A blast furnace multi-parameter coupling optimization energy-saving control system

Abstract

The present application relates to a kind of blast furnace multi-parameter coupling optimization energy-saving control system, belong to blast furnace control system technical field.System includes: raw material and cloth material determination module, for determining initial proportioning scheme and initial cloth matrix;Parameter acquisition module, for according to initial cloth matrix, raw fuel cloth is into blast furnace, obtains the multiple real-time combustion parameters of raw fuel combustion;Parameter screening module, for screening several key real-time combustion parameters of key combustion parameter type;Combustion control module, for adjusting the equipment parameter of blast furnace according to multi-parameter coupling combustion field model framework and current time blast furnace combustion energy consumption;Raw material cloth adjustment module, for adjusting initial raw material proportioning scheme and initial cloth matrix using double-layer optimization framework;Closed-loop control module, for carrying out closed-loop control during blast furnace combustion process.The present application realizes the intelligent control of the energy consumption optimization of blast furnace combustion process multi-parameter, can reduce cost and improve the combustion efficiency of blast furnace.

Description

Technical Field

[0001] This invention relates to the field of blast furnace control system technology, and in particular to a multi-parameter coupling optimization energy-saving control system for blast furnaces. Background Technology

[0002] As a crucial steel production facility, the efficiency and economy of blast furnace operation have a profound impact on global industrial production. Blast furnace optimization technology is a key means to improve production efficiency, reduce energy consumption, and minimize environmental impact. In recent years, research and development in blast furnace operation control and energy conservation has made significant progress in the directions of intelligentization, low carbonization, and high efficiency.

[0003] Traditional blast furnace control systems typically achieve energy-saving control of blast furnace combustion by adjusting the feedstock ratio or simply regulating individual equipment parameters. In other words, traditional blast furnace control systems often focus only on optimizing one aspect (such as equipment parameter control or feedstock ratio control), neglecting the interactions between different parameter types. This results in low blast furnace combustion utilization and poor energy-saving control performance.

[0004] Therefore, the present invention provides a multi-parameter coupling optimization energy-saving control system for blast furnaces. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a multi-parameter coupled optimization energy-saving control system for blast furnaces. The technical solution of this invention is as follows:

[0006] A multi-parameter coupled optimization energy-saving control system for blast furnaces includes:

[0007] The raw material and charge determination module is used to determine the initial proportioning scheme and initial charge matrix based on the historical combustion data of the blast furnace.

[0008] The parameter acquisition module is used to mix raw materials according to the initial proportioning scheme to obtain raw materials and fuels, and to distribute the raw materials and fuels into the blast furnace according to the initial distribution matrix, and to acquire real-time combustion parameters of multiple combustion parameter types of raw materials and fuels during the combustion process in the blast furnace.

[0009] The parameter filtering module is used to calculate the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the blast furnace combustion energy consumption at the current moment, and to filter the key real-time combustion parameters of several key combustion parameter types based on the correlation coefficient of each combustion parameter type;

[0010] The combustion control module is used to construct a multi-parameter coupled combustion field model framework based on the inherent relationship of all key real-time combustion parameters, and to adjust the equipment parameters of the blast furnace according to the multi-parameter coupled combustion field model framework and the current blast furnace combustion energy consumption.

[0011] The raw material distribution adjustment module is used to obtain the total blast furnace combustion energy consumption for the complete combustion of raw materials and fuels in the blast furnace. It adopts a two-layer optimization framework to adjust the initial raw material ratio scheme and the initial distribution matrix in combination with the total blast furnace combustion energy consumption.

[0012] The closed-loop control module is used to perform closed-loop control of the blast furnace combustion process based on the adjusted initial raw material ratio scheme and the initial charging matrix.

[0013] Preferably, the raw material and fabric determining module includes:

[0014] The initial proportioning calculation unit is used to construct an objective function based on the historical combustion data of the blast furnace and the performance of each raw material, and to calculate the initial proportioning scheme based on the objective function with minimum combustion energy consumption as a constraint.

[0015] The initial material distribution matrix generation unit is used to obtain the original material distribution method based on the historical combustion data of the blast furnace, and generate the initial material distribution method based on the original material distribution method and the initial proportioning scheme. The initial material distribution method is then filled into the two-dimensional material distribution orientation matrix to obtain the initial material distribution matrix.

[0016] Preferably, the parameter filtering module includes:

[0017] The combustion energy consumption calculation unit is used to calculate the blast furnace combustion energy consumption at the current moment based on the combustion information of raw materials and fuels and the real-time combustion parameters of each combustion parameter type.

[0018] The correlation coefficient calculation unit is used to calculate the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the blast furnace combustion energy consumption at the current moment.

[0019] The key combustion type screening unit is used to screen real-time combustion parameters of several combustion parameter types with correlation coefficients greater than the correlation threshold as key real-time combustion parameters of several key combustion parameter types.

[0020] Preferably, the correlation coefficient calculation unit includes:

[0021] The standardization subunit is used to standardize the real-time combustion parameters of each combustion parameter type to obtain the standard combustion parameters corresponding to each combustion parameter type.

[0022] The point mapping sub-unit is used to construct a two-dimensional combustion point representing the relationship between each standard combustion parameter and the blast furnace combustion energy consumption at the current moment. All two-dimensional combustion points are mapped to a preset standard grid. Each vertical line in the preset standard grid is used to divide the value range of blast furnace combustion energy consumption vertically, and each horizontal line is used to divide the value range of standard combustion parameter values ​​horizontally. Each horizontal and vertical line is configured with a preset probability density after the corresponding value range division.

[0023] The marginal probability calculation sub-unit is used to take the preset probability density of the horizontal line where the standard combustion parameters of each two-dimensional combustion point in the preset standard grid are located as its combustion marginal probability, and the preset probability density of the vertical line where the blast furnace combustion energy consumption is located at the current moment as its energy consumption marginal probability.

[0024] The correlation coefficient calculation subunit calculates the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the current blast furnace combustion energy consumption based on the combustion marginal probability and energy consumption marginal probability of the standard combustion parameters of each combustion parameter type.

[0025] Preferably, when the correlation coefficient calculation subunit calculates the correlation coefficient I between the real-time combustion parameter of any combustion parameter type and the current blast furnace combustion energy consumption based on the combustion marginal probability and energy consumption marginal probability of the standard combustion parameter of any combustion parameter type, it does so using formula (1):

[0026] (1);

[0027] In formula (1), X represents the value of the standard combustion parameter of the combustion parameter type, Y represents the value of the blast furnace combustion energy consumption at the current moment, p(X,Y) represents the joint probability of the standard combustion parameter of the combustion parameter type and the blast furnace combustion energy consumption at the current moment, p(X) represents the combustion marginal probability of the standard combustion parameter of the combustion parameter type, p(Y) represents the energy consumption marginal probability of the blast furnace combustion energy consumption at the current moment, log() represents the logarithmic function, and e represents the smoothing coefficient.

[0028] Preferably, the combustion control module includes:

[0029] The time-series combustion data acquisition unit is used to acquire the time-series combustion data of each key real-time combustion parameter within a preset time period with the current time as the end point. The time-series combustion data of all key real-time combustion parameters are time-aligned.

[0030] The intrinsic relationship judgment unit is used to determine the intrinsic relationship between all key real-time combustion parameters based on causal relationships according to each time-series combustion data.

[0031] A multi-parameter coupled combustion field model framework building unit is used to construct a multi-parameter coupled combustion field model framework based on the intrinsic relationship between all key real-time combustion parameters.

[0032] The combustion control unit is used to input all key real-time combustion parameters into the multi-parameter coupled combustion field model framework to obtain the energy consumption threshold. If the blast furnace combustion energy consumption is higher than the energy consumption threshold at the current moment, the equipment parameters of each device in the blast furnace are adjusted to reduce the blast furnace combustion energy consumption to below the energy consumption threshold.

[0033] Preferably, the intrinsic relationship determination unit includes:

[0034] The hysteresis coefficient configuration subunit is used to calculate the hysteresis coefficient of each pair of key real-time combustion parameters based on the statistical characteristics of each pair of time-series combustion data.

[0035] The causal relationship testing subunit is used to construct two causal relationship regression models with mutual relationship based on the lag coefficients of each pair of time-series combustion data and each pair of key real-time combustion parameters. If the statistical probability of the F-test value of any causal relationship regression model of any two key real-time combustion parameters is less than the preset critical value, the causal relationship is determined to be valid, and the intrinsic relationship between the two key real-time combustion parameters is determined. Otherwise, the causal relationship is not valid.

[0036] Preferably, the raw material fabric adjustment module includes:

[0037] The total blast furnace combustion energy consumption acquisition unit is used to acquire the cumulative amount of blast furnace combustion energy consumption of raw materials and fuels from the start of combustion to complete combustion in the blast furnace, and to obtain the total blast furnace combustion energy consumption of raw materials and fuels in complete combustion in the blast furnace.

[0038] The dual-layer optimization framework construction unit is used to construct the dual-layer optimization framework by taking the raw material ratio-blast furnace combustion energy consumption model as the outer framework, the charging matrix-blast furnace combustion energy consumption model as the inner framework, the total blast furnace combustion energy consumption as the minimum threshold, and the production standard, the initial ratio scheme and the initial charging matrix as the production constraints.

[0039] The optimization unit is used to optimize the outer frame according to the two-layer optimization framework to obtain the optimal proportion scheme, and to optimize the inner frame according to the optimal proportion scheme to obtain the optimal fabric matrix.

[0040] The adjustment unit is used to adjust the initial proportioning scheme and the initial fabric matrix through the optimal proportioning scheme and the optimal fabric matrix.

[0041] Preferably, the optimization unit includes:

[0042] The weighting sub-unit is used to assign weighting to the data domain of the raw material ratio-blast furnace combustion energy consumption model and to assign weighting to the data domain of the charge matrix-blast furnace combustion energy consumption model.

[0043] The outer optimization subunit is used to input the initial proportioning scheme and the initial material distribution matrix into the two-layer optimization framework. Based on the total blast furnace combustion energy consumption and the initial proportioning scheme, the outer loss value is calculated through the proportioning weight. The raw material proportioning-blast furnace combustion energy consumption model is superimposed with the outer loss value to obtain a new raw material proportioning-blast furnace combustion energy consumption model. The new raw material proportioning-blast furnace combustion energy consumption model is minimized according to the production constraints to obtain the optimal proportioning scheme.

[0044] The inner layer optimization subunit is used to calculate the inner layer loss value based on the total blast furnace combustion energy consumption, the initial charging matrix, and the optimal proportion scheme through the charging weight. The inner layer loss value is superimposed on the charging matrix-blast furnace combustion energy consumption model to obtain a new charging matrix-blast furnace combustion energy consumption model. The optimal spatial solution of the new charging matrix-blast furnace combustion energy consumption model is calculated based on the production constraints and the global optimization algorithm. The optimal charging matrix is ​​determined based on the optimal spatial solution.

[0045] All of the above-mentioned optional technical solutions can be combined arbitrarily, and the present invention will not provide a detailed description of the structure after each combination.

[0046] By means of the above solution, the beneficial effects of the present invention are as follows:

[0047] The parameter filtering module calculates the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the current blast furnace combustion energy consumption, and filters several key real-time combustion parameters based on the correlation coefficient. This allows for the extraction of key parameters that have a significant impact on the current blast furnace combustion energy consumption, providing an accurate data foundation for subsequent energy-saving control.

[0048] By constructing a multi-parameter coupled combustion field model framework based on the inherent relationship of all key real-time combustion parameters through the combustion control module, and adjusting the blast furnace equipment parameters according to the multi-parameter coupled combustion field model framework and the current blast furnace combustion energy consumption, the operation of the blast furnace equipment can be controlled according to the real-time combustion parameters, thereby improving the blast furnace combustion utilization rate and optimizing the energy-saving control effect.

[0049] By using a dual-layer optimization framework in conjunction with the total blast furnace combustion energy consumption adjustment module, the initial raw material ratio scheme and initial material distribution matrix are adjusted according to the feedback of the total blast furnace combustion energy consumption, thereby continuously optimizing the energy-saving control effect.

[0050] This invention fully considers the interrelationship between real-time combustion parameters. On the one hand, it optimizes the energy-saving control effect from the perspective of real-time adjustment of blast furnace equipment parameters. On the other hand, it further optimizes the energy-saving control effect from the perspective of optimizing the initial raw material ratio scheme and the initial charging matrix. It realizes intelligent control of energy consumption optimization of multiple parameters in the blast furnace combustion process, which can reduce costs and improve the combustion efficiency of the blast furnace.

[0051] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of a multi-parameter coupling optimization energy-saving control system for a blast furnace provided in an embodiment of the present invention. Detailed Implementation

[0053] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0054] like Figure 1 As shown, this embodiment of the invention provides a multi-parameter coupled optimization energy-saving control system for a blast furnace, which includes:

[0055] The raw material and charge determination module is used to determine the initial proportioning scheme and initial charge matrix based on the historical combustion data of the blast furnace.

[0056] The parameter acquisition module is used to mix raw materials according to the initial proportioning scheme to obtain raw materials and fuels, and to distribute the raw materials and fuels into the blast furnace according to the initial distribution matrix, and to acquire real-time combustion parameters of multiple combustion parameter types of raw materials and fuels during the combustion process in the blast furnace.

[0057] The parameter filtering module is used to calculate the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the blast furnace combustion energy consumption at the current moment, and to filter the key real-time combustion parameters of several key combustion parameter types based on the correlation coefficient of each combustion parameter type;

[0058] The combustion control module is used to construct a multi-parameter coupled combustion field model framework based on the inherent relationship of all key real-time combustion parameters, and to adjust the equipment parameters of the blast furnace according to the multi-parameter coupled combustion field model framework and the current blast furnace combustion energy consumption.

[0059] The raw material distribution adjustment module is used to obtain the total blast furnace combustion energy consumption for the complete combustion of raw materials and fuels in the blast furnace. It adopts a two-layer optimization framework to adjust the initial raw material ratio scheme and the initial distribution matrix in combination with the total blast furnace combustion energy consumption.

[0060] The closed-loop control module is used to perform closed-loop control of the blast furnace combustion process based on the adjusted initial raw material ratio scheme and the initial charging matrix.

[0061] Specifically, in the raw material and material distribution determination module, the blast furnace historical combustion data refers to historical data related to multiple combustion parameter types collected during the blast furnace operation over a period of time. The combustion parameter types include heat, temperature, gas flow rate, fuel consumption, etc.

[0062] The initial proportioning scheme refers to the ratio of all raw materials used in the blast furnace combustion process, such as 10% coke, 50% iron ore, and 40% limestone. The initial charging matrix refers to the matrix in which raw materials and fuels are arranged in the blast furnace according to a predetermined charging method. Each row of the initial charging matrix represents one layer of the blast furnace, and each column represents the charging parameters for one type of raw material or fuel, including the proportion of raw materials and fuels, the thickness and width of the charging layer, etc. Since the blast furnace area is generally divided into three layers (upper, middle, and lower), the initial charging matrix consists of three rows.

[0063] Specifically, in the parameter filtering module, the correlation coefficient is a numerical value representing the degree of influence of each combustion parameter type's real-time combustion parameter on the blast furnace combustion energy consumption at the current moment. A positive number indicates a positive correlation, and a negative number indicates a negative correlation. The types of real-time combustion parameters include temperature, flow rate, and fuel consumption, etc.

[0064] Specifically, taking an initial feeding matrix consisting of three rows and three columns as an example, the feeding parameter in the first column is the proportion of raw materials and fuel, and the feeding parameter in the second column is the thickness of the feeding material. If the data in the first row and first column is 30 and the data in the first row and second column is 100, then when feeding raw materials and fuel into the blast furnace according to the initial feeding matrix, the proportion of raw materials and fuel in the first layer of feeding material in the blast furnace is 30%, and the thickness of the first layer of feeding material in the blast furnace is 100mm.

[0065] Specifically, in the combustion control module, the multi-parameter coupled combustion field model framework is a multi-layered relationship model that includes the intrinsic relationships between all key real-time combustion parameters and blast furnace combustion energy consumption. This multi-parameter coupled combustion field model framework enables the prediction of energy consumption thresholds. The intrinsic relationship between each pair of combustion parameter types refers to the causal relationship between them.

[0066] Blast furnace combustion energy consumption is determined by parameters such as heat, temperature, gas flow rate, fuel consumption, and combustion information of raw materials and fuels. It represents the sum of various energy consumptions of raw materials and fuels during combustion in the blast furnace. The blast furnace combustion energy consumption at the current moment is obtained by superimposing the fuel combustion energy, blast heat energy, and heat energy from other heat sources at the current moment.

[0067] The equipment parameters of a blast furnace include oxygen supply, furnace temperature, gas pressure, and blower speed.

[0068] Specifically, in the raw material feeding adjustment module, complete combustion of raw materials and fuels in the blast furnace refers to the combustion process of raw materials and fuels inside the blast furnace reaching ideal reaction conditions. Complete combustion generally means that the remaining raw material and fuel residue in the blast furnace after combustion is negligible. Total blast furnace combustion energy consumption refers to the cumulative amount of blast furnace combustion energy consumption after complete combustion of raw materials and fuels.

[0069] A two-layer optimization framework is a two-layer framework for optimization under multiple constraints. It typically includes two optimization levels: an outer layer optimization and an inner layer optimization. In this embodiment of the invention, the optimization objective of the outer layer optimization is the initial proportioning scheme, and the optimization objective of the inner layer optimization is the initial fabric matrix.

[0070] Specifically, in the closed-loop control module, closed-loop control is the process of controlling the blast furnace combustion process based on the adjusted initial raw material ratio scheme and initial charging matrix. Through closed-loop control, the raw material ratio scheme and charging matrix of the blast furnace can be continuously adjusted to ensure that the blast furnace combustion process is continuously improved in terms of energy saving.

[0071] In one specific embodiment, the raw material and fabric determination module includes:

[0072] The initial proportioning calculation unit is used to construct an objective function based on the historical combustion data of the blast furnace and the performance of each raw material, and to calculate the initial proportioning scheme based on the objective function with minimum combustion energy consumption as a constraint.

[0073] The initial material distribution matrix generation unit is used to obtain the original material distribution method based on the historical combustion data of the blast furnace, and generate the initial material distribution method based on the original material distribution method and the initial proportioning scheme. The initial material distribution method is then filled into the two-dimensional material distribution orientation matrix to obtain the initial material distribution matrix.

[0074] Specifically, in the initial proportioning calculation unit, the performance of the raw materials refers to parameters such as calorific value, chemical composition, particle size distribution, and melting point. In this embodiment of the invention, the objective function is a function constructed based on the historical combustion data of the blast furnace and the performance of each raw material. With minimum combustion energy consumption as a constraint, the optimal solution can be obtained by minimizing the objective function, and this optimal solution constitutes the initial proportioning scheme. The historical combustion data of the blast furnace includes the type of combustion raw material, combustion energy consumption, combustion parameters, and the original charging method, etc.

[0075] In the initial charge distribution matrix generation unit, the initial charge distribution method is obtained by adjusting the original charge distribution method in detail according to the initial mix design under the guidance of experts. The two-dimensional charge distribution matrix is ​​a blank matrix used to describe the spatial distribution of raw materials and fuels within the blast furnace.

[0076] In one specific embodiment, the parameter filtering module includes:

[0077] The combustion energy consumption calculation unit is used to calculate the blast furnace combustion energy consumption at the current moment based on the combustion information of raw materials and fuels and the real-time combustion parameters of each combustion parameter type.

[0078] The correlation coefficient calculation unit is used to calculate the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the blast furnace combustion energy consumption at the current moment.

[0079] The key combustion type screening unit is used to screen real-time combustion parameters of several combustion parameter types with correlation coefficients greater than the correlation threshold as key real-time combustion parameters of several key combustion parameter types.

[0080] Specifically, in the combustion energy consumption calculation unit, the combustion information of raw materials and fuels includes the calorific value, oxygen requirement, combustion rate, and oxygen cost of the raw materials and fuels. When calculating the blast furnace combustion energy consumption at the current moment based on the combustion information of raw materials and fuels and the real-time combustion parameters of each combustion parameter type, the calculation formula can be expressed as: combustion rate × calorific value of raw materials and fuels + oxygen requirement × oxygen cost.

[0081] In the critical combustion type screening unit, the correlation threshold is a value representing the lowest correlation, determined empirically.

[0082] In one specific embodiment, the correlation coefficient calculation unit includes:

[0083] The standardization subunit is used to standardize the real-time combustion parameters of each combustion parameter type to obtain the standard combustion parameters corresponding to each combustion parameter type.

[0084] The point mapping sub-unit is used to construct a two-dimensional combustion point representing the relationship between each standard combustion parameter and the blast furnace combustion energy consumption at the current moment. All two-dimensional combustion points are mapped to a preset standard grid. Each vertical line in the preset standard grid is used to divide the value range of blast furnace combustion energy consumption vertically, and each horizontal line is used to divide the value range of standard combustion parameter values ​​horizontally. Each horizontal and vertical line is configured with a preset probability density after the corresponding value range division.

[0085] The marginal probability calculation sub-unit is used to take the preset probability density of the horizontal line where the standard combustion parameters of each two-dimensional combustion point in the preset standard grid are located as its combustion marginal probability, and the preset probability density of the vertical line where the blast furnace combustion energy consumption is located at the current moment as its energy consumption marginal probability.

[0086] The correlation coefficient calculation subunit calculates the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the current blast furnace combustion energy consumption based on the combustion marginal probability and energy consumption marginal probability of the standard combustion parameters of each combustion parameter type.

[0087] Specifically, in the standardization subunit, extremum standardization is a common method for dividing data into the range [0,1]. In this embodiment of the invention, for real-time combustion parameters of a certain combustion parameter type, when performing extremum standardization, a historical combustion parameter (generally 20 minutes) of that combustion parameter type is first obtained, where the time endpoint is the current moment; then, the real-time combustion parameters of that combustion parameter type are performed on extremum standardization using formula (2) to obtain the standard combustion parameter X corresponding to that combustion parameter type:

[0088] (2);

[0089] In formula (2), x0 represents the real-time combustion parameter of the combustion parameter type, x represents the historical combustion parameter of the combustion parameter type, min(x) represents the minimum value of the historical combustion parameter, and max(x) represents the maximum value of the historical combustion parameter.

[0090] In the point mapping subunit, the two-dimensional combustion point includes two parameters: standard combustion parameters and blast furnace combustion energy consumption at the current moment. Its structure is: (standard combustion parameters, blast furnace combustion energy consumption at the current moment).

[0091] The preset standard grid is a pre-divided grid with the origin at the lower left corner. The horizontal axis represents the value range of standard combustion parameters. As shown in the above embodiment, the value range of all standard combustion parameters is between [0,1]. The vertical axis of the preset standard grid represents the value range of blast furnace combustion energy consumption. The value range of blast furnace combustion energy consumption is determined by a large amount of historical blast furnace combustion data, and the value range of blast furnace combustion energy consumption is used as the vertical range of the preset standard grid for division. During the division, the number of vertical and horizontal lines in the preset standard grid is determined by expert review (generally 10 each).

[0092] In addition, when configuring the preset probability density after dividing the corresponding value range for each horizontal and vertical line, a large amount of historical blast furnace combustion data is obtained and processed according to normal distribution. Then, the probability density of each value range interval is calculated and configured in the preset standard grid.

[0093] Taking the preset probability density corresponding to the value range division of the vertical line configuration of the preset standard grid as an example, the value range of blast furnace combustion energy consumption is divided into multiple value range intervals according to the number of vertical lines (the range of each value range interval is left-closed and right-open). The number of data points in each value range interval of a large amount of historical blast furnace combustion data is counted. The probability density of each value range interval is calculated according to the number of data points in each value range interval and the probability density calculation formula. The probability density of each value range interval is used as the preset probability density of the vertical line on the left side of its value range interval.

[0094] When configuring the preset probability density after value range division for the horizontal lines of the preset standard grid, refer to the method for configuring the preset probability density after value range division for the vertical lines. However, since the probability density of each combustion parameter type is not necessarily the same, when configuring the preset probability density after value range division for each horizontal line, it is necessary to add a label for the combustion parameter type to the preset probability density of each horizontal line. For example, on the horizontal line with a value range of 0.4, the preset probability density of the wind speed combustion parameter is 0.6, and the preset probability density of the temperature combustion parameter is 0.8. On the bus with a value range of 16000, the preset probability density of the blast furnace combustion energy consumption at the current moment is 0.5.

[0095] To facilitate understanding, a concrete example is given below: The default horizontal range of the standard grid is [0, 1), which is divided into five intervals: [0, 02), [0.2, 0.4), [0.4, 0.6), [0.6, 0.8), and [0.8, 1.0). The default vertical range of the standard grid is [0, 20000), which is divided into five intervals: [0, 4000), [4000, 8000), and [0, 20000). If the two-dimensional combustion point of the wind speed combustion parameter is (0.5, 18000), then the horizontal line corresponding to it in the preset standard grid is 0.4 and the vertical line is 16000. Taking the above example, the preset probability density of the horizontal line 0.4 corresponding to the wind speed combustion parameter is 0.6, and the preset probability density of the vertical line 16000 corresponding to the current blast furnace combustion energy consumption is 0.5.

[0096] In a specific embodiment, the correlation coefficient calculation subunit calculates the correlation coefficient I between the real-time combustion parameter of any combustion parameter type and the current blast furnace combustion energy consumption based on the combustion marginal probability and energy consumption marginal probability of the standard combustion parameter of any combustion parameter type, using formula (1):

[0097] (1);

[0098] In formula (1), X represents the value of the standard combustion parameter of the combustion parameter type, Y represents the value of the blast furnace combustion energy consumption at the current moment, p(X,Y) represents the joint probability of the standard combustion parameter of the combustion parameter type and the blast furnace combustion energy consumption at the current moment, p(X) represents the combustion marginal probability of the standard combustion parameter of the combustion parameter type, p(Y) represents the energy consumption marginal probability of the blast furnace combustion energy consumption at the current moment, log() represents the logarithmic function; e represents the smoothing coefficient, which is an empirical value; p(X,Y)=p(X)×p(Y).

[0099] In one specific embodiment, the combustion control module includes:

[0100] The time-series combustion data acquisition unit is used to acquire the time-series combustion data of each key real-time combustion parameter within a preset time period with the current time as the end point. The time-series combustion data of all key real-time combustion parameters are time-aligned.

[0101] The intrinsic relationship judgment unit is used to determine the intrinsic relationship between all key real-time combustion parameters based on causal relationships according to each time-series combustion data.

[0102] A multi-parameter coupled combustion field model framework building unit is used to construct a multi-parameter coupled combustion field model framework based on the intrinsic relationship between all key real-time combustion parameters.

[0103] The combustion control unit is used to input all key real-time combustion parameters into the multi-parameter coupled combustion field model framework to obtain the energy consumption threshold. If the blast furnace combustion energy consumption is higher than the energy consumption threshold at the current moment, the equipment parameters of each device in the blast furnace are adjusted to reduce the blast furnace combustion energy consumption to below the energy consumption threshold.

[0104] Specifically, in the time-series combustion data acquisition unit, all time-series combustion data are aligned according to the time standard interval. If any time-series combustion data has an outlier or missing value, the value of the previous time node is used to fill the position of the outlier or missing value.

[0105] In the intrinsic relationship judgment unit, intrinsic relationship refers to the causal relationship between all key real-time combustion parameters. For example, since an increase in the temperature parameter type of real-time combustion parameter will lead to an increase in the gas emission parameter type of real-time combustion parameter, there is an intrinsic relationship between the temperature parameter type of real-time combustion parameter and the gas emission parameter type of real-time combustion parameter.

[0106] In the construction unit of the multi-parameter coupled combustion field model framework, the multi-parameter coupled combustion field model framework refers to a model that predicts the blast furnace combustion energy consumption at each moment by modeling the intrinsic relationship between all key real-time combustion parameters based on the coupling relationship of multiple parameters.

[0107] In the combustion control unit, when adjusting the equipment parameters of each piece of equipment in the blast furnace, all key real-time combustion parameters are input into the controller, which then outputs the adjusted equipment parameters for each piece of equipment in each blast furnace. The controller can be a PID controller, proportional controller, fuzzy controller, etc.

[0108] In one specific embodiment, the intrinsic relationship determination unit includes:

[0109] The hysteresis coefficient configuration subunit is used to calculate the hysteresis coefficient of each pair of key real-time combustion parameters based on the statistical characteristics of each pair of time-series combustion data.

[0110] The causal relationship testing subunit is used to construct two causal relationship regression models with mutual relationship based on the lag coefficients of each pair of time-series combustion data and each pair of key real-time combustion parameters. If the statistical probability of the F-test value of any causal relationship regression model of any two key real-time combustion parameters is less than the preset critical value, the causal relationship is determined to be valid, and the intrinsic relationship between the two key real-time combustion parameters is determined. Otherwise, the causal relationship is not valid.

[0111] Specifically, in the lag coefficient configuration sub-unit, the statistical characteristics include extreme values, mean, and variance; when calculating the lag coefficient T of the corresponding two key real-time combustion parameters based on the statistical characteristics of any two time-series combustion data, it is achieved through formula (3):

[0112] (3);

[0113] In formula (3), θ represents the optimal lag time, represents the time corresponding to the maximum matching degree between the two time-series combustion data, U(t) represents the value of the first key real-time combustion parameter at the current time, V(t-θ) represents the value of the second key real-time combustion parameter at the current time with a lag of θ, and argmax θ [] represents a decreasing function for finding the optimal lag time θ.

[0114] In calculating the maximum matching degree between the two time-series combustion data, the average value of each data point at each time step is subtracted from the average value of the two data points to obtain the deviation value of the data points at each time step. Then, the ratio of the deviation values ​​of the two time-series combustion data at each time step is calculated as the matching degree of the two time-series combustion data at each time step. Finally, the value with the largest matching degree is selected from all matching degrees as the maximum matching degree.

[0115] In the causal relationship testing subunit, the mutual relationship refers to the mutual relationship between cause and effect; the causal relationship regression model is a model of the relationship between the change of one key real-time combustion parameter and the change of another key real-time combustion parameter; the F-test value is obtained by calculating the ratio of the regression mean square to the residual mean square of the causal relationship regression model; the confidence level of the causal relationship regression model can be judged by judging the statistical probability of the F-test value and the magnitude of the preset critical value. The preset critical value is an empirical value, usually set to 0.05.

[0116] When determining the intrinsic relationship between two key real-time combustion parameters, the causal relationship between these two key real-time combustion parameters is directly taken as their intrinsic relationship.

[0117] In one specific embodiment, the raw material fabric adjustment module includes:

[0118] The total blast furnace combustion energy consumption acquisition unit is used to acquire the cumulative amount of blast furnace combustion energy consumption of raw materials and fuels from the start of combustion to complete combustion in the blast furnace, and to obtain the total blast furnace combustion energy consumption of raw materials and fuels in complete combustion in the blast furnace.

[0119] The dual-layer optimization framework construction unit is used to construct the dual-layer optimization framework by taking the raw material ratio-blast furnace combustion energy consumption model as the outer framework, the charging matrix-blast furnace combustion energy consumption model as the inner framework, the total blast furnace combustion energy consumption as the minimum threshold, and the production standard, the initial ratio scheme and the initial charging matrix as the production constraints.

[0120] The optimization unit is used to optimize the outer frame according to the two-layer optimization framework to obtain the optimal proportion scheme, and to optimize the inner frame according to the optimal proportion scheme to obtain the optimal fabric matrix.

[0121] The adjustment unit is used to adjust the initial proportioning scheme and the initial fabric matrix through the optimal proportioning scheme and the optimal fabric matrix.

[0122] Specifically, in the total blast furnace combustion energy consumption acquisition unit, the cumulative blast furnace combustion energy consumption refers to the cumulative amount of blast furnace combustion energy consumption during all combustion stages from the start of combustion to complete combustion of raw materials and fuels in the blast furnace. The combustion stages from the start of combustion to complete combustion include the initial combustion stage, the stable combustion stage, and the complete combustion stage.

[0123] In the two-layer optimization framework construction unit, the raw material ratio-blast furnace combustion energy consumption model is a model used to describe the quantitative relationship between the initial ratio scheme and the blast furnace combustion energy consumption. In this embodiment of the invention, the raw material ratio-blast furnace combustion energy consumption model can be selected as a heat balance model. The charging matrix-blast furnace combustion energy consumption model is a model used to describe the quantitative relationship between the initial charging matrix and the blast furnace combustion energy consumption. In this embodiment of the invention, the charging matrix-blast furnace combustion energy consumption model can be selected as a random forest model. Production standards include pre-set thresholds for parameters such as production cost, minimum production energy consumption, and production pollution.

[0124] In one specific embodiment, the optimization unit includes:

[0125] The weighting sub-unit is used to assign weighting to the data domain of the raw material ratio-blast furnace combustion energy consumption model and to assign weighting to the data domain of the charge matrix-blast furnace combustion energy consumption model.

[0126] The outer optimization subunit is used to input the initial proportioning scheme and the initial material distribution matrix into the two-layer optimization framework. Based on the total blast furnace combustion energy consumption and the initial proportioning scheme, the outer loss value is calculated through the proportioning weight. The raw material proportioning-blast furnace combustion energy consumption model is superimposed with the outer loss value to obtain a new raw material proportioning-blast furnace combustion energy consumption model. The new raw material proportioning-blast furnace combustion energy consumption model is minimized according to the production constraints to obtain the optimal proportioning scheme.

[0127] The inner layer optimization subunit is used to calculate the inner layer loss value based on the total blast furnace combustion energy consumption, the initial charging matrix, and the optimal proportion scheme through the charging weight. The inner layer loss value is superimposed on the charging matrix-blast furnace combustion energy consumption model to obtain a new charging matrix-blast furnace combustion energy consumption model. The optimal spatial solution of the new charging matrix-blast furnace combustion energy consumption model is calculated based on the production constraints and the global optimization algorithm. The optimal charging matrix is ​​determined based on the optimal spatial solution.

[0128] Specifically, in the weighted proportioning sub-unit, assigning weights to different value ranges is to emphasize the influence of the initial proportioning scheme and the initial charging matrix on the total blast furnace combustion energy consumption during the optimization process of the two-layer optimization framework. Among them, the proportioning weight is a historical empirical value that quantifies the influence of each raw material proportion in the initial proportioning scheme, and the charging weight is a historical empirical value that quantifies the influence of the charging parameters in each row of the initial charging matrix.

[0129] In the outer optimization sub-unit, the proportions of each raw material in the initial proportioning scheme are weighted and summed according to their respective weights. This summation is then compared with the total blast furnace combustion energy consumption to obtain the outer loss value. In this embodiment of the invention, the objective function of a commonly used minimization problem is minimized.

[0130] In the inner optimization sub-unit, the distribution parameters of each row in the initial distribution matrix are weighted and summed according to the distribution weights to obtain the weighted distribution result for each row. The ratio of the weighted distribution result for each row to the total blast furnace combustion energy consumption is calculated, and the ratios of each row are accumulated to obtain the inner layer loss value. A global optimization algorithm is used to solve the optimal spatial solution of the new distribution matrix-blast furnace combustion energy consumption model constrained by production constraints. In this embodiment, the global optimization algorithm uses particle swarm optimization. The optimal spatial solution obtained by the global optimization algorithm represents the distribution parameters of each layer of the blast furnace. Combining the distribution parameters of each layer of the blast furnace yields the optimal distribution matrix.

[0131] Based on all the above embodiments, the blast furnace multi-parameter coupling optimization energy-saving control system provided by the present invention has the following beneficial effects:

[0132] First, the raw material and charge distribution module determines the initial mix ratio and initial charge distribution matrix based on the blast furnace's historical combustion data. This allows for precise determination of the initial mix ratio and charge distribution matrix by combining historical blast furnace combustion data. Second, the parameter acquisition module acquires real-time combustion parameters of multiple combustion parameter types of raw materials and fuels in the blast furnace, providing diverse data for subsequent multi-parameter coupled optimization control of the blast furnace.

[0133] Next, the combustion control module constructs a multi-parameter coupled combustion field model framework based on the inherent relationship of all key real-time combustion parameters. Based on the multi-parameter coupled combustion field model framework and the current blast furnace combustion energy consumption, the equipment parameters of the blast furnace are adjusted. This allows the equipment parameters of the blast furnace to be adjusted through the interrelationship between real-time combustion parameters. The operation of the blast furnace equipment can be controlled according to the real-time combustion parameters, thereby improving the combustion utilization rate of the blast furnace and optimizing the energy-saving control effect.

[0134] By using a dual-layer optimization framework in conjunction with the total blast furnace combustion energy consumption adjustment module, the initial raw material ratio scheme and initial feeding matrix are adjusted according to the total blast furnace combustion energy consumption feedback, thereby continuously optimizing the energy-saving control effect.

[0135] Finally, the closed-loop control module performs closed-loop control of the blast furnace combustion process based on the adjusted initial raw material ratio scheme and the initial charging matrix, thus realizing the closed-loop control of the blast furnace combustion process.

[0136] In summary, this invention fully considers the interrelationships between real-time combustion parameters when performing energy-saving optimization control of blast furnaces. On the one hand, it optimizes the energy-saving control effect by adjusting the equipment parameters of the blast furnace in real time. On the other hand, it further optimizes the energy-saving control effect by optimizing the initial raw material ratio scheme and the initial charging matrix. This achieves intelligent control of energy consumption optimization of multiple parameters in the blast furnace combustion process, which can reduce costs and improve the combustion efficiency of the blast furnace.

[0137] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A multi-parameter coupled optimization energy-saving control system for blast furnaces, characterized in that, include: The raw material and charge determination module is used to determine the initial proportioning scheme and initial charge matrix based on the historical combustion data of the blast furnace. The parameter acquisition module is used to mix raw materials according to the initial proportioning scheme to obtain raw materials and fuels, and to distribute the raw materials and fuels into the blast furnace according to the initial distribution matrix, and to acquire real-time combustion parameters of multiple combustion parameter types of raw materials and fuels during the combustion process in the blast furnace. The parameter filtering module is used to calculate the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the blast furnace combustion energy consumption at the current moment, and to filter the key real-time combustion parameters of several key combustion parameter types based on the correlation coefficient of each combustion parameter type; The combustion control module is used to construct a multi-parameter coupled combustion field model framework based on the inherent relationship of all key real-time combustion parameters, and to adjust the equipment parameters of the blast furnace according to the multi-parameter coupled combustion field model framework and the current blast furnace combustion energy consumption. The raw material distribution adjustment module is used to obtain the total blast furnace combustion energy consumption for the complete combustion of raw materials and fuels in the blast furnace. It adopts a two-layer optimization framework to adjust the initial raw material ratio scheme and the initial distribution matrix in combination with the total blast furnace combustion energy consumption. The raw material fabric adjustment module includes: The total blast furnace combustion energy consumption acquisition unit is used to acquire the cumulative amount of blast furnace combustion energy consumption of raw materials and fuels from the start of combustion to complete combustion in the blast furnace, and to obtain the total blast furnace combustion energy consumption of raw materials and fuels in complete combustion in the blast furnace. The dual-layer optimization framework construction unit is used to construct the dual-layer optimization framework by taking the raw material ratio-blast furnace combustion energy consumption model as the outer framework, the charging matrix-blast furnace combustion energy consumption model as the inner framework, the total blast furnace combustion energy consumption as the minimum threshold, and the production standard, the initial ratio scheme and the initial charging matrix as the production constraints. The optimization unit is used to optimize the outer frame according to the two-layer optimization framework to obtain the optimal proportion scheme, and to optimize the inner frame according to the optimal proportion scheme to obtain the optimal fabric matrix. The adjustment unit is used to adjust the initial proportioning scheme and the initial fabric matrix through the optimal proportioning scheme and the optimal fabric matrix; The closed-loop control module is used to perform closed-loop control of the blast furnace combustion process based on the adjusted initial raw material ratio scheme and the initial charging matrix.

2. The blast furnace multi-parameter coupled optimization energy-saving control system according to claim 1, characterized in that, The raw material and fabric determination module includes: The initial proportioning calculation unit is used to construct an objective function based on the historical combustion data of the blast furnace and the performance of each raw material, and to calculate the initial proportioning scheme based on the objective function with minimum combustion energy consumption as a constraint. The initial material distribution matrix generation unit is used to obtain the original material distribution method based on the historical combustion data of the blast furnace, and generate the initial material distribution method based on the original material distribution method and the initial proportioning scheme. The initial material distribution method is then filled into the two-dimensional material distribution orientation matrix to obtain the initial material distribution matrix.

3. The blast furnace multi-parameter coupled optimization energy-saving control system according to claim 1, characterized in that, The parameter filtering module includes: The combustion energy consumption calculation unit is used to calculate the blast furnace combustion energy consumption at the current moment based on the combustion information of raw materials and fuels and the real-time combustion parameters of each combustion parameter type. The correlation coefficient calculation unit is used to calculate the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the blast furnace combustion energy consumption at the current moment. The key combustion type screening unit is used to screen real-time combustion parameters of several combustion parameter types with correlation coefficients greater than the correlation threshold as key real-time combustion parameters of several key combustion parameter types.

4. The blast furnace multi-parameter coupled optimization energy-saving control system according to claim 3, characterized in that, The correlation coefficient calculation unit includes: The standardization subunit is used to standardize the real-time combustion parameters of each combustion parameter type to obtain the standard combustion parameters corresponding to each combustion parameter type. The point mapping sub-unit is used to construct a two-dimensional combustion point representing the relationship between each standard combustion parameter and the blast furnace combustion energy consumption at the current moment. All two-dimensional combustion points are mapped to a preset standard grid. Each vertical line in the preset standard grid is used to divide the value range of blast furnace combustion energy consumption vertically, and each horizontal line is used to divide the value range of standard combustion parameter values ​​horizontally. Each horizontal and vertical line is configured with a preset probability density after the corresponding value range division. The marginal probability calculation sub-unit is used to take the preset probability density of the horizontal line where the standard combustion parameters of each two-dimensional combustion point in the preset standard grid are located as its combustion marginal probability, and the preset probability density of the vertical line where the blast furnace combustion energy consumption is located at the current moment as its energy consumption marginal probability. The correlation coefficient calculation subunit calculates the correlation coefficient between the real-time combustion parameters of each combustion parameter type and the current blast furnace combustion energy consumption based on the combustion marginal probability and energy consumption marginal probability of the standard combustion parameters of each combustion parameter type.

5. The blast furnace multi-parameter coupled optimization energy-saving control system according to claim 4, characterized in that, The correlation coefficient calculation subunit calculates the correlation coefficient I between the real-time combustion parameter of any combustion parameter type and the current blast furnace combustion energy consumption based on the combustion marginal probability and energy consumption marginal probability of the standard combustion parameter of any combustion parameter type, using formula (1): （1）； In formula (1), X represents the value of the standard combustion parameter of the combustion parameter type, Y represents the value of the blast furnace combustion energy consumption at the current moment, p(X,Y) represents the joint probability of the standard combustion parameter of the combustion parameter type and the blast furnace combustion energy consumption at the current moment, p(X) represents the combustion marginal probability of the standard combustion parameter of the combustion parameter type, p(Y) represents the energy consumption marginal probability of the blast furnace combustion energy consumption at the current moment, log() represents the logarithmic function, and e represents the smoothing coefficient.

6. The blast furnace multi-parameter coupled optimization energy-saving control system according to claim 1, characterized in that, The combustion control module includes: The time-series combustion data acquisition unit is used to acquire the time-series combustion data of each key real-time combustion parameter within a preset time period with the current time as the end point. The time-series combustion data of all key real-time combustion parameters are time-aligned. The intrinsic relationship judgment unit is used to determine the intrinsic relationship between all key real-time combustion parameters based on causal relationships according to each time-series combustion data. A multi-parameter coupled combustion field model framework building unit is used to construct a multi-parameter coupled combustion field model framework based on the intrinsic relationship between all key real-time combustion parameters. The combustion control unit is used to input all key real-time combustion parameters into the multi-parameter coupled combustion field model framework to obtain the energy consumption threshold. If the blast furnace combustion energy consumption is higher than the energy consumption threshold at the current moment, the equipment parameters of each device in the blast furnace are adjusted to reduce the blast furnace combustion energy consumption to below the energy consumption threshold.

7. The blast furnace multi-parameter coupled optimization energy-saving control system according to claim 6, characterized in that, The intrinsic relationship judgment unit includes: The hysteresis coefficient configuration subunit is used to calculate the hysteresis coefficient of each pair of key real-time combustion parameters based on the statistical characteristics of each pair of time-series combustion data. The causal relationship testing subunit is used to construct two causal relationship regression models with mutual relationship based on the lag coefficients of each pair of time-series combustion data and each pair of key real-time combustion parameters. If the statistical probability of the F-test value of any causal relationship regression model of any two key real-time combustion parameters is less than the preset critical value, the causal relationship is determined to be valid, and the intrinsic relationship between the two key real-time combustion parameters is determined. Otherwise, the causal relationship is not valid.

8. The blast furnace multi-parameter coupled optimization energy-saving control system according to claim 1, characterized in that, The optimization unit includes: The weighting sub-unit is used to assign weighting to the data domain of the raw material ratio-blast furnace combustion energy consumption model and to assign weighting to the data domain of the charge matrix-blast furnace combustion energy consumption model. The outer optimization subunit is used to input the initial proportioning scheme and the initial material distribution matrix into the two-layer optimization framework. Based on the total blast furnace combustion energy consumption and the initial proportioning scheme, the outer loss value is calculated through the proportioning weight. The raw material proportioning-blast furnace combustion energy consumption model is superimposed with the outer loss value to obtain a new raw material proportioning-blast furnace combustion energy consumption model. The new raw material proportioning-blast furnace combustion energy consumption model is minimized according to the production constraints to obtain the optimal proportioning scheme. The inner layer optimization subunit is used to calculate the inner layer loss value based on the total blast furnace combustion energy consumption, the initial charging matrix, and the optimal proportion scheme through the charging weight. The inner layer loss value is superimposed on the charging matrix-blast furnace combustion energy consumption model to obtain a new charging matrix-blast furnace combustion energy consumption model. The optimal spatial solution of the new charging matrix-blast furnace combustion energy consumption model is calculated based on the production constraints and the global optimization algorithm. The optimal charging matrix is ​​determined based on the optimal spatial solution.

Images