed. The sintering process is analyzed based on the uploaded multi-source data and quality inspection data, and control commands are generated. The control commands are then sent to the edge preprocessing end. The edge preprocessing end sends out the control commands and dynamically adjusts the fan speed, sintering machine speed and fuel addition amount. The model parameters are dynamically adjusted based on the quality inspection data. Step S33: Use a genetic algorithm and energy consumption prediction model to perform energy consumption analysis on the uploaded multi-data, generate control commands, and send the control commands to the edge preprocessing end. The edge preprocessing end will then send the control commands to dynamically adjust the operating parameters of each device. The parameters of the energy consumption prediction model will be dynamically adjusted based on the uploaded quality inspection data. Step S34: Using a PID control algorithm, a correlation model is established between negative pressure, ignition effect, and sintering quality. Real-time monitoring data from high-frequency negative pressure sensors, igniter temperature and gas flow data, and low-frequency sintering quality inspection information from the sintered ore surface are analyzed for negative pressure ignition to generate control commands. These commands are then sent to the edge preprocessing end, which dynamically adjusts the flue gate opening, gas supply, and air ratio. The correlation model parameters are dynamically adjusted based on the uploaded quality inspection data. Step S35: Using a fuzzy adaptive control algorithm and mathematical model, analyze the uploaded high-frequency wind pressure and air volume sensor real-time data, sintering material layer permeability detection data, and low-frequency sinter internal quality inspection information to generate control commands. Send the control commands to the edge preprocessing end. The edge preprocessing end sends the control commands to dynamically adjust the fan speed and damper opening. The algorithm rules and parameters are dynamically adjusted according to the uploaded quality inspection data. Step S36: Using a neural network control algorithm, a collaborative control model is established among the sintering machine, the feeding machine, the crusher, and the cooler. The uploaded multi-source data and quality inspection data are analyzed to generate control commands, which are then sent to the edge preprocessing end. The edge preprocessing end then sends the control commands to dynamically adjust the operating speeds of the sintering machine, the feeding machine, the crusher, and the cooler. The weights and thresholds of the multi-objective optimization algorithm are dynamically adjusted based on the uploaded quality inspection data.

5. The intelligent sintering method based on multi-source data analysis according to claim 4, characterized in that: The specific steps for S4 are as follows: Step S41: Using a multi-objective optimization algorithm, based on the process mechanism and historical data, an intelligent decision-making model is constructed, and the control instructions and production objectives of each algorithm are comprehensively considered to generate the optimal production control strategy. Step S42: Use rule-based reasoning, case-based reasoning, and fuzzy reasoning to analyze and judge the output of multi-source data and analysis models. Adjust decision-making strategies flexibly according to different production conditions and constraints to generate the optimal decision-making scheme. Step S43: The generated control strategy and decision scheme are sent to the edge preprocessing end, and then the edge preprocessing end sends the control strategy and decision scheme to the actuator.

6. The intelligent sintering method based on multi-source data analysis according to claim 5, characterized in that: The specific steps for S5 are as follows: Step S51: Compare the uploaded actual production data with the data expected for decision-making, and provide feedback to the decision-making model and reasoning mechanism based on the comparison results; Step S52: Set up a reward mechanism, establish a reward function by comprehensively considering multiple factors of production, obtain a reward signal, and continuously optimize the control strategy based on the reward signal.

7. A smart sintering system based on multi-source data analysis, characterized in that: It includes a data acquisition module, an edge preprocessing module, an actuator module, a data analysis module, an intelligent decision-making module, a decision evaluation and feedback module, and a control distribution module; The data acquisition module includes sensor sub-modules and quality inspection equipment sub-modules distributed throughout the various stages of sintering production; by collecting production data and product quality inspection data during sintering production, it provides multi-source data for sintering production analysis. The edge preprocessing module includes an edge computing submodule, a local storage submodule, and an edge controller submodule. It removes noise and outliers through preprocessing to improve data quality, shortens the analysis time of multi-source data analysis servers, and provides temporary storage to ensure that data is not lost in the event of network or server failure, and that data is retransmitted after the network or server is restored. The actuator module sets the operating parameters of the sintering process and controls the smooth execution of the sintering production process through preset parameters; It also receives control commands from the edge controller submodule to precisely control the production equipment in real time; The data storage and analysis module includes a data storage submodule, a batching accuracy analysis submodule, a sintering process analysis submodule, an energy consumption control analysis submodule, a negative pressure ignition control analysis submodule, an air pressure and air volume control analysis submodule, and a machine speed and four-machine linkage analysis submodule. By analyzing multi-source data, the production equipment is adjusted in real time based on the analysis results, and the model parameters in each analysis submodule are dynamically adjusted in real time based on quality inspection data. The intelligent decision-making module includes a production control decision-making submodule and a decision-making scheme submodule; The decision evaluation and feedback module provides feedback and optimization to the decision-making model and reasoning mechanism by comparing the differences between actual production results and expected goals. The control distribution module distributes control commands generated by the data analysis module and optimal production control commands and optimal decision-making schemes generated by the intelligent decision-making module to the edge controller submodule via the industrial optical network.

8. The intelligent sintering system based on multi-source data analysis according to claim 7, characterized in that: The sensor submodule collects production data in real time during the sintering process by installing various sensors. The quality inspection equipment submodule obtains product quality inspection data by periodically conducting quality inspections on sintered products; The edge computing submodule uses filtering algorithms to remove noise from uploaded multi-source data, and interpolation algorithms and statistical analysis methods to remove outliers and fill in missing values. Principal component analysis, correlation analysis, and wavelet transform were used to extract key feature information related to sintering quality, efficiency, and cost from the processed data. The local storage submodule temporarily saves preprocessed data to the local storage device when the network is interrupted or the multi-source data analysis server fails. After the network interruption or multi-source data analysis server failure is restored, the data is retransmitted. It supports fast query and analysis of local data. The edge controller submodule receives control commands from the control module and then sends them to the actuator module. The data storage submodule is used to store data uploaded by edge preprocessing and quality inspection equipment. It adopts distributed file system and database technology for data storage and performs long-term storage and management of multi-source data; including a relational database for storing structured data. No relational databases are used to store unstructured data; The ingredient accuracy analysis submodule uses regression analysis and PID control algorithms to establish a correlation model between ingredient ratios and finished product chemical composition. It performs ingredient accuracy analysis on the actual ingredient quantity data from the uploaded multi-source data and the finished product chemical composition data from the quality inspection data. By comparing the deviation between the ingredient set value and the actual ingredient quantity in real time, it generates control commands and dynamically adjusts the model parameters based on the quality inspection data. The sintering process analysis submodule uses fuzzy logic control and adaptive control algorithms to construct a multi-factor coupled sintering process control model. It analyzes the sintering process of uploaded multi-source data and quality inspection data, generates control commands, and dynamically adjusts the model parameters based on the quality inspection data. The energy consumption control and analysis submodule uses a genetic algorithm and an energy consumption prediction model to perform energy consumption analysis on the uploaded multi-data, generate control commands, and dynamically adjust the parameters of the energy consumption prediction model based on the quality inspection data. The negative pressure ignition control and analysis submodule uses a PID control algorithm to establish a correlation model between negative pressure, ignition effect, and sintering quality. It performs negative pressure ignition analysis on real-time monitoring data from high-frequency negative pressure sensors, temperature and gas flow data from the igniter, and low-frequency quality inspection information of the sintered ore surface layer, generates control commands, and dynamically adjusts the parameters of the correlation model based on the uploaded quality inspection data. The wind pressure and air volume control analysis submodule uses a fuzzy adaptive control algorithm and mathematical model to analyze the uploaded high-frequency wind pressure and air volume sensor real-time data, sintering material layer permeability detection data and low-frequency sintering internal quality inspection information, generate control commands, and dynamically adjust the algorithm rules and parameters according to the quality inspection data. The four-machine linkage analysis submodule uses a neural network control algorithm to establish a collaborative control model between the sintering machine, the feeding machine, the crusher, and the cooler. It performs four-machine linkage analysis on the uploaded multi-source data and quality inspection data, generates control commands, and dynamically adjusts the weights and thresholds of the multi-objective optimization algorithm based on the quality inspection data. The production control decision-making submodule adopts a multi-objective optimization algorithm, constructs an intelligent decision-making model based on process mechanism and historical data, and comprehensively considers the control instructions and production objectives of each algorithm to generate the optimal production control strategy. The decision-making scheme submodule uses rule-based reasoning, case-based reasoning, and fuzzy reasoning to analyze and judge the output of multi-source data and analysis models. Based on different production conditions and constraints, it flexibly adjusts decision-making strategies to generate the optimal decision-making scheme.

9. An electronic device comprising a memory (102), a processor (101), a display module (103), and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the intelligent sintering method based on multi-source data analysis as described in any one of claims 1 to 6.

10. A readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent sintering method based on multi-source data analysis as described in any one of claims 1 to 6.

Images