Mechanism-data driven converter energy consumption and carbon emission prediction modeling method

Abstract

The invention discloses a mechanism-data driven converter energy consumption and carbon emission prediction modeling method. On the basis of analyzing the relationship between substances and energy flow in the converter steelmaking process, a theoretical analysis method is adopted, and a mechanism model of nitrogen, argon, compressed air and electricity consumption in the converter steelmaking process is established. A data driving mode is used, an MIC method is adopted for feature selection, and then a WOA-XGBoost method is used for constructing a data model of oxygen consumption, converter gas recovery, steam recovery and molten steel yield. And finally, combining the mechanism and the data model to obtain a converter ton steel energy consumption and ton steel carbon emission model to realize prediction. The method can provide support for iron and steel enterprises to realize energy conservation and carbon reduction, and is of great significance to green development of the enterprises.

Description

Technical Field

[0001] The present invention relates to the field of iron and steel metallurgy industry, and in particular to a converter energy consumption and carbon emission prediction modeling method. Background Art

[0002] Iron and steel metallurgy is a typical high-energy consumption and high-emission industry. The converter steelmaking process is one of the key processes that affect the total energy consumption and carbon emissions of iron and steel metallurgy. Its smelting process consumes multiple energy media such as electricity, oxygen, and nitrogen, produces a large amount of converter gas and steam, and emits carbon dioxide. It is one of the main processes for energy consumption, carbon dioxide emissions, and secondary energy recovery in iron and steel metallurgy. Using technical means such as industrial Internet of Things and big data analysis combined with theoretical methods, the influence of converter steelmaking process parameters on energy consumption and carbon emissions is analyzed, and a converter energy consumption and carbon emission model is established. This helps enterprises predict the energy consumption and carbon dioxide emissions of the converter steelmaking process, and provides support for enterprises to formulate energy-saving and carbon reduction plans. It has important research significance for promoting the green development of enterprises. Summary of the invention

[0003] The present invention provides a mechanism-data driven converter energy consumption and carbon emission prediction modeling method. Combining theoretical principles and data analysis methods, a converter energy consumption per ton of steel and carbon emission per ton of steel model is established to achieve the prediction of converter energy consumption and carbon emission.

[0004] The technical solution of the present invention is: a mechanism-data driven converter energy consumption and carbon emission prediction modeling method, comprising the following steps:

[0005] Step 1: Analyze the relationship between material and energy flow in the converter steelmaking process to obtain the material consumption and energy consumption of the converter steelmaking process.

[0006] Step 2: Use theoretical principles to analyze the production and consumption of energy media such as oxygen, nitrogen, and electricity in converter steelmaking, and establish mechanism models for nitrogen, argon, compressed air, and electricity.

[0007] Step 3: Use a data-driven approach to establish a data model for oxygen consumption, converter gas recovery, steam recovery, and molten steel production in converter steelmaking.

[0008] Step 4: Combine the mechanism and data model of each energy medium and molten steel output to establish a model of converter energy consumption per ton of steel and carbon emissions per ton of steel to achieve prediction;

[0009] Preferably, the specific process of the energy medium production and consumption analysis and the construction of the partial energy medium mechanism model in step 2 is as follows:

[0010] (1) Oxygen consumption

[0011] Oxygen is the most consumed energy medium in converter steelmaking. It is mainly used to oxidize impurity elements in molten iron, reduce carbon content, and provide a large amount of heat. The consumption of oxygen can be expressed as:

[0012]

[0013] Where m O2 is the oxygen consumption mass of the converter, kg; m iron_O2 is the mass of oxygen consumed by element oxidation in molten iron, kg; m scrap_O2 is the mass of oxygen consumed by oxidation of elements in scrap steel, kg; m lin_O2 The mass of oxygen consumed by lining oxidation, kg; m dust_O2 The mass of oxygen consumed by the oxidation of iron in smoke, kg; m BOFG_O2 is the mass of free oxygen in the furnace gas, kg; m ore_O2 is the mass of oxygen replaced in the ore, kg; k oxygen is the oxygen conversion coefficient to standard coal, kgce / m 3 .

[0014] The principle of element oxidation in molten iron and scrap steel is as follows:

[0015] The reaction formula is:

[0016] C+O2→CO2

[0017] C+1 / 2O2→CO

[0018] Si+O2→SiO2

[0019] Mn+1 / 2O2→MnO

[0020] 2P+5 / 2O2→P2O5

[0021] 2Fe+O2→2FeO

[0022] 2Fe+3 / 2O2→Fe2O3

[0023] Therefore, the element oxidation consumption in molten iron and scrap steel can be calculated by calculating the oxidation mass of each element and the oxygen consumption according to the chemical reaction balance:

[0024]

[0025] Among them, μ i m is the oxygen consumption coefficient of the ith element oxidation; iron_i is the mass of the oxidized element i in the molten iron, kg; m scrap_i is the mass of the oxidized element i in the scrap steel, kg.

[0026] The converter lining will also be partially oxidized during the oxygen blowing process, mainly the carbon element in the lining. Part of the C is oxidized to CO, and part of the C is oxidized to CO2. Therefore, the amount of oxygen consumed by the oxidation lining is calculated as follows:

[0027]

[0028] Where λ is the proportion of C oxidized to CO; m lin_C is the mass of oxidized C element in the furnace lining, kg.

[0029] The Fe element in the converter dust will be oxidized by oxygen into FeO and Fe2O3, which will also cause a certain amount of oxygen consumption. The calculation formula for oxygen consumption is obtained based on the oxidation mechanism of the Fe element:

[0030]

[0031] Among them, m dust_FeO is the mass of FeO in converter dust, kg; m dust_Fe2O3 is the mass of Fe2O3 in converter dust, kg.

[0032] (2) Nitrogen consumption

[0033] In converter steelmaking, nitrogen can be used to seal oxygen gun holes, feed pipes, and converter ports to prevent converter flue gas from overflowing and converter flames from leaking. Blowing nitrogen into the converter in the early and middle stages of blowing can remove impurities in the molten steel. In addition, nitrogen can also be used for slag splashing and furnace protection. Therefore, for the consumption of nitrogen in the converter steelmaking process, it is necessary to consider the total consumption of nitrogen at different pressures in multiple scenarios. The calculation method is as follows:

[0034]

[0035] Among them, E nitrogen is the total nitrogen consumption during converter blowing process; k nitrogen is the standard coal conversion coefficient of nitrogen, kgce / m 3 ;Q i-nitrogen is the flow rate of nitrogen used in different scenarios, m 3 / min;t i-nitrogen The time when nitrogen is used in different scenarios, min;

[0036] (3) Argon consumption

[0037] In converter steelmaking, argon is usually introduced into the molten steel in the later stage of blowing. This is because argon, as an inert gas, does not participate in metallurgical reactions, and can discharge gas impurities such as N2 and H2 while stirring the molten steel, playing the role of "gas washing". Therefore, the consumption model of argon can be expressed as:

[0038] Eargon =k argon Q argon ·t argon

[0039] Among them, E argon is the total argon consumption in the converter blowing process, kgce; k argon is the argon conversion coefficient to standard coal, kgce / m 3 ;Q argon is the flow rate of argon gas, which can be regarded as a constant value, m 3 / min;t argon is the time of passing argon gas, min.

[0040] (4) Compressed air consumption

[0041] During steelmaking, compressed air drives valves to control the opening and closing of each gas supply pipeline to control the delivery of gaseous media. If the compressed air flow rate is considered stable each time the pneumatic valve is used, the total consumption of compressed air in a smelting cycle can be calculated by accumulating the amount of compressed air consumed by each opening and closing of the pneumatic valve using the following formula:

[0042]

[0043] Among them, E air is the total compressed air consumption during converter blowing, kgce; k air is the standard coal coefficient of compressed air, kgce / m 3 ;Q i-air is the flow rate of compressed air when the pneumatic valve is used for the i-th time, m 3 / s;t i-air is the time for compressed air to be introduced when the pneumatic valve is used for the i-th time, s.

[0044] (5) Power consumption

[0045] This invention mainly studies the energy consumption of the converter in a smelting cycle, so the equipment considered mainly includes the converter, oxygen lance, etc. Assuming that the power of each device during operation is approximately constant, the consumption of electric energy can be calculated by statistically analyzing the power consumption of each device separately, as follows:

[0046]

[0047] Among them, E electricity is the total power consumption of converter blowing process, kgce; k electricity is the coefficient of electricity conversion to standard coal, kgce / (kW·h); P i is the power of each power-consuming device of the converter, kW; t i is the operating time of each power-consuming equipment of the converter, min.

[0048] (6) Water consumption

[0049] Water is also a huge energy medium in the metallurgical industry. In the converter process, water is mainly used to cool the converter, oxygen lance and other equipment. This part of water usually belongs to the factory cooling water circulation system and is recycled in the steelmaking process. Cooling water is also used in the converter flue gas recovery to recover the heat carried by the flue gas and convert it into steam for external use. In addition, a part of the water is used for dust removal of converter flue gas. The present invention mainly considers the energy consumption used in the converter within a smelting cycle and the secondary energy generated by the blowing process, so the consumption of water is not considered here for the time being.

[0050] (7) Converter gas recovery

[0051] During the oxygen blowing process of converter steelmaking, various elements in the molten iron are oxidized to produce gases, also known as converter gas. Among them, carbon reacts with oxygen to produce CO, which is the main component of the gas and has a high recovery value. During the recovery process, when the gas enters the purification and recovery system, it will inhale some air, causing some CO to burn when it encounters oxygen and release heat. The gas at this time is called flue gas. After passing through the purification and recovery system, converter gas is obtained. The main component of converter gas is CO. In addition, it also includes CO2, N2 and a small amount of O2. When the CO content of converter gas is greater than 25% and the O2 content is less than 1%, the converter gas is recovered, otherwise it is passed into the chimney for ignition and dispersion.

[0052] (8) Steam recovery

[0053] During the converter steelmaking process, oxygen reacts chemically with impurity elements such as carbon, silicon, manganese, and phosphorus in the molten iron to generate a large amount of high-temperature furnace gas. The temperature of these furnace gases is usually around 1400°C and contains a large amount of heat energy. These high-temperature furnace gases enter the waste heat recovery system, where the heat exchanger transfers the heat of the furnace gas to water, thereby heating the water to generate steam. In this process, the water absorbs the heat in the exhaust gas and gradually evaporates, and the generated steam can be collected and used.

[0054] Preferably, step 3 establishes a data model of oxygen consumption, converter gas recovery, steam recovery and molten steel production in converter steelmaking, and the specific process is as follows:

[0055] (1) Collect data on parameters such as molten iron quality, scrap steel quality, molten iron temperature and composition, as well as oxygen consumption, converter gas recovery, steam recovery, and molten steel output during the converter steelmaking process.

[0056] (2) Preprocess the collected data, remove outliers and missing values, and normalize the data.

[0057] (3) The MIC method was used to screen the features of oxygen consumption, converter gas recovery, steam recovery and molten steel output, and the WOA-XGBoost method was used to establish data models respectively, and finally experimental verification was carried out.

[0058] Preferably, step 4 combines the mechanism and data-driven model to establish a converter energy consumption per ton of steel and carbon emission per ton of steel. The specific process is as follows:

[0059] The converter energy consumption is usually expressed as the comprehensive energy consumption of the consumption of each energy medium and the secondary energy generated. The specific calculation is the amount of standard coal converted from the production and consumption of various energy media, and the unit product energy consumption is used as an indicator to measure the converter energy consumption:

[0060] E BOF =(E oxygen +E nitrogen +E argon +E air +E electricity +E water -E LDG -E steam ) / m steel

[0061] Among them, E BOF is the converter energy consumption per ton of steel, kgce / t; m steel is the mass of a furnace of molten steel, t.

[0062] Converter carbon emissions usually study the carbon dioxide emissions within a smelting cycle (one furnace) and are evaluated based on the carbon dioxide generated by producing 1 ton of molten steel. Carbon emissions can be divided into fuel carbon, material carbon and energy carbon:

[0063] CE total =CE fue +CE mat +CE ene

[0064] The carbon emission model of converter per ton of steel can be expressed as: CE BOF =CE total / m steel .

[0065] Among them, CE total is the total carbon emissions, tCO2; CE fue is the carbon emission from fossil fuel combustion, tCO2; CE mat Material consumption and generated CO2, tCO2; CE ene Indirect emissions caused by consumption and generation of energy media, tCO2.

[0066] Converter smelting mainly provides heat through oxidation reaction rather than burning fossil fuels. Therefore, direct emissions from fossil fuel combustion, i.e. CE fue =0.

[0067] Material carbon mainly includes CE of consumed materials mat-con and output product emissions CE mat-pro :

[0068] CE mat =CE mat-con -CE mat-pro

[0069]

[0070] Among them, P i is the net consumption of the ith material of the converter (including raw materials such as molten iron and scrap steel and fluxes such as limestone and dolomite), t; EF i is the CO2 emission factor of the i-th material, tCO2 / t. AD j is the output of the jth output product (including molten steel, converter slag, etc.), t; EF j is the CO2 emission factor of the jth output product, tCO2 / t.

[0071] Energy carbon in converter smelting is also divided into energy consumption and emission CE ene-con Mainly includes oxygen, nitrogen, argon, compressed air and other energy sources, as well as recycled energy emission CE ene-rec , including converter gas and steam, namely:

[0072] CE ene =CE ene-con -CE ene-rec

[0073] CE ene-con =k kgce-c ·(E nitrogen +E argon +E air +E oxygen +E electricity )·10 -3

[0074] CE ene-rec =k kgce-c ·(E LDG +E steam )·10 -3

[0075] Among them, k kgce-c is the carbon emission coefficient of standard coal, tCO2 / tce. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 Mechanism-Data-driven Modeling Framework Diagram

[0077] Figure 2 Relationship diagram of converter material and energy flow

[0078] Figure 3 Data modeling framework diagram

[0079] Figure 4 WOA-XGBoost modeling effect diagram

[0080] Figure 5 XGBoost modeling effect diagram

[0081] Figure 6 RF modeling effect diagram

[0082] Figure 7 SVR modeling effect diagram DETAILED DESCRIPTION

[0083] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely explain the technical solutions in the embodiments of the present invention. The described examples are only part of the examples of the present invention, not all of the examples. Without departing from the above technical ideas of the present invention, various replacements and changes are made according to the common technical knowledge and customary means in the field, which should be included in the protection scope of the present invention.

[0084] In this embodiment, 902 sets of converter steelmaking data were collected in a steel plant in Chongqing, including the consumption of molten iron, scrap steel, oxygen, etc., molten iron entering furnace temperature, oxygen blowing intensity, converter gas and steam recovery, molten steel output, etc. Part of the data is shown in Table 1.

[0085] Table 1 Converter parameters and energy consumption data

[0086]

[0087] The collected data is preprocessed and feature selected, and finally the model is built using data-driven methods and frameworks. Figure 3 The outliers and missing values ​​in the data set were removed and then normalized. The MIC method was used to screen out the features that have a greater impact on oxygen consumption, converter gas recovery, steam recovery and molten steel production, as shown in Table 2.

[0088] Table 2 MIC feature selection weight ratio

[0089]

[0090] In order to further verify the effectiveness and superiority of the adopted data-driven modeling method and data model, WOA-XGBoost, XGBoost, RF, and SVR methods were used to establish oxygen consumption, converter gas recovery, steam recovery, and molten steel production prediction models. The model training results are shown in Figure 2. Figure 4 , Figure 5 , Figure 6 and Figure 7 In order to intuitively show the performance of different prediction models, the evaluation index results of each model are shown in Table 3. Obviously, the model established by the WOA-XGBoost method has better indicators than several other prediction models.

[0091] Table 3 Evaluation index results of different models

[0092]

[0093] Combining the mechanism model with the four data models established by the WOA-XGBoost method, the converter energy consumption per ton of steel and carbon emission per ton of steel models are finally obtained, and the two models are experimentally verified, and the obtained model evaluation indicators are shown in Table 4. The experimental results show that the converter energy consumption per ton of steel and carbon emission per ton of steel prediction model indicators established by the present invention perform well, the model has high accuracy, and can effectively realize the prediction of energy consumption and carbon emissions in the converter steelmaking production process.

[0094] Table 4 Evaluation index results of the model of energy consumption per ton of steel and carbon emission per ton of steel

[0095]

Claims

1. A mechanism-data driven converter energy consumption and carbon emission prediction modeling method, characterized in that: The following steps are involved: Step 1: Analyze the relationship between material and energy flow in the converter steelmaking process to obtain the material consumption and energy consumption of the converter steelmaking process. Step 2: Use theoretical principles to analyze the production and consumption of energy media such as oxygen, nitrogen, and electricity in converter steelmaking, and establish mechanism models for nitrogen, argon, compressed air, and electricity. Step 3: Use a data-driven approach to establish a data model for oxygen consumption, converter gas recovery, steam recovery, and molten steel production in converter steelmaking. Step 4: Combine the mechanism and data model of each energy medium and molten steel output to establish a model of converter energy consumption per ton of steel and carbon emissions per ton of steel to achieve prediction.

2. The lightweight design method for the feed system structure of an internal gear honing machine tool according to claim 1, characterized in that In step 3, a data model for oxygen consumption, converter gas recovery, steam recovery, and molten steel production in the converter is established, following the following process: (1) Collect data on parameters such as molten iron quality, scrap steel quality, molten iron temperature and composition, as well as oxygen consumption, converter gas recovery, steam recovery, and molten steel output during the converter steelmaking process. (2) Preprocess the collected data, remove outliers and missing values, and normalize the data. (3) The MIC method was used to screen the features of oxygen consumption, converter gas recovery, steam recovery and molten steel output, and the WOA-XGBoost method was used to establish data models respectively, and finally experimental verification was carried out.

Images