An optimization method, equipment and storage medium for steelmaking process in the iron and steel industry

Abstract

The present invention relates to the technical field of low-carbon process optimization in the iron and steel industry. The present invention provides a method, equipment and storage medium for optimizing the steelmaking process in the iron and steel industry, including the following steps: establishing a parameterized process of steelmaking technology, defining material, energy consumption, carbon emission coefficient and cost parameters, and constructing a parameterized model of the steelmaking process; based on the parameterized model of the steelmaking process, defining the state space and action space of the steelmaking process, setting a multi-objective reward function and a corresponding constraint violation penalty term to obtain a reinforcement learning model; training the reinforcement learning model based on the Q-learning algorithm, updating the Q value through the interaction loop of state-action-reward, and dynamically adjusting the selection strategy; performing simulation verification on the selection strategy output by the reinforcement learning model. Thus, the problems of poor dynamic adaptability, single optimization objective and unstable technology transition existing in the existing research methods for low-carbon paths in the iron and steel industry are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of low-carbon process optimization in the steel industry, and more specifically, to a method, device, and storage medium for optimizing the steelmaking process in the steel industry. Background Art

[0002] As a core component of the global industrial system, the steel industry has long faced challenges of high energy consumption and high carbon emission intensity.

[0003] Currently, the research methods for the low-carbon path in the steel industry are mainly divided into three categories:

[0004] 1) Bottom-up models (such as MESSAGE and AIM / Enduse models) optimize technical paths based on the principle of cost minimization, but they lack adaptability to dynamic market changes.

[0005] 2) Material flow analysis methods reveal emission reduction potential by tracking material flows throughout the life cycle, but they lack the ability to dynamically predict technology diffusion.

[0006] 3) Top-down models (such as computable general equilibrium models) focus on the interaction of the macroeconomic system, but they are difficult to handle multi-objective optimization problems.

[0007] The limitations of the existing technologies are mainly reflected in:

[0008] Poor dynamic adaptability: Traditional models rely on preset parameters and explicit constraint equations and are difficult to cope with dynamic environments such as policy adjustments and market fluctuations.

[0009] Single optimization objective: Most models only focus on single-objective optimization of cost or carbon emissions and lack the ability to co-optimize multiple objectives (such as cost, carbon emission reduction, and market share).

[0010] Unstable technology transition: The system stability of technology path switching is not considered, which may lead to production fluctuations or resource waste. Summary of the Invention

[0011] The purpose of the present invention is to provide a method, device, and storage medium for optimizing the steelmaking process in the steel industry, aiming to solve the problems of poor dynamic adaptability, single optimization objective, and unstable technology transition existing in the existing research methods for the low-carbon path in the steel industry.

[0012] Embodiments of the present invention are achieved through the following technical solutions:

[0013] A method for optimizing the steelmaking process in the steel industry includes the following steps:

[0014] Build a parametric process that includes three steelmaking technologies: blast furnace-converter steelmaking, scrap steel steelmaking, and direct reduced iron steelmaking. Define material, energy consumption, carbon emission coefficients, and cost parameters, and construct a parametric model for the steelmaking process;

[0015] Based on the parametric model of the steelmaking process, define the state space and action space of the steelmaking process, set a multi-objective reward function and the corresponding constraint violation penalty term to obtain a reinforcement learning model; among them, the state space includes market share, total cost, carbon emissions, output, and technical stability; the action space is the selection strategy of the three steelmaking technologies;

[0016] Train the reinforcement learning model based on the Q-learning algorithm, update the Q value through the interaction cycle of state-action-reward, and dynamically adjust the selection strategy;

[0017] Perform simulation verification on the selection strategy output by the reinforcement learning model. If the simulation result does not meet the expected optimization goal, adjust the corresponding parameters of the reinforcement learning model through the offline gradient descent method and retrain the reinforcement learning model until the optimization goal is met.

[0018] Optionally, introduce a market price fluctuation factor, dynamically adjust the unit price of energy and materials, optimize the equipment efficiency parameters based on historical data, and correct the energy and material consumption coefficients of each process link.

[0019] Optionally, the specific process of introducing a market price fluctuation factor, dynamically adjusting the unit price of energy and materials, optimizing the equipment efficiency parameters based on historical data, and correcting the energy and material consumption coefficients of each process link is as follows:

[0020] Build a market database, collect historical price data and supply-demand relationship data of energy and materials;

[0021] Use time series analysis or machine learning algorithms to build a price fluctuation prediction model to quantify the fluctuation range and frequency of the unit price of energy and materials;

[0022] According to the price fluctuation prediction model, generate energy unit price fluctuation factors and material unit price fluctuation factors, and update the fluctuation factors through the sliding window method to ensure that the price parameters are synchronized with the real-time market dynamics;

[0023] Construct an equipment efficiency evaluation index system, including energy conversion efficiency, material utilization rate, and equipment failure rate, and use the particle swarm optimization algorithm or genetic algorithm to optimize the equipment efficiency parameters with historical operation data as input;

[0024] Based on the optimized equipment efficiency parameters, recalculate the energy consumption coefficient and material consumption coefficient for each process step, use Bayesian inference method to quantify the uncertainty of the corrected parameters, construct a confidence interval, and verify the corrected parameter model through a process simulation platform to ensure that the prediction error remains ≤ 5% within the market fluctuation range of ±10%;

[0025] Set the parameter update period, trigger an immediate update when the market fluctuation exceeds the threshold, and establish a parameter version management system to record the fluctuation factor, equipment efficiency parameters, and correction basis for each adjustment.

[0026] Optionally, the specific process of obtaining the reinforcement learning model based on the steelmaking process parameter model by defining the state space and action space of the steelmaking process, setting a multi-objective reward function and the corresponding constraint violation penalty term is as follows:

[0027] Take the market share, total cost, carbon emissions, output, and technical stability as state variables, and construct a five-tuple state vector as shown in the following formula (1):

[0028]

[0029] where, represents the market share of technology at time ; represents the total steelmaking cost at time ; represents the carbon emissions at time ; represents the steel output at time ; represents the technical stability, that is, the absolute value of the change in market share between adjacent time steps;

[0030] Take the market share allocation ratios of the three steelmaking technologies as actions, as shown in the following formula (2):

[0031]

[0032] where, represents the action that the agent can choose in the state; , and represent the market share allocation ratios corresponding to the three steelmaking technologies respectively; and represent the upper and lower limits of the market share respectively;

[0033] The reward function is as shown in the following formula (3):

[0034]

[0035] and and are expressed as shown in the following equations (4), (5) and (6):

[0036]

[0037]

[0038]

[0039] wherein and and and respectively represent the weight coefficients of the corresponding state vectors; and and respectively represent the penalty coefficients of the corresponding penalty terms; represents the penalty term for carbon emission constraint; represents the penalty term for market share constraint; represents the penalty term for steel production constraint.

[0040] Optionally, the reinforcement learning model is trained based on the Q - learning algorithm, and the Q - value is updated through the interaction loop of state - action - reward, and the specific process of dynamically adjusting the selection strategy is as follows:

[0041] Initialize the Q - values of all state - action pairs to random values or zero;

[0042] Based on the current state and the ε - greedy policy, select an action , where the ε - greedy policy randomly explores the action space with probability ε and selects the action with the largest current Q - value with probability 1 - ε;

[0043] Execute the selected action , trigger the environmental state to transfer from to , and calculate the immediate reward according to the reward function, and the reward function includes a market share reward term, a cost penalty term, a carbon emission penalty term, a technical stability penalty term and a constraint violation penalty term;

[0044] Update the Q - value using Equation (7), and Equation (7) is shown as follows:

[0045]

[0046] wherein represents the Q - value; represents the Q - value learning rate; Represents the discount factor, which is used to control the impact of future rewards; Is the maximum Q-value in the next state, representing the return after the agent selects the optimal action;

[0047] The behavioral policy is iteratively updated through the following formula (8), as shown in formula (8):

[0048]

[0049] Among them, Represents the steelmaking process selected by the agent in state Under.

[0050] Optionally, gradually increase the penalty coefficient in the later stage of training , And , ensuring that the technical path selected by the agent meets the carbon emission cap, market share constraint, and production volume constraint.

[0051] Optionally, the specific process of simulating and verifying the selection strategy output by the reinforcement learning model, and if the simulation result does not reach the expected optimization goal, adjusting the corresponding parameters of the reinforcement learning model through the offline gradient descent method and retraining the reinforcement learning model is as follows:

[0052] Collect historical data, including the market share, total cost, carbon emissions, and market share change of each steelmaking process;

[0053] Construct the objective function, as shown in the following formula (9):

[0054]

[0055] Among them, the reward function Includes the market share reward term, cost penalty term, carbon emission penalty term, and technical stability penalty term;

[0056] Calculate the gradient of the objective function with respect to each weight coefficient , , And , as shown in the following formula (10):

[0057]

[0058] Among them, , , And Respectively represent taking the partial derivative with respect to the weight coefficients , , And The weight coefficients are updated using the following formula (11):

[0059]

[0060] Among them, represents the weight learning rate; , , and respectively represent the weight coefficients corresponding to market share, cost, carbon emissions, and technological stability in the reward function at the -th iteration; , , and respectively represent the corresponding updated weight coefficients; represents the discount factor at the -th time; represents the time step of historical data.

[0061] Optionally, the specific process of establishing the accounting models for equipment depreciation cost, maintenance cost, environmental protection cost, energy cost, material cost, and financial subsidy, constructing the carbon emission coefficient matrix for each process link, and calculating the total carbon emissions in combination with energy consumption data is as follows:

[0062] Let the total cost include equipment depreciation cost, equipment maintenance cost, environmental protection equipment input cost, financial subsidy, and energy cost and material cost required in different steelmaking processes, as shown in the following formula (12):

[0063]

[0064] Among them, represents the total steelmaking cost; represents the equipment depreciation cost; represents the equipment maintenance cost; represents the environmental protection equipment input cost; represents the energy cost; represents the material cost; represents the financial subsidy;

[0065] During the steelmaking process, the carbon emission expression is as shown in the following formula (13):

[0066]

[0067] Among them, represents the total carbon emissions; represents the steelmaking process; represents the steelmaking equipment; represents the type of energy; represents the carbon emission coefficient.

[0068] Based on the same inventive concept, the present invention also provides an electronic device, including a memory and a processor. The memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the above-mentioned steelmaking process optimization method in the iron and steel industry.

[0069] Based on the same inventive concept, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it realizes the above-mentioned photovoltaic power generation energy storage grid connection optimization system and method.

[0070] The technical solutions of the embodiments of the present invention have at least the following advantages and beneficial effects:

[0071] Enhanced dynamic adaptability: By constructing a coupled system of a parameterized process model and a reinforcement learning model, the real-time response ability to dynamic environments such as market fluctuations and policy adjustments is realized; through the cycle iteration of state-action-reward, the Q-learning algorithm can autonomously update the technology selection strategy, breaking through the limitations of traditional models that rely on preset parameters.

[0072] Multi-objective collaborative optimization: By introducing a five-dimensional state space including market share, cost, carbon emissions, output, and technical stability, combined with a multi-objective reward function, the collaborative optimization of economic benefits and environmental benefits is realized; compared with traditional single-objective optimization models, it can simultaneously pursue cost minimization and market competitiveness improvement under the carbon emission reduction constraint.

[0073] Guarantee of technical transition stability: Incorporating technical stability into the state space, effectively suppressing the drastic switching of technical paths through a constraint violation penalty mechanism; combined with the parameter calibration of the offline gradient descent method, it can ensure the stable operation of the production system during the technical iteration process, avoiding resource waste and production capacity fluctuations.

[0074] Improved model generalization ability: The parameterized process model uniformly models the three mainstream steelmaking processes of blast furnace-converter, scrap steel, and direct reduced iron, and the reinforcement learning framework has the ability to optimize strategies across technical paths; through simulation verification and parameter fine-tuning mechanisms, it can quickly adapt to the energy structure, policy environment, and market demands in different regions.

[0075] Intelligent decision-making support: Breaking through the static analysis mode of traditional models, it can dynamically generate the optimal technology combination plan; by quantifying the marginal emission reduction costs and market returns of each technical path, it provides a scientific decision-making basis for steel enterprises to formulate medium- and long-term low-carbon transformation plans.

[0076] Prediction of low-carbon technology diffusion: The reinforcement learning model can simulate the dynamic evolution process of technology selection strategies, reveal the diffusion laws of different technical paths in market competition, and provide a theoretical support for promoting the large-scale application of low-carbon technologies. Description of the Drawings

[0077] Figure 1 It is a schematic flow chart of the steelmaking process optimization method in the embodiment of the present invention for the iron and steel industry;

[0078] Figure 2 It is a schematic flow chart of the blast furnace-converter steelmaking process in the steelmaking process optimization method in the embodiment of the present invention for the iron and steel industry;

[0079] Figure 3 It is a schematic flow chart of the scrap steelmaking process in the steelmaking process optimization method in the embodiment of the present invention for the iron and steel industry;

[0080] Figure 4 It is a schematic flow chart of the direct reduced iron steelmaking process in the steelmaking process optimization method in the embodiment of the present invention for the iron and steel industry. Detailed implementation manners

[0081] The following are the detailed implementation manners in combination with the accompanying drawings.

[0082] Refer to Figure 1 , a steelmaking process optimization method for the iron and steel industry, comprising the following steps:

[0083] Step 1: Establish a parametric process including three steelmaking technologies of blast furnace-converter steelmaking, scrap steelmaking, and direct reduced iron steelmaking, define material, energy consumption, carbon emission coefficient, and cost parameters, and construct a parametric model of the steelmaking process.

[0084] In some embodiments, the specific process of constructing the parametric model of the steelmaking process is as follows:

[0085] Refer to Figure 2 , parametric of blast furnace-converter steelmaking:

[0086] Define the consumption of coking coal, electricity, and coal. Define the consumption of coking coal. It can be achieved by installing high-precision material metering equipment, such as belt scales, in the coking section of blast furnace-converter steelmaking to monitor the quality of coking coal entering the coke oven in real time and record the consumption per unit time; the electricity consumption can be accurately measured by installing watt-hour meters at the electricity input ports of steelmaking equipment, and classified and counted according to different equipment or equipment groups; the definition method of coal consumption is similar to that of coking coal. For coal used in other processes (such as heating), its usage is recorded through appropriate metering devices.

[0087] Establish a material balance equation for the conversion of coking coal into coke. Based on the principles of chemical reactions, analyze the compositional changes of coking coal during the coking process, and determine the conversion ratios of various elements (such as carbon, hydrogen, oxygen, etc.) in coking coal during the conversion into coke; collect relevant data during the coking process, such as the initial mass and compositional analysis data of coking coal, as well as the mass and compositional data of the produced coke. Through the law of conservation of mass, establish the mathematical relationship between the mass of coking coal and the mass of coke to form a material balance equation. For example, assuming that most of the carbon element in coking coal is converted into carbon in coke, a part of the equation can be constructed based on the mass conservation of carbon element.

[0088] Establish consumption models for coke, iron ore, electricity, and coal, and correlate the ratio relationship between the output of sinter and the input of raw materials. For the establishment of consumption models for coke, iron ore, electricity, and coal, long-term production data recording and analysis can be carried out. During the production process, simultaneously record the input amounts of these raw materials and energy in each batch of production, as well as the corresponding product output situations (such as the output of hot metal, etc.); use statistical analysis methods, such as regression analysis, to find out the internal relationships between these consumptions and production indicators, and construct consumption models. For example, analyze the average consumption amounts and their fluctuation ranges of coke, iron ore, electricity, and coal required for each production of a certain mass of hot metal; for the ratio relationship between the output of sinter and the input of raw materials, in the sintering process, record the input amounts of iron ore, fluxes (such as limestone, etc.), fuels (such as coke powder, etc.), and the mass of the produced sinter. Through the analysis of material balance and actual production data, determine their ratio relationship, such as how much mass of sinter can be produced for every input of a certain mass of iron ore and other raw materials.

[0089] Based on the consumption amounts of coke, sinter, electricity, and coal, construct a mass conservation equation for the output of hot metal. Combining the consumption models and relevant data established previously, according to the principle of mass conservation, analyze how the inputs of coke, sinter, electricity, coal, etc. are converted into the output of hot metal during the blast furnace ironmaking process; considering some possible material losses during the ironmaking process (such as part of the iron element taken away by slag, etc.), through detailed research and data statistics of the production process, quantify and correct these losses. Thus, establish a mass conservation equation with the consumption amounts of coke, sinter, electricity, and coal as independent variables and the output amount of hot metal as the dependent variable.

[0090] Quantify the consumption parameters of hot metal, scrap steel, electricity, coal, natural gas, and coke, and establish a linear relationship between crude steel output and material input. In the converter steelmaking process, through the metering devices installed on various equipment and material conveying pipelines, accurately measure the actual consumption of hot metal, scrap steel, electricity, coal, natural gas, coke, etc.; keep a detailed record of each steelmaking production, including the quantity of various materials input, energy consumption, and the quality of the finally produced crude steel; use mathematical methods such as linear regression to analyze and process these data, find out the linear relationship between the crude steel output and the input quantity of materials such as hot metal, scrap steel, electricity, coal, natural gas, coke, etc., and establish a corresponding mathematical model. Through the definition of each parameter and the establishment of relevant equations and models, it is possible to deeply understand the flow and transformation relationship of materials and energy in the blast furnace-converter steelmaking process, providing a solid theoretical basis and data support for subsequent process optimization. For example, accurate material balance equations and mass conservation equations can help technicians clearly see the changes of materials in each link, so as to make targeted improvements.

[0091] Refer to Figure 3 , parameterize scrap steel steelmaking: Based on the hot metal flow rate produced by blast furnace-converter steelmaking, establish a ratio model of hot metal to scrap steel. Collect the hot metal flow rate data produced by blast furnace-converter steelmaking under different production conditions, such as during different production shifts and different equipment operating states, and record the output of hot metal. These data can be obtained in real time through the flow monitoring equipment installed on the blast furnace-converter steelmaking production line. Analyze and process the collected hot metal flow rate data, considering the influence of factors such as the quality and specifications of scrap steel on the ratio of hot metal to scrap steel, and establish a mathematical model to describe the reasonable ratio relationship between hot metal and scrap steel. For example, a linear regression model can be used. By fitting the historical production data, determine the linear relationship between the hot metal flow rate and the scrap steel addition amount, that is, assume the hot metal flow rate is , the scrap steel addition amount is , and we can get in the form of, where and are coefficients obtained through data fitting. In actual production, according to the hot metal flow rate produced by blast furnace-converter steelmaking monitored in real time, use the established ratio model to calculate the corresponding scrap steel addition amount to ensure the reasonable input of materials in the scrap steel steelmaking process.

[0092] Define the energy consumption coefficients of electricity, coal, and natural gas, as well as the proportional relationship between crude steel output and total material input. For the determination of the energy consumption coefficient of electricity, install an electricity monitoring device on the scrap steel melting equipment to record the power consumption at different stages (such as scrap steel preheating, melting, etc.) during the scrap steel melting process in real time. Through statistical analysis of the power consumption data of a large number of production batches and in combination with the amount of crude steel produced, calculate the power consumption per unit of crude steel production, thereby determining the energy consumption coefficient of electricity. For the energy consumption coefficient of coal, record the amount of coal used during the scrap steel melting process, including information such as the type and quality of coal. Similarly, by analyzing the coal usage and crude steel output of multiple production batches, calculate the amount of coal consumed per unit of crude steel production to obtain the energy consumption coefficient of coal. For the energy consumption coefficient of natural gas, install a natural gas flow monitoring device to monitor the amount of natural gas used during the scrap steel melting process. According to a method similar to that of electricity and coal, calculate the amount of natural gas consumed per unit of crude steel production to determine the energy consumption coefficient of natural gas. Clearly define that the total material input includes hot metal, scrap steel, and other possible auxiliary materials. During the production process, accurately record the quality of hot metal, scrap steel, and auxiliary materials input each time, as well as the quality of the finally produced crude steel. Through the analysis of multiple sets of production data, establish the proportional relationship between the crude steel output and the total material input. For example, it can be obtained in the form of where, k represents the proportional coefficient obtained through data statistics and analysis, and this coefficient reflects the material conversion efficiency during the scrap steel melting process; represents the crude steel output, represents the total material input. By establishing a ratio model of hot metal and scrap steel, the amount of scrap steel added can be reasonably adjusted according to the hot metal flow rate produced by blast furnace-converter steelmaking, avoiding waste or shortage of materials, thereby improving the material utilization efficiency and reducing production costs. For example, when the hot metal flow rate is low, appropriately increase the proportion of scrap steel to ensure the continuity and stability of production, while reducing the excessive dependence on hot metal.

[0093] Refer to Figure 4 , for the parametricization of direct reduced iron steelmaking: establish the relationship between the output of iron ore pellets and the consumption of iron ore, electricity, and coal. Collect a large amount of direct reduced iron steelmaking production data, covering different production conditions, equipment operation parameters, etc. Sort out and analyze the data to determine the relationship between the input and output amounts of iron ore during the conversion of iron ore into iron ore pellets. At the same time, monitor the real-time consumption of electricity and coal during this process, and establish a mathematical model based on production data. For example, methods such as multiple linear regression can be used to determine the functional relationship between the output amount of iron ore pellets and the input amounts of iron ore, electricity, and coal, that is: where, represents the output amount of iron ore pellets, Represents the input amount of iron ore, Represents the power consumption, Represents the coal consumption.

[0094] Quantify the consumption of natural gas, electricity, and hydrogen, and establish the material balance between the output of direct reduced iron and the raw materials. During the production process of direct reduced iron, install high-precision metering equipment to accurately measure the real-time consumption of natural gas, electricity, and hydrogen. Based on the chemical reaction principle and the law of conservation of matter, establish the material balance equation between the output of direct reduced iron and the input of these raw materials. For example, assume that the main chemical reaction in the production of direct reduced iron is: xFe 2 O 3 +yCH 4 +zH 2 +mO 2 →nFe +pCO 2 +qH 2 O (The stoichiometric coefficients here x , y , z , m , n , p , q are determined according to the actual reaction). According to this reaction formula and the actual consumption of natural gas, hydrogen, etc., a material balance equation can be established, such as: , where represents the output of direct reduced iron, represents the consumption of natural gas, represents the power consumption, represents the consumption of hydrogen.

[0095] Define the ratio parameter of direct reduced iron to scrap steel, as well as the energy consumption models of natural gas, electricity, and coal, and establish a linear equation for crude steel output. During the actual production process, set different ratio schemes of direct reduced iron to scrap steel, accurately control the input amounts of the two through an automated control system, and record the corresponding crude steel output and the consumption data of natural gas, electricity, and coal. Define the ratio parameter of direct reduced iron to scrap steel, where is the mass of direct reduced iron, is the mass of scrap steel. Use methods such as grey relational analysis and principal component analysis to find out the key influencing factors between the energy consumption of natural gas, electricity, and coal and the crude steel output, as well as the ratio of direct reduced iron to scrap steel, and construct the energy consumption models. For example, the natural gas energy consumption model is , where is the crude steel output, , , are model parameters; similarly, energy consumption models for electricity and coal are constructed. Finally, through the analysis and fitting of a large amount of production data, a linear equation is established for the crude steel output in relation to factors such as the direct reduced iron input, scrap steel input, natural gas consumption, electricity consumption, and coal consumption. The precise consumption relationship and material balance model enable producers to anticipate in advance the demand for raw materials and energy during the production process, timely adjust the input volume, and avoid production interruptions or inefficiencies caused by raw material shortages or energy waste, greatly enhancing the controllability and stability of the production process.

[0096] Cost and carbon emission parameterization: Establish accounting models for equipment depreciation cost, maintenance cost, environmental protection cost, energy cost, material cost, and financial subsidies; construct a carbon emission coefficient matrix for each process link, and calculate the total carbon emissions in combination with energy consumption data.

[0097] In some embodiments, the specific process of establishing accounting models for equipment depreciation cost, maintenance cost, environmental protection cost, energy cost, material cost, and financial subsidies, constructing a carbon emission coefficient matrix for each process link, and calculating the total carbon emissions in combination with energy consumption data is as follows:

[0098] Let the total cost include equipment depreciation cost, equipment maintenance cost, environmental protection equipment investment cost, financial subsidies, and energy cost and material cost required in different steelmaking processes, as shown in the following formula (12):

[0099]

[0100] Among them, represents the total steelmaking cost; represents the equipment depreciation cost; represents the equipment maintenance cost; represents the environmental protection equipment investment cost; represents the energy cost; represents the material cost; represents the financial subsidy;

[0101] During the steelmaking process, the carbon emission expression is as shown in the following formula (13):

[0102]

[0103] Among them, represents the total carbon emissions; represents the steelmaking process; represents the steelmaking equipment; represents the energy type; represents the carbon emission coefficient.

[0104] Introduce the market price fluctuation factor, dynamically adjust the unit price of energy and materials, optimize the equipment efficiency parameters based on historical data, and correct the energy and material consumption coefficients of each process link.

[0105] In some embodiments, the specific process of introducing the market price fluctuation factor, dynamically adjusting the unit price of energy and materials, optimizing the equipment efficiency parameters based on historical data, and correcting the energy and material consumption coefficients of each process link is as follows:

[0106] Establish a market database to collect historical price data and supply-demand relationship data of energy and materials;

[0107] Adopt time series analysis or machine learning algorithms to establish a price fluctuation prediction model to quantify the fluctuation amplitude and frequency of the unit price of energy and materials;

[0108] According to the price fluctuation prediction model, generate the energy unit price fluctuation factor and the material unit price fluctuation factor , and the expression is as shown in the following formula (14):

[0109]

[0110] Among them, and respectively represent the reference prices of energy and materials ; and represent the dynamically adjusted prices corresponding to time ; Update the fluctuation factor through the sliding window method to ensure that the price parameters are synchronized with the real-time market dynamics;

[0111] Construct an equipment efficiency evaluation index system, including energy conversion efficiency, material utilization rate, and equipment failure rate. Adopt the particle swarm optimization algorithm or genetic algorithm, use historical operation data as input, and optimize the equipment efficiency parameters; establish the mapping relationship between equipment efficiency parameters and process parameters. For example: , among which, represents the corrected energy consumption; represents the reference energy consumption; represents the energy conversion efficiency.

[0112] Based on the optimized equipment efficiency parameters, recalculate the energy consumption coefficients and material consumption coefficients of each process link, use the Bayesian inference method to quantify the uncertainty of the corrected parameters, construct a confidence interval, and verify the corrected parameter model through the process simulation platform to ensure that the prediction error is ≤5% within the market fluctuation range of ±10%;

[0113] Set the parameter update period to trigger immediate updates when market fluctuations exceed the threshold, and establish a parameter version management system to record the fluctuation factors, equipment efficiency parameters, and correction bases for each adjustment.

[0114] Step 2: Based on the steelmaking process parameter model, define the state space and action space of the steelmaking process, set the multi-objective reward function and the corresponding constraint violation penalty term, and obtain the reinforcement learning model; among them, the state space includes market share, total cost, carbon emissions, output, and technical stability; the action space is the selection strategies of three steelmaking technologies.

[0115] In some embodiments, the specific process of defining the state space and action space of the steelmaking process, setting the multi-objective reward function and the corresponding constraint violation penalty term, and obtaining the reinforcement learning model based on the steelmaking process parameter model is as follows:

[0116] Take market share, total cost, carbon emissions, output, and technical stability as state variables, and construct a five-tuple state vector as shown in the following formula (1):

[0117]

[0118] Among them, represents the market share of technology at time ; represents the total steelmaking cost at time ; represents the carbon emissions at time ; represents the steel output at time ; represents technical stability, that is, the absolute value of the change in market share between adjacent time steps;

[0119] Take the market share allocation ratios of the three steelmaking technologies as actions, as shown in the following formula (2):

[0120]

[0121] Among them, represents the action that the agent can choose in the state; , and respectively represent the market share allocation ratios corresponding to the three steelmaking technologies; and respectively represent the upper and lower limits of the market share;

[0122] The reward function is as shown in the following formula (3):

[0123]

[0124] and and The expressions of

[0125]

[0126]

[0127]

[0128] wherein and and and respectively represent the weight coefficients of the corresponding state vectors; and and respectively represent the penalty coefficients of the corresponding penalty terms; represents the penalty term for carbon emission constraint; represents the penalty term for market share constraint; represents the penalty term for steel production constraint.

[0129] Step 3. Train the reinforcement learning model based on the Q - learning algorithm, update the Q - value through the state - action - reward interaction loop, and dynamically adjust the selection strategy.

[0130] In some embodiments, the specific process of training the reinforcement learning model based on the Q - learning algorithm, updating the Q - value through the state - action - reward interaction loop, and dynamically adjusting the selection strategy is as follows:

[0131] Initialize the Q - values of all state - action pairs to random values or zero;

[0132] Based on the current state and the ε - greedy policy, select an action , wherein the ε - greedy policy randomly explores the action space with probability ε and selects the action with the largest current Q - value with probability 1 - ε;

[0133] Execute the selected action , trigger the environmental state to transfer from to , and calculate the immediate reward according to the reward function, and the reward function includes a market share reward term, a cost penalty term, a carbon emission penalty term, a technical stability penalty term, and a constraint violation penalty term;

[0134] Update the Q - value using Equation (7), and Equation (7) is as follows:

[0135]

[0136] Among them, represents the Q value; represents the Q-value learning rate; represents the discount factor, which is used to control the influence of future rewards; is the maximum Q value in the next state, representing the return after the agent selects the optimal action;

[0137] The behavior policy is iteratively updated through the following formula (8), as shown in formula (8):

[0138]

[0139] Among them, represents the steelmaking process selected by the agent in state under.

[0140] In some embodiments, the penalty coefficients , and are gradually increased in the later stage of training to ensure that the technical path selected by the agent meets the carbon emission ceiling, market share constraint and production constraint.

[0141] Step 4: Simulate and verify the selection strategy output by the reinforcement learning model. If the simulation result does not reach the expected optimization goal, adjust the corresponding parameters of the reinforcement learning model by the offline gradient descent method and retrain the reinforcement learning model until the optimization goal is met.

[0142] In some embodiments, the specific process of simulating and verifying the selection strategy output by the reinforcement learning model and adjusting the corresponding parameters of the reinforcement learning model by the offline gradient descent method and retraining the reinforcement learning model when the simulation result does not reach the expected optimization goal is as follows:

[0143] Collect historical data, including the market share, total cost, carbon emissions and market share change of each steelmaking process;

[0144] Construct an objective function, as shown in the following formula (9):

[0145]

[0146] Among them, the reward function includes a market share reward term, a cost penalty term, a carbon emission penalty term and a technical stability penalty term;

[0147] Calculate the gradients of the objective function with respect to each weight coefficient , , and , as shown in the following formula (10):

[0148]

[0149] Among them, 、 、 and respectively represent taking partial derivatives with respect to the weight coefficients 、 、 and The weight coefficients are updated by the following formula (11):

[0150]

[0151] Among them, represents the weight learning rate; 、 、 and respectively represent the weight coefficients corresponding to market share, cost, carbon emissions, and technological stability in the reward function at the -th iteration; 、 、 and respectively represent the corresponding updated weight coefficients; represents the discount factor at the -th time; represents the time step of historical data.

[0152] Based on the same inventive concept, corresponding to the steelmaking process optimization method in any of the above embodiments, the present invention provides an electronic device, including a memory and a processor. The memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the steelmaking process optimization method of the embodiment.

[0153] Optionally, the above electronic device may be a server.

[0154] In addition, this embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steelmaking process optimization method of the embodiment.

[0155] It can be understood that the processor in the embodiments of the present invention may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0156] The method steps in the embodiments of the present invention may be implemented in a hardware manner or by a processor executing software instructions. The software instructions may be composed of corresponding software modules, and the software modules may be stored in a random access memory (RAM), flash memory, read-only memory (ROM), programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), registers, hard disks, removable hard disks, CD-ROMs, or any other form of storage medium well known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium may also be a component of the processor. The processor and the storage medium may be located in an ASIC.

[0157] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a storage medium or transmitted through the storage medium. The computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired manner (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or a wireless manner (such as infrared, wireless, microwave, etc.). The storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more integrated available media. The available medium may be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk (SSD)), etc.

Claims

1. A method for optimizing steelmaking process in the steel industry, characterized in that: The following steps are involved: Establish a parameterized process including three steelmaking technologies: blast furnace-converter steelmaking, scrap steelmaking, and reduced iron steelmaking. Define materials, energy consumption, carbon emission coefficient, and cost parameters, and build a parameterized model for steelmaking processes. Based on the steelmaking process parameter model, the state space and action space of the steelmaking process are defined, and the multi-objective reward function and the corresponding constraint violation penalty items are set to obtain a reinforcement learning model; the state space includes market share, total cost, carbon emissions, output and technical stability; the action space is the selection strategy of the three steelmaking technologies; The specific process of defining the state space and action space of the steelmaking process based on the steelmaking process parameter model, setting the multi-objective reward function and the corresponding constraint violation penalty term, and obtaining the reinforcement learning model is as follows: Taking market share, total cost, carbon emissions, output and technical stability as state variables, a five-tuple state vector is constructed, as shown in the following formula (1): in, Display Technology In time market share; Indicates time Total steelmaking cost; Indicates time carbon emissions; Indicates time of steel production; represents technological stationarity, i.e., the absolute value of the change in market share between adjacent time steps; The market share distribution ratio of the three steelmaking technologies is taken as the action, as shown in the following formula (2): in, Indicates that the agent is Optional actions in the state; , and Respectively represent the market share distribution ratios corresponding to the three steelmaking technologies; and Respectively represent the upper and lower limits of market share; The reward function is shown in the following formula (3): , and The expressions of are shown in equations (4), (5) and (6): in, , , and Respectively represent the weight coefficients of the corresponding state vectors; , and Respectively represent the penalty coefficients of the corresponding penalty items; represents the penalty term of carbon emission constraint; represents the penalty term of market share constraint; represents the penalty term for steel production constraint; The reinforcement learning model is trained based on the Q-learning algorithm, and the Q value is updated through the interactive cycle of state-action-reward to dynamically adjust the selection strategy; The specific process of training the reinforcement learning model based on the Q-learning algorithm, updating the Q value through the interactive cycle of state-action-reward, and dynamically adjusting the selection strategy is as follows: Initialize the Q-values ​​of all state-action pairs to random values ​​or zero; Based on the current state and ε-greedy strategy selects actions , where the ε-greedy strategy randomly explores the action space with ε probability and selects the action with the largest current Q value with 1-ε probability; Execute the selected action , triggering the environment state from Transfer to , and calculate the instant reward according to the reward function, wherein the reward function includes a market share reward term, a cost penalty term, a carbon emission penalty term, a technical stability penalty term, and a constraint violation penalty term; The Q value is updated using formula (7), which is as follows: in, represents the Q value; represents the Q value learning rate; represents the discount factor used to control the impact of future rewards; The maximum Q value in the next state, indicating the reward after the agent chooses the best action; The behavior strategy is iteratively updated through the following formula (8), as shown in formula (8): in, Indicates that the agent is in state The steelmaking process selected below; The selection strategy of the reinforcement learning model output is simulated and verified. If the simulation result does not meet the expected optimization goal, the corresponding parameters of the reinforcement learning model are adjusted through the offline gradient descent method and the reinforcement learning model is retrained until the optimization goal is met.

2. The method for optimizing steelmaking process in the steel industry according to claim 1, characterized in that: Introduce market price fluctuation factors, dynamically adjust energy unit prices and material unit prices, optimize equipment efficiency parameters based on historical data, and correct the energy and material consumption coefficients of each process link.

3. The method for optimizing steelmaking process in the steel industry according to claim 2, characterized in that: The specific process of introducing market price fluctuation factors, dynamically adjusting energy unit prices and material unit prices, optimizing equipment efficiency parameters based on historical data, and correcting energy and material consumption coefficients of each process link is as follows: Establish a market database to collect historical price data and supply and demand relationship data of energy and materials; Use time series analysis or machine learning algorithms to establish a price fluctuation prediction model to quantify the fluctuation range and frequency of energy unit prices and material unit prices; Generate energy unit price fluctuation factors and material unit price fluctuation factors based on the price fluctuation prediction model, and update the fluctuation factors through the sliding window method to ensure that the price parameters are synchronized with the real-time market dynamics; Construct an equipment efficiency evaluation index system, including energy conversion efficiency, material utilization rate, and equipment failure rate, and use particle swarm optimization algorithm or genetic algorithm to optimize equipment efficiency parameters with historical operation data as input; Based on the optimized equipment efficiency parameters, the energy consumption coefficient and material consumption coefficient of each process link are recalculated, and the uncertainty of the revised parameters is quantified using the Bayesian inference method, and the confidence interval is constructed. The revised parameter model is verified through the process simulation platform to ensure that the prediction error remains ≤5% within the ±10% market fluctuation range; Set the parameter update cycle, trigger instant updates when market fluctuations exceed the threshold, and establish a parameter version management system to record the fluctuation factors, equipment efficiency parameters and correction basis for each adjustment.

4. The method for optimizing steelmaking process in the steel industry according to claim 1, characterized in that: Gradually increase the penalty coefficient in the later stages of training , and , ensuring that the technical path selected by the intelligent agent meets the carbon emission cap, market share constraints and output constraints.

5. The method for optimizing steelmaking process in the steel industry according to claim 1, characterized in that: The selection strategy for the output of the reinforcement learning model is simulated and verified. If the simulation result does not reach the expected optimization goal, the specific process of adjusting the corresponding parameters of the reinforcement learning model by the offline gradient descent method and retraining the reinforcement learning model is as follows: Collect historical data, including market share, total cost, carbon emissions and market share changes of various steelmaking processes; Construct the objective function as shown in the following formula (9): Among them, the reward function It includes market share bonus, cost penalty, carbon emission penalty and technical stability penalty; Calculate the objective function for each weight coefficient , , and The gradient of is shown in the following formula (10): in, , , and Respectively represent the weight coefficient , , and Calculate partial derivatives; update the weight coefficients using the following formula (11): in, Represents the weight learning rate; , , and Respectively represent The weight coefficients corresponding to market share, cost, carbon emissions and technological stability in the reward function at the iteration; , , and Respectively represent the corresponding weight coefficients after update; Indicates The discount factor of times; Represents the number of time steps of historical data.

6. The method for optimizing steelmaking process in the steel industry according to claim 1, characterized in that: Establish an accounting model for equipment depreciation cost, maintenance cost, environmental protection cost, energy cost, material cost and financial subsidies, construct a carbon emission coefficient matrix for each process link, and calculate the total carbon emissions in combination with energy consumption data. The specific process is as follows: Assume that the total cost includes equipment depreciation cost, equipment maintenance cost, environmental protection equipment investment cost, financial subsidies, and the energy cost and material cost required in different steelmaking processes, as shown in the following formula (12): in, represents the total cost of steelmaking; represents the depreciation cost of equipment; represents the equipment maintenance cost; represents the investment cost of environmental protection equipment; represents the energy cost; The material cost expressed; It means financial subsidy; During the steelmaking process, the carbon emission expression is shown in the following formula (13): in, represents the total amount of carbon emissions; Indicates the steelmaking process; Indicates steelmaking equipment; Indicates the type of energy; Represents the carbon emission factor.

7. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the steelmaking process optimization method for the steel industry according to any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for optimizing the steelmaking process in the steel industry according to any one of claims 1 to 6 is implemented.

Images