Method for continuously smelting copper concentrate through cooperation of three continuous furnaces

Abstract

The invention relates to the technical field of ore smelting, in particular to a method for continuously smelting copper concentrate through cooperation of three continuous furnaces, which comprises the following steps: 1) adding smelting raw materials into a side blowing furnace, and feeding produced matte into a converting furnace; (2) grade and material amount intelligent monitoring is conducted on the matte, and crude copper is produced; (3) the crude copper is alternately loaded into different anode furnaces for refining, and furnace condition indexes in the refining process are intelligently monitored; 4) constructing a metallurgical digital-analog collaborative engine model according to the monitored data; 5) constructing a three-dimensional twinborn scene; and 6) completing twinborn mapping. According to the method, by comprehensively sensing the furnace condition state, perfecting the control parameter adjustment basis and based on the smelting mechanism and the intelligent method, intelligent decision making and precise control over process parameters are achieved, when the working conditions of raw materials and front procedures are changed, recognition is conducted in advance, preprocessing is achieved, the furnace condition and product stability is improved, and the production cost is reduced. Therefore, heat utilization rate in the furnace is improved, fuel consumption is reduced, and energy conservation and environmental protection are achieved.

Description

Technical Field

[0001] This invention relates to the field of ore smelting technology, and in particular to a method for continuous smelting of copper concentrate using three furnaces in a coordinated manner. Background Technology

[0002] Copper concentrate is the main raw material for copper smelting, containing copper and various other metallic and non-metallic components. During the smelting process, the copper concentrate needs to be pretreated to bring it to a suitable state for smelting. Pretreatment typically includes steps such as crushing, grinding, and beneficiation to improve copper recovery and reduce impurity content. Subsequently, the pretreated copper concentrate is fed into a smelting furnace for high-temperature melting, separating copper from other impurities to ultimately obtain crude copper.

[0003] In copper smelting, traditional CNC methods rely primarily on fixed process parameter settings. Stable operation is achieved by pre-setting and controlling key parameters such as furnace temperature, airflow, and feed rate. Specifically, operators set furnace temperature curves based on experience, controlling the heating rate and temperature range at different stages. Simultaneously, airflow is adjusted to maintain a suitable redox atmosphere within the furnace, ensuring effective separation of copper from other impurities. Regarding feed rate control, pretreated copper concentrate is added to the furnace at predetermined times and quantities according to a pre-defined feeding plan. However, this traditional CNC method lacks the ability to perceive and dynamically adjust actual smelting conditions in real time. It struggles to optimize process parameters promptly based on fluctuations in copper concentrate composition and changes in furnace reactions, resulting in limitations on smelting efficiency and room for improvement in product quality stability.

[0004] Chinese patent CN120296156A discloses an oxygen-enriched side-blown copper smelting process control system based on LLM and RAG. It relies on Large Language Model (LLM) and Retrieval Enhanced Generation (RAG) technology to generate control suggestions through semantic vector mapping and an expert experience base, thus recommending smelting parameters. This technology has the following drawbacks: 1) Strong dependence on static knowledge: The RAG module's retrieval relies on pre-stored expert experience and literature, failing to dynamically respond to real-time changes in multi-source heterogeneous data during smelting (such as fluctuations in raw material composition and abrupt changes in sensor data), leading to a disconnect between control suggestions and actual operating conditions; 2) Lack of multi-furnace coordination mechanism: Designed only for a single oxygen-enriched side-blown furnace, it does not consider the cross-process transfer of materials / energy in a three-furnace system (smelting-blowing-refining), failing to solve the problem of multi-furnace parameter coupling (such as the chain reaction of side-blown furnace slag grade fluctuations on top-blown furnace oxygen consumption); 3) Insufficient model interpretability: The "black box" reasoning logic of LLM makes it difficult to provide traceable metallurgical evidence, preventing process engineers from verifying the scientific validity of parameter recommendations and reducing system credibility.

[0005] Chinese patent CN119689859A discloses a data-driven intelligent control system and method for continuous smelting in multiple furnaces. Based on a continuous smelting simulation model and a fuzzy rule base, it optimizes the process by dynamically adjusting the parameters of multiple input and multiple output equipment. This technology has the following shortcomings: 1) Insufficient data fusion depth: It only collects conventional data such as temperature and oxygen content, lacking online sensing capabilities for key variables (raw materials, matte and side-blown slag composition, anode furnace flue gas composition, etc.), resulting in insufficient input feature dimensions for the simulation model and limited prediction accuracy; 2) Lack of multi-furnace collaboration mechanism: While a continuous smelting simulation model is constructed through a machine learning module, the data acquisition and analysis process does not fully cover the three-furnace (smelting-blowing-refining) operation module, lacking collaborative guidance significance. Summary of the Invention

[0006] To address the above problems, the present invention provides a method for continuous smelting of copper concentrate using three furnaces, comprising the following steps: 1) After the smelting raw materials are proportioned, they are added to the side-blown furnace. Based on the model calculation, the required oxygen-enriched air is delivered to the side-blown furnace, and the produced matte enters the smelting furnace. 2) Intelligent monitoring of the grade and quantity of matte produced in step 1), collection of furnace condition data, calculation of the required flux, oxygen and cold material to be added through model, and production of crude copper; 3) The crude copper produced in step 2) is alternately loaded into different anode furnaces for refining, and the furnace condition indicators are intelligently monitored during the refining process; 4) Construct a metallurgical numerical simulation collaborative engine model based on the monitored data to predict the oxidation endpoint; 5) Construction of 3D Twin Scene Based on the twin digital model layer, it integrates different OT data, 3D model data, and equipment data. Through OPC-UA protocol conversion, it realizes real-time forwarding of OT layer data to IT system. Through DataSmith and programmatic model generation, it imports 3D related equipment for use by the engine layer. It builds a high-fidelity and parametric model of the furnace body in the 3D engine and connects the data interface to the scene. 6) Complete twin mapping Based on high-precision 3D models and multiphysics finite element simulation, a dynamic and visualized digital twin scene is realized, which maps the smelting process status in real time and intelligently adjusts the process parameters under different raw material ratio inputs, thereby realizing dynamic optimization and closed-loop control of production strategies.

[0007] Specifically, in step 1), the raw materials for smelting are periodically tested in the laboratory for grade. Combined with data from the planned batching sheet and belt scale, the heterogeneous multi-source data is integrated to dynamically calculate the amount of material fed into the furnace and the overall grade data. An online automatic raw material detector is installed at the side-blown furnace inlet. In step 2), an online matte composition detector is installed at the matte inlet. In step 3), an online flue gas analyzer is installed in the anode furnace. These detection devices enable real-time acquisition of key data during the smelting process. Laboratory grade testing provides a basic data reference for the overall smelting process. The raw material quantity and overall grade data obtained by integrating the planned batching sheet and belt scale data provide accurate material information for subsequent smelting stages. The online automatic raw material detector at the side-blown furnace inlet provides timely feedback on the raw material status, ensuring that the raw materials entering the side-blown furnace meet smelting requirements. The online matte composition detector installed at the matte inlet accurately detects the composition of the matte, allowing for adjustments to smelting parameters based on actual conditions. The online flue gas analyzer on the anode furnace can analyze the flue gas generated by the anode furnace, providing important data support for the refining process, thereby ensuring the stable and efficient operation of the entire three-furnace co-processing continuous copper concentrate smelting process.

[0008] Based on the above equipment configuration, support is provided for predicting oxidation endpoints by integrating and modeling the detected data. In step 4), the metallurgical numerical modeling collaborative engine model is as follows: For a stable production process, the equilibrium relationship of a certain element E in the system is as follows: ; Among them, element E is Cu, Fe, S, O, and C. The grade of element E in the material fed into the furnace. The amount of material fed into the furnace. The grade of element E in the material fed into the furnace. This refers to the amount of material fed into the furnace.

[0009] The method for continuous smelting of copper concentrate in a three-furnace complex of the present invention comprehensively senses the furnace condition and improves the basis for adjusting control parameters. Based on the smelting mechanism and intelligent methods, it realizes intelligent decision-making and precise control of process parameters. When the operating conditions of raw materials and upstream processes change, it can identify and pre-treat in advance, improve the stability of furnace conditions and products, thereby improving the heat utilization rate in the furnace, reducing fuel consumption, and saving energy and protecting the environment. Attached Figure Description

[0010] Figure 1 This is a diagram of a four-level digital twin architecture in a specific implementation of the present invention.

[0011] Figure 2 This is a diagram of the digital control system architecture for three-furnace collaborative continuous smelting in a specific implementation of the present invention.

[0012] Figure 3 This is the calculation process for the top-blown furnace metallurgical model in a specific implementation of the present invention.

[0013] Figure 4 This is the calculation process for the side-blown furnace metallurgical model in a specific implementation of the present invention.

[0014] Figure 5 This is the calculation process of the intelligent prediction model for the oxidation endpoint of the anode furnace in a specific implementation of the present invention.

[0015] Figure 6 This is the parameter optimization and verification process based on the orthogonal experimental design approach for the specific implementation of this invention.

[0016] Figure 7 This is the test report for the produced crude copper. Detailed Implementation

[0017] The present invention will be described below with reference to examples. These examples are only used to explain the present invention and are not intended to limit the scope of the present invention.

[0018] This invention achieves twin-synergistic optimization of control parameters in the three-furnace smelting process by constructing a multi-source sensing system and a metallurgical mechanism-data-driven fusion model, forming an integrated control mode of "sensing-decision-execution" throughout the entire process, and providing core technical support for performance improvement and energy consumption optimization of continuous copper smelting process.

[0019] I. System Architecture The method of this invention constructs a four-level digital twin architecture of "physical layer - perception layer - model layer - application layer" to achieve closed-loop optimization of the entire process of three-furnace smelting-blowing-refining, such as... Figure 1 As shown, the functions of each level of the digital twin architecture are as follows.

[0020] Physical layer: The smelting furnace (oxygen-enriched side-blown), the blowing furnace (top-blown oxygen lance), and the refining furnace (anode furnace) are connected in series via chutes, and the furnaces work together.

[0021] Sensing layer: Deploy a multi-source sensor network covering the feeding, reaction, slag discharge, and flue gas processes.

[0022] Model Layer: A metallurgical numerical modeling collaborative engine is built, integrating metallurgical mechanism models and data-driven models to perform optimization decision analysis on key control parameters of the smelting process. The model input parameters of each core furnace are decoupled and coordinated, supporting both independent analysis and computation, as well as coherent collaborative analysis.

[0023] Application Layer: The twin application layer receives output instructions from the model, dynamically maps them to the virtual furnace, and realizes predictive analysis of the control results.

[0024] Under this collaborative optimization method, the specific operational steps of the three-furnace process for continuous copper concentrate smelting are as follows: 1) Based on the copper concentrate batching ratio set by the technical team, the batching system calculates the key grades and moisture content of the mixed copper concentrate, and calculates the required amounts of flux, coke, and intermediate products to be added; the material preparation process prepares various copper concentrates, fluxes, and auxiliary fuels according to the calculated ratio, and then transports them to the furnace top silo by belt conveyor; the mixed materials are then fed into the side-blown furnace from the top of the smelting furnace through the three feed ports via a plow-type unloader by a belt conveyor. 2) Based on the model calculation of the required air volume and oxygen volume, oxygen-enriched air is delivered into the furnace through the side blowing hole, where a chemical reaction occurs, controlling the temperature of the smelting process, producing matte, and separating the smelting slag in the furnace. The matte settles to the bottom layer and is discharged through the siphon port, then enters the blowing furnace through the chute, and the high-temperature flue gas is discharged to the subsequent section through the flue. 3) Intelligent monitoring of the grade and quantity of continuously charged matte is carried out. The furnace condition data is collected and the required flux, oxygen and cold material are calculated through the model to achieve the target grade of crude copper and control the process temperature in the furnace. The produced crude copper enters the anode furnace through the chute. The produced smelting slag is granulated by the granulation device and then recycled. The high temperature flue gas is discharged to the subsequent section through the flue. 4) According to the system plan, crude copper is alternately loaded into different anode furnaces for refining operations. The furnace condition indicators are intelligently monitored during the operation, the end point of the oxidation operation is predicted, and intelligent operation in the furnace is realized.

[0025] This digital twin-based collaborative optimization method, through Figure 2 The digital management and control system architecture shown is implemented in detail. For example... Figure 2 As shown, the system architecture comprises four layers from bottom to top: Data acquisition and sensing layer: Corresponding sensor networks are deployed for side-blown smelting furnaces, top-blown smelting furnaces, and anode refining furnaces to collect key process parameters such as oxygen concentration, oxygen lance pressure, and oxidation-reduction temperature in real time, forming the basis of system sensing.

[0026] Data Fusion and Processing Center: Receives multi-source heterogeneous data from the perception layer, and outputs a pre-processed, high-quality, time-consistent dataset through data cleaning, feature engineering, standardization, and other processing procedures, providing reliable input for upper-layer models.

[0027] Multi-furnace collaborative optimization engine: As the core of the system, it utilizes the fused data provided by the lower layer to execute material balance tracking, heat balance calculation and intelligent optimization algorithms to generate the optimal set of control parameters for the collaborative operation of three furnaces.

[0028] Control execution and feedback: The decision instructions of the optimization engine are issued to the actuators of each furnace, and the process execution effect is monitored in real time. Feedback data (such as the deviation between actual operating conditions and expectations) is returned to the upper engine, thereby realizing closed-loop optimization control of the entire process.

[0029] II. Implementation of Multi-Source Heterogeneous Data Fusion 1) Multi-dimensional data acquisition is achieved, eliminating blind spots in perception and providing high-fidelity input for the model. Based on the acquisition of all real-time production data from the existing DCS, online detection data of raw materials entering the side-blown furnace, online detection data of matte and side-blown slag composition, and flue gas detection data during the anode furnace feeding period are introduced for automatic verification, providing scientific guidance for parameter control in the smelting process.

[0030] ① Material sensing for furnace entry An online automatic raw material detector is installed at the feed inlet of the side-blown furnace to realize real-time detection of concentrate grade (Cu / Fe / SiO2 content, etc.) with an accuracy of ±0.5%. The raw materials entering the furnace are regularly tested for grade in the laboratory. Combined with the planned batching sheet and belt scale data, the above heterogeneous multi-source data are integrated to dynamically calculate the amount of material entering the furnace and the overall grade data.

[0031] ②Perception of work process status An online copper matte composition analyzer is installed at the top-blown furnace feed chute to achieve high-frequency copper matte grade analysis (Cu / Fe / SiO2 content, etc.) with an accuracy of ±0.5%. An online detector for side-blown slag composition is installed at the slag discharge port of the side-blown furnace to quickly detect the composition of the side-blown slag and provide data support for the feedback control of the side-blown furnace. The selection of an appropriate online flue gas analyzer for the anode furnace enables accurate detection of flue gas under different SO2 concentration conditions, providing a basis for judging the sulfur content of crude copper and the operating status of the anode furnace.

[0032] ③ Soft measurement of key process indicators Continuous copper smelting operations use chutes for material transfer. Due to the characteristics of high-temperature melt, the melt flow rate is difficult to measure directly. Taking the amount of matte charged as the analysis object, its fluctuation will affect the calculation and control of the top-blown furnace operation parameters. Based on the furnace condition parameters of the top-blown furnace and the liquid level data of the side-blown furnace, a coupled prediction model is established to output the estimated value of the amount of matte charged. The amount of oxygen supplied to the furnace is affected by many factors inside the furnace and does not match the theoretical amount of oxygen required for the reaction based on the model calculation. Therefore, a conversion is required to construct an oxygen utilization rate calculation model, which can optimize and correct the theoretical oxygen consumption.

[0033] 2) Construct a data fusion pipeline to address the heterogeneity issue between time-series and non-time-series data and generate a unified feature vector. This data processing process includes two core elements: First, focusing on data quality optimization, this involves removing abnormal noise and redundant information through data cleaning, and filling in missing values ​​using data compensation strategies to ensure the reliability of the data foundation. Second, overcoming the challenge of multi-source data collaboration, this involves integrating data features of different dimensions and granularities using multi-scale data fusion technology, and simultaneously achieving temporal uniformity of continuous copper smelting operation data through a data time axis sliding matching mechanism, ensuring the temporal consistency and feature integrity of the fused data.

[0034] ① Data cleaning Based on the verification and analysis of data source quality, a special model for data preprocessing is constructed to address key issues such as sensor environmental noise interference, inherent noise of process characteristics, and the need for complementary measurement ranges. ② Feature fusion Time-series data (temperature, belt scale) are processed by extracting frequency domain features using DB wavelet transform, while non-time-series data (grade, composition) are normalized using Z-score. Data is managed in a unified and standardized manner, avoiding the storage of data in multiple formats. For data calculations with inconsistent time axes, a calculation period can be set, and operations such as accumulation and averaging can be performed on the data according to the calculation period to ensure consistent units and data matching.

[0035] For time-series data (such as temperature and belt scale data), the DB wavelet transform method is used to extract its frequency domain features; for non-time-series data (such as grade and composition data), the Z-score normalization method is used to achieve data standardization; process data management specifications are established to achieve unified data management and avoid management redundancy caused by multi-format data storage; for systems with high latency and large lag characteristics, a time shift window is set to accurately match and dynamically adjust the data time axis to eliminate the impact of latency on data time-series consistency.

[0036] III. Construction of a Collaborative Engine for Metallurgical Numerical Modeling 1) Based on the smelting process mechanism, a metal-heat balance equation is constructed, and the optimal parameters for multi-furnace coordination are solved under the constraints of the law of conservation of matter and the first law of thermodynamics.

[0037] ① For a stable production process, the equilibrium relationship of a certain element E in the system is: ; Among them, element E is Cu, Fe, S, O, and C. The grade of element E in the material fed into the furnace. The amount of material fed into the furnace. The grade of element E in the material fed into the furnace. This refers to the amount of material fed into the furnace.

[0038] ② The thermal balance of a copper smelting furnace follows the law of conservation of energy, where the sum of heat input equals the sum of heat output. The main sources of heat input include heat released by chemical reactions and physical heat, while heat output mainly includes heat carried away by materials, consumption of cold materials, and losses of cooling water and heat.

[0039] 2) Supplement the dynamic parameter optimization mechanism to improve the model's adaptability to raw material fluctuations. Based on phase analysis, and combined with the fluctuation range of key influencing factors such as concentrate grade / copper matte grade, comprehensively consider the furnace condition rationality index, conduct operating condition tests according to a matrix, and output production benchmark values ​​under different operating conditions after data preprocessing and statistical analysis. Based on the final test results, use a parameter fitting algorithm to update the oxygen utilization rate and material loss rate coefficients.

[0040] The metallurgical numerical modeling collaborative engine constructs a dedicated metallurgical calculation model for each core furnace body in the three-furnace complex. Its specific implementation path is as follows: Figures 3-5 As shown.

[0041] Implementation of top-blown and side-blown furnace models: such as Figure 3 and Figure 4 As shown, the model construction of top-blown furnaces and side-blown furnaces follows the same logical framework. First, the input-output phase analysis provides the basis for model calculation; the core metallurgical calculation model (MB: material balance model, HB: heat balance model) is used for simulation calculation to meet specific functional requirements such as batching calculation, flux calculation, and air-oxygen calculation; the data support required for calculation and the verification and optimization of results are obtained through the industrial data platform to realize the on-site time-series data acquisition, processing and analysis at the edge, forming a closed loop from edge equipment data to model decision-making and then feedback to production.

[0042] The calculation process of the intelligent prediction model for the oxidation endpoint of the anode furnace is as follows: Figure 5 As shown, the established industrial data platform can monitor furnace condition data in real time, enabling the classification and monitoring of furnace operating conditions. Furthermore, the platform provides a complete technology chain for dataset construction, including data acquisition, preprocessing, outlier labeling, and data transformation. By training the dataset, a predictive model for the oxidation endpoint can be built. Production data indicators can be used to classify the anode furnace operating conditions. Based on the trained predictive model, intelligent prediction of key process indicators (such as the anode furnace oxidation endpoint) can be achieved, providing core parameters for precise control.

[0043] After completing the basic model construction, the metallurgical numerical modeling collaborative engine employs a systematic model verification and optimization method to ensure that its calculation parameters and output results accurately match actual production conditions. The core process is as follows: Figure 6As shown in the figure. This method uses orthogonal experimental design to efficiently obtain representative production data of key process parameters at different levels, and then calibrates and verifies the model, effectively solving the problem of establishing a reliable benchmark due to complex working conditions and large data fluctuations in smelting sites.

[0044] IV. Application of Digital Twins in Smelting Processes 1) Construction of 3D Twin Scene Based on the twin model layer, it integrates different OT data, 3D model data, equipment data, etc., and realizes real-time forwarding of OT layer data to IT system through OPC-UA protocol conversion. It imports 3D related equipment through DataSmith and programmatic model generation for use by the engine layer.

[0045] ① Construct a high-fidelity and parametric model of the furnace body in a 3D engine; ② Data interface access scenarios 2) Twin mapping Based on high-precision 3D models and multiphysics finite element simulation, it enables dynamic visualization of digital twin scenarios, accurately and in real time maps the smelting process status, and intelligently adjusts the process parameters under different raw material ratio inputs, thereby realizing dynamic optimization and closed-loop control of production strategies.

[0046] ① To meet the needs of on-site smelting production, different smelting production statistics interfaces can be flexibly configured, including real-time status of the production process, data curves, key data statistics results, etc. ② Construct a parameterized model of the furnace body for simulation calculation, bind the recommended air and oxygen values ​​to the oxygen lance particle emitter in real time, simulate the airflow field distribution, and map the oxygen utilization rate to the melt stirring flow rate color level; ③ Integrate the mechanism calculation model to create an intelligent prediction and optimization system for the smelting process.

[0047] Based on the above method, this invention provides the following demonstration of actual operation process: the numerical display of the control interface of the three-furnace continuous copper smelting collaborative system at a certain moment, as shown in Tables 1 to 6. Table 1 shows the equipment parameter input control, including preset input values ​​and real-time collected actual values; Table 2 shows the target values ​​and actual values ​​of the simulation calculation results; Table 3 shows the target values ​​and actual values ​​of key intermediate quantities; Table 4 shows the control target setpoints and actual values; Table 5 shows the calculated values ​​and actual values ​​of control parameters; and Table 6 shows the theoretical values ​​and correction values ​​of control conditions. From the preset values ​​and instantaneous values ​​displayed in Tables 1 to 6, the status of the three-furnace continuous copper smelting can be intuitively understood. Anomalies can be detected and adjusted in real time, thereby ensuring stable smelting operations.

[0048] Table 1. Preset and Actual Values ​​of Equipment Parameters

[0049] Table 2. Target and Actual Values ​​of Simulation Results

[0050] Table 3. Target and Actual Values ​​of Key Intermediates

[0051] Table 4. Control target setpoints and actual values

[0052] Table 5. Calculated and Actual Values ​​of Control Parameters

[0053] Table 6. Theoretical and Corrected Values ​​of Control Conditions

[0054] Figure 7 The test report for the crude copper produced using the above system shows that the copper content is high, while the contents of other elements are within a suitable range.

[0055] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for continuous smelting of copper concentrate using three furnaces in a coordinated manner, characterized in that, Includes the following steps: 1) After the smelting raw materials are proportioned, they are added to the side-blown furnace. Based on the model calculation, the required oxygen-enriched air is delivered to the side-blown furnace, and the produced matte enters the smelting furnace. 2) Intelligent monitoring of the grade and quantity of matte produced in step 1), collection of furnace condition data, calculation of the required flux, oxygen and cold material to be added through model, and production of crude copper; 3) The crude copper produced in step 2) is alternately loaded into different anode furnaces for refining, and the furnace condition indicators are intelligently monitored during the refining process; 4) Construct a metallurgical numerical simulation collaborative engine model based on the monitored data to predict the oxidation endpoint; 5) Construction of 3D Twin Scene Based on the twin digital model layer, it integrates different OT data, 3D model data, and equipment data. Through OPC-UA protocol conversion, it realizes real-time forwarding of OT layer data to IT system. Through DataSmith and programmatic model generation, it imports 3D related equipment for use by the engine layer. It builds a high-fidelity and parametric model of the furnace body in the 3D engine and connects the data interface to the scene. 6) Complete twin mapping Based on high-precision 3D models and multiphysics finite element simulation, a dynamic and visualized digital twin scene is realized, which maps the smelting process status in real time and intelligently adjusts the process parameters under different raw material ratio inputs, thereby realizing dynamic optimization and closed-loop control of production strategies.

2. The method according to claim 1, characterized in that, In step 1), the smelting raw materials are regularly tested in the laboratory for grade. Combined with the planned batching list and belt scale data, the above heterogeneous multi-source data are integrated to dynamically calculate the amount of material fed into the furnace and the comprehensive grade data. The side-blown furnace feed port is equipped with an online automatic raw material detector.

3. The method according to claim 1, characterized in that, In step 2), an online copper matte composition analyzer is installed at the copper matte inlet.

4. The method according to claim 1, characterized in that, In step 3), the anode furnace is equipped with an online flue gas analyzer.

5. The method according to any one of claims 1 to 4, characterized in that, In step 4), the metallurgical numerical modeling collaborative engine model is as follows: For a stable production process, the equilibrium relationship of a certain element E in the system is as follows: ； Among them, element E is Cu, Fe, S, O, and C. The grade of element E in the material fed into the furnace. The amount of material fed into the furnace. The grade of element E in the material fed into the furnace. This refers to the amount of material fed into the furnace.

Images