Control device and control method

Abstract

To realize a technology capable of executing control for properly determining an operation amount of a blast furnace without waiting for an observation result of an observation amount that is observable with a relatively slow time response to an operation or disturbance to the blast furnace.SOLUTION: A state observation amount estimation unit of a control device includes: a representative value calculation unit that calculates a representative value of an observation amount observable with a relatively slow time response in a focused period, from an observation amount observable with a relatively slow time response in the focused period, the focused period being a period across one or a plurality of TAPs when periods with different states in which disturbances are superimposed are distinguished as TAPs according to a tapping cycle of a blast furnace; an observation amount correction unit that corrects the observation amount observable with a relatively slow time response in the focused period such that the representative value and a time-averaged value of the observation amount observable with a relatively slow time response in the focused period coincide with each other; and a state estimation unit that estimates a state of the blast furnace using the corrected observation amount observable with a relatively slow time response in the focused period and a third state space model.SELECTED DRAWING: Figure 7

Description

Technical Field

[0001] The present invention relates to a control device and a control method, and relates to a control device and a control method capable of executing control for appropriately determining the operation amount of a blast furnace without waiting for the observation result of an observable quantity that can be observed with a relatively slow time response to operations and disturbances on the blast furnace.

Background Art

[0002] In recent years' blast furnace processes, the use of inexpensive iron ores and operation under low reductant ratio conditions have been targeted, and the need to stabilize the furnace condition of the blast furnace has been increasing.

[0003] However, operations using inexpensive and low-quality raw materials or low reductants tend to lead to unstable furnace conditions and may cause fluctuations in production volume and a decrease in furnace heat. For example, it is known that various types of disturbances occur due to charged raw materials such as iron ore and coke, causing fluctuations in blast furnace operation. Therefore, in order to achieve stable blast furnace operation, it is necessary to operate considering these disturbances.

[0004] In addition, in blast furnace operation, various operations are performed on the blast furnace, such as material input and condition change. In order to optimize blast furnace operation, changes resulting from the operations are observed as observable quantities, and control is performed to adjust the operation amount of the next operation by feeding back the observable quantities.

[0005] As a technique related to blast furnace operation, a method for predicting the hot metal temperature in a blast furnace using a physical model capable of calculating the state inside the blast furnace has been proposed (see Non-Patent Document 1).

[0006] For example, in Patent Document 1, in view of the fact that in blast furnace operation, the influence of disturbances derived from charged materials is often first measured only after a certain period of time has elapsed, the error between the output variable of the physical model and its measured value in a past predetermined interval is fitted by the response of the output variable when the output variable is changed step by step, so that the error of the output variable can be compensated, thereby improving the prediction accuracy of the hot metal temperature. A technique is disclosed.

[0007] In addition, in Patent Document 2, when predicting the hot metal temperature using a physical model, in view of the fact that the prediction accuracy of the hot metal temperature may decrease due to the influence of disturbances such as the reducibility of iron ore and gas maldistribution that are difficult to model, based on the time change rate of the error in the past interval of output variables other than the hot metal temperature, the difference between the calculated value and the actual value is calculated as the calculation error of the hot metal temperature, and by removing the influence of the prediction error resulting from the change in the manipulated variable of the manipulated variable from the calculation time point, a technique is disclosed for improving the prediction accuracy of the change amount of the hot metal temperature and thus improving the prediction accuracy of the hot metal temperature.

Prior Art Documents

Patent Documents

[0008]

Patent Document 1

Patent Document 2

Non-Patent Documents

[0009]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0010] However, among the observed quantities in which the results of operations and disturbances are reflected and observed, there are observed quantities that are observed at a relatively late timing after the operation (observed quantities that can be observed with a relatively slow time response to operations and disturbances), and observed quantities that are observed at a relatively early timing (observed quantities that can be observed with a relatively fast time response to operations and disturbances). Therefore, even if the amount of operation for the next operation is adjusted by simply feeding back the observed quantity observed at a relatively early timing, it is not possible to perform appropriate control that reflects the observed quantity observed at a relatively late timing. Also, even if the amount of operation for the next operation is adjusted by simply feeding back the observed quantity observed at a relatively late timing, there is a problem that the response is too late and it becomes too late to take action.

[0011] For example, the molten iron temperature, which is an important indicator in operation, is observed at a relatively late timing with respect to the influence of operations and disturbances. For example, even if an operation is performed to bring the molten iron temperature closer to the desired value when it is observed that the molten iron temperature has deviated from the desired value, it will take a long time for the operation to bring the molten iron temperature closer to the desired value as a result. In this case, the state of the blast furnace with a large deviation from the desired temperature will continue for a long time, causing problems in operation.

[0012] In the technology of Patent Document 1, the difference between the observed quantity observed at a relatively late timing and the observed quantity observed at a relatively early timing with respect to the influence of operations and disturbances is not recognized, and the above-described problems in operation cannot be solved.

[0013] Also, the change in the actual value of the molten iron temperature is not caused only by factors that can be explained by the rate of change over time of operation indicators such as gas utilization rate and RAR. For example, it is also caused by disturbance changes affected by various factors such as changes in the heat exchange coefficient between solids and liquids in the lower part of the furnace, changes in the movement of the molten metal flow at the furnace bottom, variations in heat extraction on the hearth, heat extraction in the runner after tapping, and measurement errors during temperature measurement.

[0014] Therefore, even if a regression model is constructed with the rate of change of the operating index as the explanatory variable and the calculation error of the hot metal temperature with various disturbances superimposed thereon as the target variable, as in the method of Patent Document 2, there is a problem that it is difficult to accurately calculate the regression coefficient.

[0015] One aspect of the present invention aims to realize a technique that enables execution of control for appropriately determining the operation amount of a blast furnace without waiting for the observation results of an observable quantity having a relatively slow time response to operations and disturbances to the blast furnace.

Means for Solving the Problems

[0016] To solve the above problems, a control device according to one aspect of the present invention is a control device that controls the process of a blast furnace, and includes an observable quantity that can be observed with a relatively fast time response to operations and disturbances to the blast furnace and a first state space model that estimates the state of the blast furnace corresponding to the observable quantity that can be observed with the relatively fast time response, and a disturbance estimator. After an operation to the blast furnace, a disturbance estimation unit that estimates a disturbance value from the observable quantity observed when a first time has elapsed, a heat change amount calculation unit that calculates a heat change amount accompanying the occurrence of the disturbance from the estimated disturbance value, a temperature change amount calculation unit that calculates a change amount of the hot metal temperature from the heat change amount, a second state space model that estimates the state of the blast furnace corresponding to an observable quantity including an observable quantity having a relatively slow time response to operations and disturbances to the blast furnace and an observable quantity including the observable quantity having the relatively slow time response, a third state space model is generated using the change amount of the hot metal temperature, and using the third state space model, an observable quantity observed when a second time longer than the first time has elapsed after an operation to the blast furnace and a state observation quantity estimation unit that estimates the state of the blast furnace corresponding to the observable quantity, and an operation amount determination unit that determines an operation amount related to the next operation to the blast furnace based on the estimation result of the state observation quantity estimation unit. It has a state observation quantity estimation unit. When distinguishing periods with different superimposed disturbance states as TAPs according to the cycle of the hot metal discharged from the blast furnace, the state observation quantity estimation unit calculates a representative value of the observable quantity with a relatively slow time response in the target period, which is a period spanning one or more TAPs, from the observable quantity that can be observed with a relatively slow time response in the target period; and an observable quantity correction unit that corrects the observable quantity that can be observed with a relatively slow time response in the target period so that the representative value and the time average value of the observable quantity that can be observed with a relatively slow time response in the target period match; and a state estimation unit that estimates the state of the blast furnace using the corrected observable quantity that can be observed with a relatively slow time response in the target period and the third state space model.

[0017] To solve the above problems, a control method according to an aspect of the present invention is a control method of a control device that controls the process of a blast furnace, including an observable quantity that can be observed with a relatively fast time response to operations and disturbances to the blast furnace and a first state space model for estimating the state of the blast furnace corresponding to the observable quantity that can be observed with the relatively fast time response, and an external disturbance estimator. After an operation to the blast furnace, an external disturbance estimation step of estimating an external disturbance value from the observable quantity observed when a first time has elapsed, a heat change amount calculation step of calculating a heat change amount accompanying the occurrence of the external disturbance from the estimated external disturbance value, a temperature change amount calculation step of calculating a change amount of the hot metal temperature from the heat change amount, a second state space model for estimating the state of the blast furnace including an observable quantity that can be observed with a relatively slow time response to operations and disturbances to the blast furnace and an observable quantity that can be observed with the relatively slow time response, generating a third state space model using the change amount of the hot metal temperature, and using the third state space model, a state observable quantity estimation step of estimating an observable quantity observed when a second time longer than the first time has elapsed after an operation to the blast furnace and the state of the blast furnace based on the observable quantity, and an operation amount determination step of determining an operation amount related to the next operation to the blast furnace based on the estimation result of the state observable quantity estimation step. The state observable quantity estimation step includes a representative value calculation step of calculating a representative value of the observable quantity that can be observed with a relatively slow time response in a target period, which is a period spanning one or more TAPs when periods with different superimposed states of external disturbances are distinguished as TAPs according to the tapping cycle from the blast furnace, and an observable quantity correction step of correcting the observable quantity that can be observed with a relatively slow time response in the target period so that the representative value and the time average value of the observable quantity that can be observed with a relatively slow time response in the target period match, and a state estimation step of estimating the state of the blast furnace using the corrected observable quantity that can be observed with a relatively slow time response in the target period and the third state space model.

Effect of the Invention

[0018] According to one aspect of the present invention, it is possible to execute control for appropriately determining the operation amount of the blast furnace without waiting for the observation result of an observable quantity that has a relatively slow time response to the operation and disturbance of the blast furnace.

Brief Description of the Drawings

[0019]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Embodiments for Carrying Out the Invention

[0020] Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the drawings. First, the outline of the blast furnace process will be described.

[0021] Fig. 1 is a diagram for explaining the outline of the blast furnace process. In the blast furnace process, as shown in Fig. 1, sintered ore and coke are charged from the top of the blast furnace 101 so as to create alternating layers, and hot air and pulverized coal as a reducing agent are blown in from the blast tuyeres at the bottom of the furnace. The hot air gasifies the pulverized coal and coke, and high-temperature reducing gases such as carbon monoxide and hydrogen rise up inside the furnace, melting the sintered ore while removing oxygen, and the molten iron comes into contact with the carbon in the coke and is reduced, becoming molten iron containing just under 5% carbon, which accumulates in the basin at the bottom of the furnace.

[0022] The molten iron is taken out from a tap hole installed at the side of the hearth and transported to the next steelmaking process. Gases such as coke oven gas (COG) or natural gas (NG) may be blown into the furnace through the blast tuyeres (or other tuyeres).

[0023] In addition, in blast furnace operation, operation control indexes are set, the time transition of the operation control indexes is monitored, and operation operations for the blast furnace process are performed with the corresponding operation amount to control the furnace condition. Therefore, setting appropriate operation control indexes is extremely important for stable operation of the blast furnace. Details of the operation control indexes will be described later, but the operation control indexes include indexes that should be kept at set target values ​​(in this embodiment, control amount) and indexes that are mainly used to grasp changes in the furnace condition (in this embodiment, observation amount).

[0024] Figure 2 is a diagram showing the relationship between the blast furnace process 1 and the manipulated variable u and the observed variable y. As shown in Figure 2, the blast furnace process 1 inputs the manipulated variable u and outputs the observed variable y. The state of the blast furnace process 1 is represented by x.

[0025] The operation amount u is a value related to the air blowing amount, the amount of pulverized coal (the amount of pulverized coal injection PCI or the pulverized coal ratio PCR), or the like.

[0026] The observed quantity y includes the hot metal temperature, gas composition (e.g., CO concentration, CO2 concentration, N2 concentration, H2 concentration), gas utilization rate (e.g., CO utilization rate, H2 utilization rate), amount of carbon solution loss (CSL), CO utilization rate, H2 utilization rate, tapping amount, hot metal Si, hot metal temperature, etc. Among these observed quantities y, the hot metal temperature, hot metal Si, and tapping amount are often regarded as controlled quantities with set target values.

[0027] In recent blast furnace processes, the use of inexpensive iron ores and operation under conditions of low reductant ratio are being pursued, and the need to stabilize the blast furnace condition is increasing.

[0028] However, operations using inexpensive and low-quality raw materials or low reductants tend to make the furnace condition unstable and may cause fluctuations in production volume and a decrease in furnace heat. For example, it is known that various types of disturbances occur due to charged raw materials such as iron ore and coke, causing fluctuations in blast furnace operation. Therefore, in order to achieve stable blast furnace operation, it is necessary to operate considering these disturbances.

[0029] In blast furnace operation, various operations are performed on the blast furnace, such as material input and condition changes. In order to optimize blast furnace operation, changes resulting from the operations are observed as observed quantities, and control is performed to adjust the operation amount of the next operation by feeding back the observed quantities.

[0030] Figure 3 is a diagram for explaining the outline of operation input in blast furnace operation. As shown in Figure 3, in process control, disturbance and state estimation are performed, and the estimated value d^ of the disturbance and the estimated value x^ of the blast furnace state are calculated. Here, the notations x^ and d^ are assumed to have ^ (hat) attached above x and d respectively. For characters with a hat attached, the same notation will be used hereafter.

[0031] Based on the calculated estimated value d^ of the disturbance and the estimated value x^ of the blast furnace state, the operation amount u is determined and operation input is performed.

[0032] The operations corresponding to the operation amount u affect the blast furnace process. The blast furnace process includes a fast reaction P1 and a slow reaction P2 with respect to operations and disturbances. The fast reaction P1 means that the time taken to obtain a response to an operation or disturbance is relatively short, and the slow reaction P2 means that the time taken to obtain a response to an operation or disturbance is relatively long. Due to the fast reaction P1, for example, the gas utilization rate, CSL, etc. change. Also, due to the slow reaction P2, for example, the hot metal temperature changes. Note that, although details will be described later, the fast reaction P1 corresponds to chemical reaction processes such as the reduction reaction of ore and the consumption reaction of coke, and the slow reaction P2 roughly corresponds to the heat transfer and heat conduction phenomena of the reaction heat generated by the fast reaction, respectively.

[0033] The values of indicators such as the gas utilization rate, CSL, and hot metal temperature changed by the fast reaction P1 and the slow reaction P2 are fed back to the process control as observed quantities. That is, disturbance and state estimation are performed based on the observed quantities, and the estimated value d^ of the disturbance and the estimated value x^ of the state of the blast furnace are calculated again. Then, a new operation amount u is determined based on the calculated estimated value d^ of the disturbance and the estimated value x^ of the state of the blast furnace, and an operation input is performed.

[0034] As can be seen from Figure 3, the observed quantities observed as a result of the operation include the observed quantities related to the indicators that change due to the fast reaction P1 with respect to the operation and disturbance, and the observed quantities related to the indicators that change due to the slow reaction P2 with respect to the operation and disturbance. That is, among the observed quantities fed back to the process control, the indicators changed by the slow reaction P2 are actually the indicators that changed in response to the operation and disturbance input quite a while ago in terms of time, and the indicators changed by the fast reaction P1 are actually the indicators that changed in response to the operation and disturbance input a little while ago in terms of time.

[0035] Observation quantities related to indicators (e.g., gas utilization rate, CSL) that change due to a quick response P1 to operations or disturbances can be observed at a relatively short time after the operation or the occurrence of the disturbance, and such observation quantities can be said to be observation quantities that can be observed with a relatively quick time response to operations or disturbances. On the other hand, observation quantities related to indicators (e.g., hot metal temperature) that change due to a slow response P2 to operations or disturbances can be observed at a relatively long time after the operation or the occurrence of the disturbance, and such observation quantities can be said to be observation quantities that can be observed with a relatively slow time response to operations or disturbances.

[0036] Therefore, if the observation quantity observed with a relatively slow time response to operations or disturbances is fed back and reflected in the operation quantity for the next time, it is impossible to realize a prompt control operation.

[0037] For example, the change in hot metal temperature, which is an important indicator in operation, is observed at a relatively slow timing. Even if an operation is performed to bring the hot metal temperature closer to the desired value (target temperature) when it is observed that the hot metal temperature has deviated from the desired value, it will take a long time for the hot metal temperature to approach the target temperature as a result of that operation. In this case, the state of the blast furnace with a large deviation from the target temperature will continue for a long time.

[0038] <Embodiment> (State space model of blast furnace) In this embodiment, the state of the blast furnace is obtained using a parameter that relates an observation quantity that can be observed with a relatively quick time response to operations or disturbances and an observation quantity that can be observed with a relatively slow time response to operations or disturbances. That is, the state of the blast furnace can be quickly obtained without waiting for the observation result of the observation quantity that can be observed with a relatively slow time response to operations or disturbances, and the observation quantity corresponding to the state of the blast furnace can also be obtained.

[0039] In this embodiment, a discrete-time linear state-space model represented by formulas (1) and (2) for expressing blast furnace operation is used for the control of the blast furnace process 1. The discrete-time linear state-space model includes an equation representing the discrete-time change in which the state (state variable) of the blast furnace process 1 changes from x0(k) to x0(k + 1) when the operation amount u(k) is applied at the current time k.

[0040] Here, x0(k) represents the state of the blast furnace and is the state of the factors in the blast furnace that affect the observed quantity, and various reaction heats, pressures, etc. are included in the state of the blast furnace. y0(k) represents the observed quantity of the general blast furnace, including the observed quantity that can be observed with a relatively slow time response (after a long time has passed) to the operation of the blast furnace with respect to the operation and disturbance. u(k) represents the operation amount to the blast furnace, and k is an index indicating the time minute.

[0041] Examples of the operation amount u(k) include the blast volume and the amount of pulverized coal injection.

[0042] Examples of y0(k) include the Si in hot metal and the hot metal temperature.

[0043] Also, the coefficients A0, B0, C0, and D0 are assumed to be determined in advance.

[0044]

Equation

[0045]

Equation

[0046] Also, it is assumed that the values of the coefficients A, B, C, and D are specified in advance.

[0047] Since the factors and observed quantities of the blast furnace state are different between the equations (1) and (2) and the equations (3) and (4), subscripts are attached to the variables and constants of the state space model to distinguish between the two. On the other hand, since the manipulated variable u(k) is common to the equations (1) and (2) and the equations (3) and (4), it is not distinguished by subscripts.

[0048] Note that the equation (1) and the equation (3) may be the same equation or different equations.

[0049]

Number

[0050]

Number

[0051] That is, due to the influence of various disturbances that change the blast furnace operation, such as changes in the properties of the raw materials charged from the top of the furnace, the calculated values of the observed quantities deviate from the measured values. For example, due to the occurrence of a disturbance, the main chemical reaction amounts in the furnace change unexpectedly, resulting in a deviation. Examples of the main chemical reaction amounts in the furnace include chemical reaction amounts such as the indirect reduction reaction amount, direct reduction reaction amount, hydrogen reduction reaction amount, coke consumption amount, and coke combustion reaction amount.

[0052] Therefore, in order to calculate the state and observed quantities of the blast furnace more accurately, it is necessary to feedback the actual measured values in the blast furnace affected by the disturbance to the state space model as shown in the equations (1) to (4) and reflect them in the calculated values of the state space model.

[0053] At this time, if feedback is attempted after waiting for an observable quantity that can be observed with a relatively slow time response to operations and disturbances to be actually observed, as described above, the state of the blast furnace with a large deviation from the target temperature will continue for a long time. Therefore, as the observable quantity for feedback, it is desirable to use an observable quantity that can be observed with a relatively fast time response to operations and disturbances.

[0054] Here, consider the case where disturbances are taken into account for expressions (3) and (4).

[0055] When considering disturbances, a known disturbance estimator observer for calculating the disturbance estimate d^(k) based on expressions (3) and (4) is applied. Thereby, expressions (5), (6), and (7) are obtained.

[0056]

Equation

[0057]

Equation

[0058]

Equation

[0059] Note that the configuration procedure and design method of the observer in the discrete-time system are described in Non-Patent Document 1 and the like.

[0060] By using the value of the actually observed observed quantity y(k) and solving equations (5) to (7), the value of the disturbance estimated value d^(k) can be obtained. Here, as the value of the observed quantity y(k), for example, observed quantities that can be observed with a relatively fast time response to operations and disturbances such as gas composition, CSL, CO utilization rate, H2 utilization rate, and tapping amount can be actually observed and used.

[0061] Then, by using the state space model (equations (3) and (4)) for estimating the observed quantity that can be observed with a relatively fast time response to operations and disturbances and the disturbance estimator, the disturbance value is estimated from the observed quantity observed at a point in time after a relatively short time has elapsed after the operation on the blast furnace.

[0062] As described above, since the deviation between the calculated value and the measured value is considered to be due to the disturbance, as shown in equation (6), it becomes possible to obtain an observed quantity close to the measured value by using this disturbance estimated value d^(k). However, this disturbance estimated value d^(k) is obtained from the observed quantity that can be observed with a relatively fast time response to operations and disturbances. For this reason, even if an attempt is made to appropriately reflect the disturbance estimated value d^(k) in the state space model (equations (1) and (2)) including the observed quantity that can be observed with a relatively slow time response to operations and disturbances after the operation, it is difficult because the calculation procedure is not obvious.

[0063] Therefore, in accordance with the calculation procedure described below, a parameter for relating the observed quantity that can be observed with a relatively fast time response to operations and disturbances and the observed quantity that can be observed with a relatively slow time response to operations and disturbances was introduced.

[0064] As described above, the disturbance corresponding to the disturbance estimated value d^(k) is considered to correspond to the change in the chemical reaction resulting from changes in the properties of the raw materials charged from the top of the blast furnace. That is, due to disturbance factors such as changes in the properties of the raw materials charged from the top of the blast furnace, changes in the amount of indirect reduction reaction, direct reduction reaction, hydrogen reduction reaction, coke consumption, coke combustion reaction, etc. in the furnace occur.

[0065] Since such changes in the amount of chemical reaction occur relatively rapidly, they are considered to commonly affect both the observable quantity that can be observed with a relatively fast time response to operations and disturbances, and the observable quantity that can be observed with a relatively slow time response to operations and disturbances.

[0066] Therefore, the changes in such chemical reactions are further examined.

[0067] Regarding the chemical reaction factors occurring inside the blast furnace, particularly, let the change amount of the chemical reaction factors via the gas flow be Δs. During the stable period of the furnace condition, since the process fluctuations are relatively small, the disturbance estimated value d^(k) can be linearly represented as in Equation (8).

[0068]

Equation

[0069] Also, since the change in the chemical reaction factors in the furnace is accompanied by a change in heat, based on whether each chemical reaction is an exothermic reaction or an endothermic reaction, the relationship of Equation (9) can be obtained from the relationship between each chemical reaction factor and the generated heat quantity.

[0070]

Equation

[0071] Note that the value of the reaction heat coefficient vector h in Equation (9) can be specifically calculated and obtained by using, for example, a model (blast furnace mathematical model) capable of calculating the time change of chemical reactions in the blast furnace, as described above. That is, at the operating state (operating point) serving as a reference for linear approximation, by giving changes in various boundary condition setting values that can be set in the model and correspond to disturbances around that operating point, each chemical reaction factor in the furnace is perturbed, and by calculating the perturbation amount from the operating point, the value of the reaction heat coefficient vector h in Equation (9) can be calculated.

[0072] Furthermore, by substituting Equation (8) into Equation (9), Equations (10) and (11) are obtained.

[0073]

Equation

[0074]

Equation

[0075] By doing so, the amount of heat change ΔQ accompanying the occurrence of the disturbance can be calculated from the disturbance estimated value d^(k).

[0076] The relationship between the amount of heat change ΔQ and the temperature change amount ΔT of the hot metal temperature can be shown as in Equation (12) from the nominal value of the heat exchange coefficient between the solid and the liquid and the amount of hot metal tapped in the blast furnace interior (for example, the cohesive zone).

[0077] [Number] (12) Here, β is the conversion coefficient between the amount of heat change and the amount of temperature change, and can be calculated from literature values known in the field of chemical engineering or calculations using a blast furnace mathematical model.

[0078] In this way, the change amount of the hot metal temperature can be estimated from the amount of heat change.

[0079] However, in reality, when heat is generated by chemical reactions in the blast furnace, the hot metal temperature does not immediately change. Instead, time delays (referred to as waste time), such as the time for the charged materials to descend in the furnace and the residence time of the hot metal at the lower part of the furnace, occur. Furthermore, due to the large heat capacity in the blast furnace, there is a time delay in transferring heat to the hot metal temperature. As a result, it takes a long time for the hot metal temperature to change.

[0080] Therefore, considering the waste time and the time constant of the heat transfer phenomenon, the time change of ΔT is expressed as in Equation (13) as the dynamics of a first-order lag system including the waste time. In Equation (13), after expressing it as a discrete-time model discretized in time with an appropriate calculation period, the time delay of the influence of each heat of reaction can be set individually.

[0081] [Number] (13) Here, hj and Δsj are the elements of the heat of reaction coefficient vector h and the change amount vector Δs of the chemical reaction factor in the furnace, respectively, and ns is the number of elements. k lag,j indicates the waste time set for each element of the vector. α is a coefficient determined according to the time constant and is assumed to be specified in advance.

[0082] Also, Equation (13) can be transformed into Equation (14) using Equations (8) to (11).

[0083] [Number] (14) Here, \(v_i\) and \(d_i\) are the elements of the vectors \(v\) in Equation (10) and the disturbance estimate \(\hat{d}^{(k)}\), respectively, and \(n_d\) is the number of elements. \(k'_{lag,i}\) is the dead time set for each element of the vector.

[0084] That is, the disturbance estimate used to calculate the change in the hot metal temperature is calculated using the observed values observed \(k'_{lag,i}\) time units in the past. At this time, as the observed values observed in the past, observed values that can be observed with a relatively fast time response to operations and disturbances are used. This is because the disturbances affecting the change in the hot metal temperature \(\Delta T\) are considered to commonly affect both the observed values that can be observed with a relatively fast time response to operations and disturbances and the observed values that can be observed with a relatively slow time response to operations and disturbances.

[0085] By doing so, it becomes possible to calculate the change in the hot metal temperature \(\Delta T\) from the thermal change amount accompanying the occurrence of the disturbance. At this time, the change in the hot metal temperature is obtained from the observed values that can be observed with a relatively fast time response to operations and disturbances, and the obtained change in the hot metal temperature can be applied to the state of the blast furnace including the observed values that can be observed with a relatively slow time response to operations and disturbances. That is, the change in the hot metal temperature functions as a parameter that relates the observed values that can be observed with a relatively fast time response to operations and disturbances and the observed values that can be observed with a relatively slow time response to operations and disturbances.

[0086] By adding the influence of the disturbance as used in Equations (5) to (7) and the change in the hot metal temperature \(\Delta T\) to Equations (1) and (2) that represent the state of the blast furnace including the observed values that can be observed with a relatively slow time response to operations and disturbances, the state space model of the blast furnace can be expressed as in Equations (15) to (17).

[0087]

Equation

[0088]

Equation

[0089]

Number

[0090] In this way, the state of the entire blast furnace can be quickly known from the observable quantity that can be observed at a relatively early timing and the temperature change, without waiting for the observation result of the observable quantity that can be observed at a relatively late timing. As a result, it becomes possible to accurately calculate observable quantities that change relatively slowly, such as hot metal temperature and hot metal Si, from the state space model of the blast furnace.

[0091] By the way, disturbances caused by factors different from the actual furnace heat (for example, dripping hot metal) may be superimposed on the temperature of the hot metal tapped from the blast furnace. For example, immediately after the start of tapping, a situation may continue where a significantly low temperature is measured due to heat extraction by the trough used when flowing out the hot metal from the tapping hole opened at the lower part of the blast furnace. In addition, there may be cases due to factors other than heat extraction in the trough. As shown in FIG. 4, for example, due to the influence of the molten metal flow in the lower part of the furnace, the temperature may rise with a right shoulder upward as time elapses from the start of tapping. In FIG. 4, the cycle in which the hot metal temperature rises with a right shoulder upward every periodic time unit is repeated because the tapping hole for tapping the hot metal is switched every time a certain period elapses. That is, the state in which disturbances are superimposed is different for each time (each cycle) from the start of hot metal tapping from one tapping hole until the start of hot metal tapping from the next tapping hole. Hereinafter, the periods in which the state of disturbance superposition is different according to the tapping cycle of the blast furnace will be distinguished and described as TAP.

[0092] In such a situation, the measured molten iron temperature may be lower or higher than the actual furnace heat depending on the timing of the temperature measurement. And if such measured values are directly reflected in the above formulas (15) to (17), the calculated value of the molten iron temperature may be lower or higher than the actual furnace heat.

[0093] Therefore, by going through the following procedures 1 and 2, the measured values are reflected in the above formulas (15) to (17).

[0094] (Procedure 1) Elimination of disturbance In Procedure 1, a plurality of measured values obtained as a result of measuring the temperature multiple times at one TAP are averaged. Thereby, the disturbance superimposed at one TAP can be eliminated. By implementing this Procedure 1 for each TAP, the representative measured value at each TAP can be correctly extracted.

[0095] Note that due to the influence of the complex molten metal flow at the lower part of the furnace, disturbances may be superimposed across a plurality of TAPs. Therefore, in Procedure 1, the plurality of measured values obtained across a plurality (about 2 to 3) of TAPs may be time-averaged. Thereby, the variation in the disturbances between TAPs superimposed across a plurality of TAPs can also be eliminated.

[0096] (Procedure 2) Reflection of representative measured value The representative measured value obtained through Procedure 1 cannot be directly reflected in formulas (15) to (17). This is because the representative measured value is a time-averaged value, whereas for formulas (15) to (17), the actual measured value of the molten iron temperature at each moment needs to be reflected each time. Therefore, in Procedure 2, the same constant is added to or subtracted from each estimated value so that the representative value calculated in Procedure 1 and the time-averaged value of the estimated values of a plurality of observed quantities obtained by applying the representative value and the plurality of measured values used for calculating the representative value to formulas (15) to (17) match. And the obtained value is used as an observable quantity that can be observed with a relatively slow time response when the second time has elapsed.

[0097] By doing so, an estimated value of the observed quantity from which the influence of the time-varying disturbance has been removed can be obtained.

[0098] (Configuration of the control device) FIG. 6 is a block diagram showing a configuration example of a control device for controlling a blast furnace process. The control device 200 shown in FIG. 6 is a device that observes operation indices and controls the blast furnace process. As shown in FIG. 6, the control device 200 includes an estimation unit 201 and an operation control unit 202.

[0099] The estimation unit 201 estimates the observed quantity to be observed in the future and the state of the blast furnace corresponding to the observed quantity based on various operation indices, and outputs an estimated value. Information indicating the operation amount related to the operation on the blast furnace is supplied to the estimation unit 201 from an operation unit (not shown).

[0100] The operation control unit 202 controls the determination of the operation amount related to the next operation on the blast furnace so as to optimize the operation of the blast furnace based on the estimation result of the estimation unit 201. For example, when it is estimated that the hot metal temperature rises above the target temperature, the determination of the operation amount related to the next operation is controlled so that the hot metal temperature decreases by the next operation. The operation amount is, for example, the blast volume, the amount of pulverized coal injected, etc.

[0101] Also, as shown in FIG. 6, the estimation unit 201 includes a disturbance estimation unit 211, a heat change amount calculation unit 212, a temperature change amount calculation unit 213, and a state observation quantity estimation unit 214.

[0102] The disturbance estimation unit 211 uses a state space model (Equations (3) and (4)) for estimating the observed quantity observable with a relatively fast time response to the operation on the blast furnace and the disturbance and the state of the blast furnace corresponding to the observed quantity observable with a relatively fast time response, and a disturbance estimator, and estimates the disturbance value from the observed quantity observed when the first time has elapsed after the operation on the blast furnace.

[0103] Hereinafter, the state space model represented by Expression (3) and Expression (4) will be appropriately referred to as the first state space model. Observable quantities that can be observed with a relatively fast time response to operations and disturbances on the blast furnace according to the first state space model include gas composition, carbon solution loss amount (CSL), CO utilization rate, H2 utilization rate, hot metal output, and the like.

[0104] That is, the disturbance estimation unit 211 obtains a disturbance estimation value d^(k) by solving the above-described Expressions (5) to (7) using the operation amount u(k) related to the input operation and the value of the actually measured observable quantity y(k). Then, a linearized expression as shown in the above-described Expression (8) is generated using the obtained disturbance estimation value d^(k).

[0105] The heat change amount calculation unit 212 calculates the heat change amount associated with the occurrence of the disturbance from the disturbance value estimated by the disturbance estimation unit 211. That is, the heat change amount calculation unit 212 calculates the heat change amount ΔQ by Expression (10) using the expression (8) generated by the disturbance estimation unit 211.

[0106] The temperature change amount calculation unit 213 calculates the change amount of the hot metal temperature from the heat change amount calculated by the heat change amount calculation unit 212. That is, the temperature change amount calculation unit 213 calculates a vector v representing the relationship between the disturbance estimation value d^(k) shown in Expression (11) and the heat change amount associated with the occurrence of the disturbance, and calculates the change amount ΔT(k) of the hot metal temperature as shown in Expression (14).

[0107] The state observable quantity estimation unit 214 generates another state space model (Expressions (19), (20), and (21)) using the state space model (Expressions (1) and (2)) that estimates the state of the blast furnace corresponding to the observable quantity including the observable quantity that can be observed with a relatively slow time response to operations and disturbances on the blast furnace and the observable quantity that can be observed with a relatively slow time response, and the change amount ΔT(k) of the hot metal temperature calculated by the temperature change amount calculation unit 213. Using this state space model, the observable quantity observed when a second time longer than the first time has elapsed after the operation on the blast furnace and the state of the blast furnace corresponding to the observable quantity are estimated.

[0108] Note that the state - space model represented by Expression (1) and Expression (2) will be appropriately referred to as the second state - space model. Observable quantities with a relatively slow time response to operations and disturbances related to the second state - space model are, for example, hot metal Si, hot metal temperature, etc.

[0109] That is, the state - observation quantity estimation unit 214 adds the influence of disturbances as used in Expressions (5) to (7) and the change amount ΔT of the hot metal temperature to Expressions (1) and (2) representing the state of the blast furnace that also includes observable quantities with a relatively slow time response to operations and disturbances, thereby generating a state - space model represented by Expressions (15) to (17). Further, the state - observation quantity estimation unit 214 enables the calculation of the estimated values of the observable quantities at future prediction times kp = 1, 2, …, Hp in the state - space model represented by Expressions (15) to (17). That is, a state - space model represented by Expressions (19), (20), and (21) is generated.

[0110] Note that the state - space model represented by Expressions (19), (20), and (21) will be appropriately referred to as the third state - space model.

[0111] Then, the observable quantity (hot metal temperature) to be observed in the future and the state of the blast furnace corresponding to the observable quantity are estimated by this third state - space model. That is, as the observable quantity y0^(k + kp) represented by the above - mentioned Expression (21), the estimated value of the hot metal temperature at time k + kp is calculated. Also, the estimated value of the state x0^(k + kp) of the blast furnace represented by the above - mentioned Expression (19) is calculated.

[0112] The state - observation quantity estimation unit 214 according to this embodiment includes a representative - value calculation unit 214a, an observable - quantity correction unit 214b, and a state - estimation unit 214c.

[0113] The representative value calculation unit 214a calculates a representative value of the observable quantity observable with a relatively slow time response in the target period, which is a period spanning one or more TAPs. The representative value calculation unit 214a according to the present embodiment calculates any one of the average value, the trimmed average value, or the median value of the observable quantity observable with a relatively slow time response in the target period as the representative value.

[0114] The observable quantity correction unit 214b corrects the observable quantity observable with a relatively slow time response in the target period so that the representative value calculated by the representative value calculation unit 214a and the time average value of the observable quantity observable with a relatively slow time response in the target period match. The observable quantity correction unit 214b according to the present embodiment adds or subtracts the same constant to each value of the estimated value of the observable quantity in the target period so that the representative value and the time average value of the estimated value of the observable quantity obtained by applying the observable quantity observable with a relatively slow time response in the target period to the third state space model match. Then, the observable quantity correction unit 214b corrects the value obtained by adding or subtracting the constant to be the observable quantity observable with a relatively slow time response observed when the second time has elapsed.

[0115] The state estimation unit 214c estimates the state of the blast furnace using the observable quantity observable with a relatively slow time response in the target period corrected by the observable quantity correction unit 214b and the third state space model.

[0116] The estimation unit 201 supplies the estimation result of the state observable quantity estimation unit 214 to the operation control unit 202.

[0117] Note that the state observable quantity estimation unit 214 may be included in the operation control unit 202. In this case, the estimation unit 201 may supply the change amount ΔT(k) of the hot metal temperature calculated by the temperature change amount calculation unit 213 to the operation control unit 202.

[0118] (Flow of control processing) Next, the flow of the control process by the control device 200 will be described. FIG. 7 is a flowchart for explaining an example of the control process.

[0119] In step S101, the disturbance estimation unit 211 uses a first state space model that estimates an observable quantity that can be observed with a relatively fast time response to operations and disturbances on the blast furnace and the state of the blast furnace corresponding to the observable quantity, and a disturbance estimator, and estimates a disturbance value from the observable quantity observed when a first time has elapsed after the operation on the blast furnace.

[0120] In step S102, the temperature change amount calculation unit 213 calculates a heat change amount associated with the occurrence of a disturbance from the disturbance value estimated in the process of step S101.

[0121] In step S103, the temperature change amount calculation unit 213 estimates a change amount of the hot metal temperature from the heat change amount calculated in the process of step S102.

[0122] In step S104, after the operation on the blast furnace, an observable quantity observed when a second time longer than the first time has elapsed and the state of the blast furnace based on the observable quantity are estimated.

[0123] Specifically, in step S104a, the state observable quantity estimation unit 214 generates a third state space model using a second state space model that estimates an observable quantity including an observable quantity that can be observed with a relatively slow time response to operations and disturbances on the blast furnace and the state of the blast furnace corresponding to the observable quantity, and the change amount of the hot metal temperature calculated in the process of step S103.

[0124] Then, in step S104b, the state observable quantity estimation unit 214 calculates a representative value of the observable quantity that can be observed with a relatively slow time response in the target period, which is a period spanning one or more TAPs, from the observable quantity that can be observed with a relatively slow time response in the target period. In step S104b according to this embodiment, the state observation quantity estimation unit 214 calculates, as a representative value, any one of the average value, trimmed average value, or median value of the observation quantities observable with a relatively slow time response during the target period.

[0125] Then, in step S104c, the state observation quantity estimation unit 214 corrects the observation quantity observable with a relatively slow time response during the target period so that the representative value matches the time average value of the observation quantity observable with a relatively slow time response during the target period. In step S104c according to this embodiment, the state observation quantity estimation unit 214 adds or subtracts the same constant to each value of the estimated value of the observation quantity during the target period so that the representative value matches the time average value of the estimated value of the observation quantity during the target period obtained by applying the observation quantity observable with a relatively slow time response during the target period to the third state space model. Then, the state observation quantity estimation unit 214 performs correction such that the obtained value is the observation quantity observable with a relatively slow time response observed when the second time has elapsed.

[0126] Then, in step S104d, using the corrected observation quantity observable with a relatively slow time response during the target period and the third state space model, the state of the blast furnace observed when the second time has elapsed is estimated.

[0127] In step S105, the operation control unit 202 determines the operation amount related to the next operation so as to optimize the operation of the blast furnace based on the estimation result of the process in step S104.

[0128] In this way, the control process is executed.

[0129] (Effect of the embodiment) As described above, according to the present embodiment, the amount of change in the molten iron temperature is obtained from an observable quantity that can be observed with a relatively fast time response to operations and disturbances, and the obtained amount of change in the molten iron temperature is applied to the state of the blast furnace including observable quantities that can be observed with a relatively slow time response to operations and disturbances. Therefore, when an observable quantity that can be observed with a relatively fast time response to operations and disturbances is observed, it becomes possible to estimate an observable quantity that can be observed with a relatively slow time response to operations and disturbances.

[0130] As a result, according to the present embodiment, by feeding back the change in an observable quantity (for example, gas utilization rate, CSL) observed at a relatively early timing, it becomes possible to determine an operation amount so that an observable quantity (for example, molten iron temperature) observed at a relatively late timing approaches a target value. That is, it becomes possible to execute control for appropriately determining the operation amount of the blast furnace without waiting for the observation result of an observable quantity that can be observed at a relatively late timing.

[0131] <Example of Realization by Software> The functions of the control device 200 can be realized by a program for causing a computer to function as the device, and by programs for causing a computer to function as each block of the device.

[0132] FIG. 8 is a block diagram illustrating the physical configuration of a computer used as the control device 200. As shown in FIG. 8, the control device 200 can be configured by a computer including a bus 510, a processor 501, a main memory 502, an auxiliary memory 503, a communication interface 504, and an input / output interface 505. The processor 501, the main memory 502, the auxiliary memory 503, the communication interface 504, and the input / output interface 505 are connected to each other via the bus 510. An input device 506 and an output device 507 are connected to the input / output interface 505.

[0133] As the processor 501, for example, a CPU (Central Processing Unit), a microprocessor, a digital signal processor, a microcontroller, or a combination thereof, etc. are used.

[0134] As the main memory 502, for example, a semiconductor RAM (random access memory), etc. are used.

[0135] As the auxiliary memory 503, for example, a flash memory, an HDD (Hard Disk Drive), an SSD (Solid State Drive), or a combination thereof, etc. are used. In the auxiliary memory 503, a program for causing the processor 501 to execute the operations of the above-described control device 200 is stored. The processor 501 expands the program stored in the auxiliary memory 503 onto the main memory 502 and executes each instruction included in the expanded program.

[0136] The communication interface 504 is an interface for connecting to a network.

[0137] As the input / output interface 505, for example, a USB interface, a short-range communication interface such as infrared or Bluetooth (registered trademark), or a combination thereof is used.

[0138] As the input device 506, for example, a keyboard, a mouse, a touch pad, a microphone, or a combination thereof, etc. are used. As the output device 507, for example, a display, a printer, a speaker, or a combination thereof is used.

[0139] When the functions of the control device 200 are realized by a program for causing a computer to function as the device, by executing the above program with the processor 501 and the main memory 502, each function described in the above embodiments is realized.

[0140] The above program may be recorded on one or more computer-readable recording media, rather than temporarily. This recording medium may or may not be provided in the above device. In the latter case, the above program may be supplied to the above device via any wired or wireless transmission medium.

[0141] In addition, part or all of the functions of each of the above blocks can also be realized by a logic circuit. For example, an integrated circuit in which a logic circuit functioning as each of the above control blocks is formed is also included in the scope of the present invention.

[0142] Also, part or all of the functions of each of the above blocks may operate in the above device, or may operate in another device (for example, an edge computer or a cloud server, etc.).

[0143] Note that the present invention is not limited to the above-described embodiments, and various modifications are possible within the scope shown in the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention.

Explanation of Reference Numerals

[0144] 1 Blast furnace process 200 Control device 201 Estimation unit 202 Operation control unit 211 Disturbance estimation unit 212 Heat change amount calculation unit 213 Temperature change amount calculation unit 214 State observation amount estimation unit 214a Representative value calculation unit 214b Observation amount correction unit 214c State estimation unit

Claims

1. A control device for controlling the process of a blast furnace, comprising: a first state space model for estimating an observable quantity observable with a relatively fast time response to operations and disturbances in the blast furnace and the state of the blast furnace corresponding to the observable quantity observable with the relatively fast time response, and a disturbance estimator, and a disturbance estimation unit for estimating a disturbance value from the observable quantity observed when a first time has elapsed after an operation on the blast furnace; a heat change amount calculation unit for calculating a heat change amount accompanying the occurrence of a disturbance from the estimated disturbance value; a temperature change amount calculation unit for calculating a change amount of the hot metal temperature from the heat change amount; a second state space model for estimating an observable quantity including an observable quantity observable with a relatively slow time response to operations and disturbances in the blast furnace and the state of the blast furnace corresponding to the observable quantity including the observable quantity observable with the relatively slow time response, and generating a third state space model using the change amount of the hot metal temperature; a state observable quantity estimation unit for estimating an observable quantity observed when a second time longer than the first time has elapsed after an operation on the blast furnace and the state of the blast furnace corresponding to the observable quantity using the third state space model; an operation amount determination unit for determining an operation amount related to the next operation on the blast furnace based on the estimation result of the state observable quantity estimation unit; and comprising: wherein the state observable quantity estimation unit: when distinguishing a period TA in which the state of superposition of disturbances is different according to the tapping cycle from the blast furnace, a representative value calculation unit for calculating a representative value of the observable quantity observable with a relatively slow time response in the target period, which is a period spanning one or more TAPs, from the observable quantity observable with a relatively slow time response in the target period; an observable quantity correction unit for correcting the observable quantity observable with a relatively slow time response in the target period so that the representative value and the time average value of the observable quantity observable with a relatively slow time response in the target period match; and a state estimation unit for estimating the state of the blast furnace using the corrected observable quantity observable with a relatively slow time response in the target period and the third state space model; A control device having the above components.

2. The representative value calculation unit: The control device according to claim 1, wherein any one of an average value, a trimmed average value, or a median value of the observable quantity observable with a relatively slow time response in the target period is used as the representative value.

3. The observable quantity correction unit: The time average value of the estimated values of the observed quantities in the target period obtained by applying the representative value and the observed quantities observable with a relatively slow time response in the target period to the third state space model, and adding or subtracting the same constant to each value of the estimated values of the observed quantities in the target period so that they match, The control device according to claim 1 or 2, wherein correction is performed such that the obtained value is an observed quantity observable with a relatively slow time response observed when the second time has elapsed.

4. A control method for a control device that controls the process of a blast furnace, Using a first state space model that estimates the state of the blast furnace corresponding to the observed quantities observable with a relatively fast time response to operations and disturbances to the blast furnace and the observed quantities observable with the relatively fast time response, and a disturbance estimator, a disturbance estimation step of estimating a disturbance value from the observed quantities observed when the first time has elapsed after an operation to the blast furnace, A heat change amount calculation step of calculating a heat change amount accompanying the occurrence of a disturbance from the estimated disturbance value, A temperature change amount calculation step of calculating a change amount of the hot metal temperature from the heat change amount, A second state space model that estimates the state of the blast furnace corresponding to the observed quantities including the observed quantities observable with a relatively slow time response to operations and disturbances to the blast furnace and the observed quantities observable with the relatively slow time response, and generating a third state space model using the change amount of the hot metal temperature, A state observation quantity estimation step of estimating the observed quantities observed when the second time longer than the first time has elapsed after an operation to the blast furnace and the state of the blast furnace based on the observed quantities using the third state space model, An operation amount determination step of determining an operation amount related to the next operation to the blast furnace based on the estimation result of the state observation quantity estimation step, comprising The state observation quantity estimation step is When distinguishing a period in which the state of superposition of disturbances is different as TAP according to the cycle of tapping from the blast furnace, A representative value calculation step of calculating a representative value of the observed quantities observable with a relatively slow time response in the target period from the observed quantities observable with a relatively slow time response in the target period, which is a period spanning one or more TAPs, An observed quantity correction step of correcting the observed quantities observable with a relatively slow time response in the target period so that the representative value and the time average value of the observed quantities observable with a relatively slow time response in the target period match, A state estimation step of estimating the state of the blast furnace using an observable quantity observable with a relatively slow time response during the corrected period of interest and the third state space model; A control method having the above. **Claim 5** The representative value calculation step includes: The control method according to claim 4, wherein the representative value is any one of an average value, a trimmed average value, or a median value of an observable quantity observable with a relatively slow time response during the period of interest. **Claim 6** The observable quantity correction step includes: Adding or subtracting the same constant to each value of the estimated value of the observable quantity during the period of interest so that the representative value and the time average value of the estimated value of the observable quantity obtained by applying the observable quantity observable with a relatively slow time response during the period of interest to the third state space model match; The control method according to claim 4 or 5, wherein correction is performed such that the obtained value is an observable quantity observable with a relatively slow time response observed when the second time has elapsed.

Images