Mapping and matching method and system for silicon-manganese alloy batching data and smelting furnace data

By constructing element balance model and correlation algorithm, the time delay characteristics of ingredients and smelting data are found and adjusted, the time delay problems of ingredients and smelting data in ferroalloy production are solved, real-time data matching and analysis are realized, and production efficiency and data are improved. Utilization.

CN114169689BActive Publication Date: 2024-11-12SHANGHAI LIGHT RING ENERGY TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111339611.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-11-12
Publication Date
2024-11-12
Estimated Expiration
2041-11-12

AI Technical Summary

Technical Problem

The time lag of batching and smelting furnace data in ferroalloy production leads to systematic errors in production data statistics and analysis, and the lack of effective data mapping and matching technology, which leads to the batching adjustment for several hours or even dozens of hours, making real-time analysis impossible.

Method used

By constructing a physical and chemical balance model of ingredients elements, using the correlation algorithm to find the hysteresis characteristics and delay range of data, determine the hysteresis index system for hysteresis time, gradually converge the optimal lag time between the ingredients and the furnace operation section, and establish an abnormal furnace The evaluation and judgment criteria of the secondary data are carried out, and the injection and real-time calculation of the batching data and furnace data are carried out, and the element balance model is regularly corrected.

Benefits of technology

Real-time process adjustment of the ingredients manager is realized, data utilization rate is improved, costs are reduced, data analysis and production improvement is simplified, in-depth research on the smelting process is avoided, and data analysis and indicator statistics are carried out conveniently and quickly.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114169689B_ABST
    Figure CN114169689B_ABST
Patent Text Reader

Abstract

The present invention provides a mapping and matching method and system for silicon-manganese alloy batching data and smelting furnace data, including: step 1: constructing a balance model of physical balance and chemical balance of batching elements, and finding the hysteresis characteristics and delay range of batching data and furnace data through a correlation algorithm; step 2: selecting the hysteresis time within the delay range for mapping and matching, and determining the evaluation index system of different hysteresis time periods; step 3: taking the physical and chemical balance of elements of raw materials and products as the limit condition, gradually converging the optimal lag time between the batching section and the furnace operation section under different working conditions; step 4: establishing abnormal furnace data evaluation and judgment criteria; step 5: performing real-time calculation on the batching data and furnace data, and using the calculation results to guide the batching. By adopting the method of the present invention, there is no need to add new hardware, and a large amount of data in existing production, testing, quality and energy consumption links is fully utilized, which has the advantages of improving the utilization rate of existing data and low cost.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data integration for ferroalloy smelting, and in particular to a mapping and matching method and system for silicon-manganese alloy batching data and smelting furnace data. Background Art

[0002] The two most important links in the ferroalloy production process are process batching and submerged arc furnace operation. Process batching is based on the properties of different manganese ores, coke and auxiliary materials, and is adjusted regularly based on the parameter feedback of the electric operation section. Both links are non-continuous and stable, and there is a natural time delay of several hours to dozens of hours between the two, resulting in systematic errors in production data statistics and subsequent data analysis:

[0003] 1. Batching: According to production needs and equipment characteristics, professional process engineers of ferroalloy production enterprises calculate and regularly adjust the proportion of manganese ore, coke and auxiliary ore through traditional batching tools and the current mineral properties, inventory, raw material test indicators, furnace test results and furnace characteristics of various ores, and change materials and additional materials in time; the batching link is a quasi-real-time assembly line operation, in which various ores are transported to the distribution layer of the submerged arc furnace in batches through automated loading carts, and data can be processed continuously in sections.

[0004] 2. Batching process: The batching process mainly relies on personal experience and there is no unified quantitative standard. The main reason is that each time the material is replaced, the material ratio is changed, and the batching data changes caused by the materials, it takes several hours or even more than ten hours to get effective feedback from the parameters of the furnace.

[0005] 3. Electrical operation: The furnace operator controls the primary and secondary current, voltage, and power factor within the ideal range by plugging and unplugging graphite electrodes and transformer gears according to the furnace conditions. The furnace is started and the products and slag are discharged after the power consumption of a single furnace reaches the set value. The products and slag of each furnace need to be sampled and tested, and the test results are fed back to the process engineer. There is also a lag of several hours in obtaining product test indicators.

[0006] 4. Overall process of ingredient adjustment: material ratio setting → furnace discharge index → ​​material ratio adjustment → index reaching the standard → material ratio maintenance. Since there is a large time lag between each link in the process, it takes dozens of hours to adjust the material ratio each time until the product index tends to be stable.

[0007] 5. Index statistics: It is impossible to accurately determine the corresponding relationship between ingredients and furnaces, and it is impossible to effectively analyze the real-time indicators of ferroalloy smelting. Only monthly level statistics of hundreds of furnaces can more accurately reflect the advantages and disadvantages of furnace conditions, ingredient indicators, energy consumption indicators, and output indicators.

[0008] At present, the overall ferroalloy ingredients and smelting business process is basically as follows: Figure 1 shown.

[0009] Based on the characteristics of ferroalloy proportioning, production and testing, there is no application technology in China that maps and matches the proportioning process of silicon-manganese alloy with the data of each smelting furnace.

[0010] Patent document CN202011048159 discloses a process for rapidly converting ordinary silicon-manganese alloy smelted in an electric arc furnace to a high-silicon silicon-manganese alloy, comprising the following steps: (1) batching; (2) preparation before conversion; (3) adjustment of the type of coke; (4) one day before conversion, ensuring that high-silicon silicon-manganese materials can be simultaneously fed into the electric arc furnace; (5) the slag alkalinity during the conversion transition period is 1.1 times that of conventional smelting; (6) the iron-tapping time is not less than 40 minutes; (7) dephosphorization outside the electric arc furnace. When the slag and iron liquid level in the ladle reaches 1 / 3, lime powder or fluorite powder is added until the P content in the high-silicon silicon-manganese alloy meets the standard and the addition is stopped.

[0011] Patent document CN202020833552 discloses a batching mechanism based on silicon-manganese alloy production, including a feed port opened on a smelting furnace, two rotating grooves are provided on the side walls of the feed port, and connecting blocks are rotatably connected in the two rotating grooves through rotating rods, and a sealing door is installed between the two connecting blocks. A mixing chamber is fixedly installed on the smelting furnace, and a partition is fixedly installed in the mixing chamber. Two feed seats are fixedly installed on the mixing chamber, and two receiving grooves are symmetrically provided on the inner wall of the mixing chamber, and single-layer baffles are rotatably connected in the two receiving grooves through round rods, and the two single-layer baffles are respectively attached to the two side walls of the partition.

[0012] Patent document CN112766815A (application number: CN202110162693.4) ​​discloses an online batching system for recycled aluminum based on a global optimal algorithm, including a server, a client, an industrial computer and a spectrometer. The server establishes a communication connection with the industrial computer, and the industrial computer establishes a communication connection with the spectrometer. The client is installed on the industrial computer; the server performs the following operations: obtaining the first parameter information to be batched input by the user through the client; the first parameter information includes: one or more combinations of brand information, furnace group information, and brand target value; based on the brand information and / or furnace group information, obtaining a batching plan; based on the brand target value and the batching plan, determining the type of raw materials and the quantity of each type of raw materials.

[0013] However, the above patent is about equipment modification and process adjustment in the field of ferroalloy production technology, and does not involve data matching algorithms for specific ingredients and production processes. Summary of the invention

[0014] In view of the defects in the prior art, the purpose of the present invention is to provide a mapping and matching method and system for silicon-manganese alloy batching data and smelting furnace data.

[0015] The mapping and matching method of silicon-manganese alloy batching data and smelting furnace data provided by the present invention includes:

[0016] Step 1: Based on the equipment data and historical operation data, a balance model of physical and chemical balance of batching elements is constructed to calculate the distribution coefficients of batching into products, into slag, and volatilization. Based on the quasi-continuous batching data and discontinuous out-of-furnace data, the hysteresis characteristics and delay range of batching data and furnace data are found through correlation algorithms;

[0017] Step 2: Select the hysteresis time within the delay range for mapping and matching, and determine the evaluation index system for different hysteresis periods;

[0018] Step 3: Taking the physical and chemical balance of raw materials and products as the limiting condition, gradually converge the optimal lag time between the batching section and the furnace operation section under different working conditions;

[0019] Step 4: Establish evaluation and judgment criteria for abnormal heat data;

[0020] Step 5: Perform real-time calculations on batching data and furnace data, use the calculation results to guide batching, and regularly revise the element balance model based on historical operating data.

[0021] Preferably, the elemental analysis and conventional proportion, additional mass and material change records of the three major categories of raw materials, namely, various types of manganese ore, coke and auxiliary materials, are used as the input basis for the balance calculation;

[0022] A comprehensive analysis of the components of the outgoing products, slag, dust and flue gas of the ore-fired furnace is carried out as a basis for physical and chemical balancing.

[0023] Preferably, the optimal time lag interval of the key element time is obtained according to the Pearson correlation coefficient, the Spearman correlation coefficient and the Kendall correlation coefficient.

[0024] Preferably, the index evaluation system of multiple delay windows is examined, and the silicon-manganese alloy production index under each window is calculated respectively;

[0025] Judge the physical and chemical balance of elements and the continuity of indicators under multiple windows, and output the optimal batching and blast furnace data matching results.

[0026] Preferably, abnormal process data are eliminated by statistically using the process capability index CPK and threshold exceeding limits.

[0027] The mapping and matching system of silicon-manganese alloy batching data and smelting furnace data provided by the present invention includes:

[0028] Module M1: Based on equipment data and historical operation data, build a balance model of physical and chemical balance of batching elements, calculate the distribution coefficients of batching into products, into slag, and volatilization, and find the hysteresis characteristics and delay range of batching data and furnace data through correlation algorithm based on quasi-continuous batching data and discontinuous out-of-furnace data;

[0029] Module M2: Select the hysteresis time within the delay range for mapping and matching, and determine the evaluation index system for different hysteresis periods;

[0030] Module M3: Taking the physical and chemical balance of raw materials and products as the limit condition, gradually converge the optimal lag time between the batching section and the furnace operation section under different working conditions;

[0031] Module M4: Establishing evaluation and judgment criteria for abnormal heat data;

[0032] Module M5: Perform real-time calculations on batching data and furnace data, use the calculation results to guide batching, and regularly revise the element balance model based on historical operating data.

[0033] Preferably, the elemental analysis and conventional proportion, additional mass and material change records of the three major categories of raw materials, namely, various types of manganese ore, coke and auxiliary materials, are used as the input basis for the balance calculation;

[0034] A comprehensive analysis of the components of the outgoing products, slag, dust and flue gas of the ore-fired furnace is carried out as a basis for physical and chemical balancing.

[0035] Preferably, the optimal time lag interval of the key element time is obtained according to the Pearson correlation coefficient, the Spearman correlation coefficient and the Kendall correlation coefficient.

[0036] Preferably, the index evaluation system of multiple delay windows is examined, and the silicon-manganese alloy production index under each window is calculated respectively;

[0037] Judge the physical and chemical balance of elements and the continuity of indicators under multiple windows, and output the optimal batching and blast furnace data matching results.

[0038] Preferably, abnormal process data are eliminated by statistically using the process capability index CPK and threshold exceeding limits.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] (1) By adopting the method of the present invention, a direct soft association between electrical operation data and batching data is established, and batchers can make real-time process adjustments through the model;

[0041] (2) By adopting the method of the present invention, there is no need to add new hardware, and a large amount of data in existing production, testing, quality and energy consumption links is fully utilized, which has the advantages of improving the utilization rate of existing data and low cost;

[0042] (3) By adopting the method of the present invention, in-depth research on the smelting process is avoided, which is conducive to manufacturers to conveniently and quickly carry out data analysis, indicator statistics and production improvement. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0044] Figure 1 It is the overall process flow chart of ferroalloy batching and smelting business at present;

[0045] Figure 2 is a flow chart of the mapping algorithm of the present invention;

[0046] Figure 3 This is a test effect diagram of the algorithm model of the present invention;

[0047] Figure 4 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0048] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0049] Example:

[0050] The mapping and matching method of silicon-manganese alloy batching data and smelting furnace data provided by the present invention includes: quantifying the experience of process engineers, combining the correlation analysis technology of batching station and ore-arc furnace smelting data to determine the material lag time window, using the element physical and chemical balance and key feature continuity theory to develop a mapping algorithm, generating the optimal lag time in a statistical sense, and using the CPK algorithm to discover and eliminate abnormal batching and furnace data, such as Figure 4 , the specific steps are as follows:

[0051] Step 1, balance model: According to the equipment characteristics and historical operation data, the physical and chemical balance models of manganese, iron, silicon, various alkali metals and other non-metallic elements are debugged to accurately calculate the distribution coefficients of various elements into products, into slag and volatilization; using quasi-continuous batching data and discontinuous furnace data, the correlation algorithm is used to find the time lag characteristics and delay range of batching data and furnace data;

[0052] Step 2, mapping algorithm design: The equilibrium model calculates the hysteresis characteristics, selects no less than three hysteresis times (such as Q25, Q50, Q75) to participate in the mapping algorithm, and determines the evaluation index system under different hysteresis periods.

[0053] Step 3, algorithm tuning: Based on the evaluation index system established by the mapping algorithm, the time lag characteristics of batching data and furnace data mining, the optimal continuity of evaluation indicators of ferroalloy raw materials and products is taken as the optimal goal, and the elemental physical and chemical balance of raw materials and products is taken as the limit condition, and the optimal lag time between the batching section and the furnace operation section under different working conditions is gradually converged.

[0054] Step 4, abnormal heat data processing: After model training and tuning are completed, CPK and threshold statistical methods are used to establish abnormal heat data evaluation and judgment criteria;

[0055] Step 5, deployment and execution: After the algorithm is deployed, accurate mapping and real-time calculation of batching and furnace data are achieved. The real-time calculation results are used to guide batching. The algorithm model also has the function of regular correction using past operating data.

[0056] The main elements (including Mn, Fe, Si, Ca, Mg, Al, C, P, S and other elements) of the three major categories of raw materials, namely, various types of manganese ore, coke and auxiliary materials, were fully analyzed and the conventional proportions, additional mass and material change records were selected as the input basis for the balance calculation. The full analysis of the components on one side of the blast furnace, such as the out-of-furnace products, slag, dust ash and flue gas, was selected as the core factor for physical and chemical balancing. The distribution coefficients of various elements into the product, into the slag and volatilization can be automatically generated and manually adjusted. Through the correlation algorithms such as Pearson, spearman, Kendall, etc., the optimal time lag interval of the key element time is obtained.

[0057] In step 2, the index evaluation system of multiple delay windows is examined, and the silicon manganese alloy production index under each window is calculated respectively. In step 3, a set of judgment criteria can be set to judge the elemental physical and chemical balance and index continuity index under multiple windows, and output the optimal batching and submersible arc furnace data matching results. The output result is not a fixed delay time, but a time vector set that is expressed in multiple element dimensions and changes dynamically. In step 4, a processing flow plan for abnormal furnace data is proposed, and abnormal process data is eliminated through the statistical method of CPK and threshold over-limit. Step 5 includes a mapping matching correction process, and the correction target is a stage-by-stage element balance.

[0058] Specifically, according to the ideas and steps of the present invention, by collecting data from 7 submerged arc furnaces for more than half a year, the following calculation example is given:

[0059] 1. First, create an element balance model based on short-time lag evaluation (through correlation analysis or similar data algorithms) based on the collected laboratory data, electrical operation data and empirical coefficients;

[0060] 2. Establish a set of indicator assessment system that integrates batching and electrical operation data;

[0061] 3. Based on the CPK statistical algorithm, establish the stability and balance judgment criteria of the indicator evaluation system;

[0062] 4. Based on the experience of each factory, establish a large window balance coefficient correction model;

[0063] 5. Generate a mapping model, including abnormal data removal and automatic model update.

[0064] The process of the mapping algorithm is as follows Figure 2 shown.

[0065] The test results of the generated algorithm model are as follows: Figure 3 As shown in the figure, only 1.8% of abnormal data are removed, and the error rate of 138 indicators in the indicator evaluation system is below 1%. After the algorithm has been running for a long time, if the process engineer or electrical engineer finds that the error has increased, the model can be adjusted according to the characteristics of the equipment, and the new mapping model can be updated by repeating the large window correction.

[0066] The present invention does not involve equipment modification, but mainly improves the data mismatch problem in the ferroalloy production process from the aspects of data analysis and data transparency. Combining industrial practice, equilibrium theory and big data analysis, it is found that the ferroalloy production process can improve exchange settlement and improve production efficiency.

[0067] Those skilled in the art know that, in addition to implementing the system, device and its various modules provided by the present invention in a purely computer-readable program code, it is entirely possible to implement the same program in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers and embedded microcontrollers by logically programming the method steps. Therefore, the system, device and its various modules provided by the present invention can be considered as a hardware component, and the modules included therein for implementing various programs can also be considered as structures within the hardware component; the modules for implementing various functions can also be considered as both software programs for implementing the method and structures within the hardware component.

[0068] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A mapping and matching method for silicon-manganese alloy batching data and smelting heat data, characterized in that: include: Step 1: Based on the equipment data and historical operation data, a balance model of physical and chemical balance of batching elements is constructed to calculate the distribution coefficients of batching into products, into slag, and volatilization. Based on the quasi-continuous batching data and discontinuous out-of-furnace data, the hysteresis characteristics and delay range of batching data and furnace data are found through correlation algorithms; Step 2: Select the hysteresis time within the delay range for mapping and matching, and determine the evaluation index system for different hysteresis periods; Step 3: Based on the evaluation index system established by the mapping algorithm, the time lag characteristics of batching data and furnace data mining, the optimal continuity of the evaluation indexes of ferroalloy raw materials and products is taken as the optimal goal, and the elemental physical and chemical balance of raw materials and products is taken as the limit condition, and the optimal lag time between the batching section and the furnace operation section under different working conditions is gradually converged; Step 4: After model training and tuning are completed, use CPK and threshold statistical methods to establish abnormal furnace data evaluation and judgment criteria; Step 5: After the algorithm is deployed, accurate mapping and real-time calculation of batching and furnace data are achieved. The real-time calculation results are used to guide batching. The algorithm model also has the function of regular correction using past operation data. Investigate the index evaluation system of multiple delay windows and calculate the silicon manganese alloy production index under each window respectively; Judge the physical and chemical balance of elements and the continuity of indicators under multiple windows, and output the optimal batching and submerged arc furnace data matching results; Abnormal process data are eliminated by statistical methods of process capability index CPK and threshold exceeding limit; The complete analysis of the main elements and conventional proportions, additional mass and material change records of the three major types of raw materials, namely, various types of manganese ore, coke and auxiliary materials, are selected as the input basis for the balance calculation; the complete analysis of the components on one side of the blast furnace, including the out-of-furnace products, slag, dust and flue gas, is selected as the core factor for physical and chemical balancing; the distribution coefficients of various elements into the product, into the slag and into volatilization can be automatically generated and manually adjusted; the optimal time lag interval of the key element time is obtained through the Pearson, spearman and Kendall correlation algorithms.

2. A mapping and matching system for silicon-manganese alloy batching data and smelting furnace data, characterized in that: include: Module M1: Based on equipment data and historical operation data, build a balance model of physical and chemical balance of batching elements, calculate the distribution coefficients of batching into products, into slag, and volatilization, and find the hysteresis characteristics and delay range of batching data and furnace data through correlation algorithm based on quasi-continuous batching data and discontinuous out-of-furnace data; Module M2: Select the hysteresis time within the delay range for mapping and matching, and determine the evaluation index system for different hysteresis periods; Module M3: Based on the evaluation index system established by the mapping algorithm, the time lag characteristics of batching data and furnace data mining, the optimal continuity of evaluation indicators of ferroalloy raw materials and products is taken as the optimal goal, and the elemental physical and chemical balance of raw materials and products is taken as the limit condition, and the optimal lag time between the batching section and the furnace operation section under different working conditions is gradually converged; Module M4: After model training and tuning are completed, CPK and threshold statistical methods are used to establish abnormal furnace data evaluation and judgment criteria; Module M5: After the algorithm is deployed, accurate mapping and real-time calculation of batching and furnace data are achieved. The real-time calculation results are used to guide batching. The algorithm model also has the function of regular correction using past operation data. Investigate the index evaluation system of multiple delay windows and calculate the silicon manganese alloy production index under each window respectively; Judge the physical and chemical balance of elements and the continuity of indicators under multiple windows, and output the optimal batching and submerged arc furnace data matching results; Abnormal process data are eliminated by statistical methods of process capability index CPK and threshold exceeding limit; The complete analysis of the main elements and conventional proportions, additional mass and material change records of the three major types of raw materials, namely, various types of manganese ore, coke and auxiliary materials, are selected as the input basis for the balance calculation; the complete analysis of the components on one side of the blast furnace, including the out-of-furnace products, slag, dust and flue gas, is selected as the core factor for physical and chemical balancing; the distribution coefficients of various elements into the product, into the slag and into volatilization can be automatically generated and manually adjusted; the optimal time lag interval of the key element time is obtained through the Pearson, spearman and Kendall correlation algorithms.