JP4945332B2 - Exhaust gas treatment device controller, exhaust gas treatment system, plant with exhaust gas treatment system, and control method and control program for exhaust gas treatment device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable feed back control which can cope with idle time when reducing a contaminating substance content by injecting ammonium into a gas containing nitrogen oxide and the like. <P>SOLUTION: An apparatus 25 for controlling an exhaust gas treating device is provided with an internal state estimating means 41 for calculating an internal state quantity estimated value x<SB>P</SB>estimating the internal state quantity using a dynamic characteristics model prepared by modeling a variation over time of an internal state quantity variable in an exhaust gas treating device, and a feed back control variable accumulator 48 for calculating a feed back control variable u<SB>B</SB>by adding a state feed back control variable u<SB>S</SB>obtained by multiplying an internal state quantity estimating value x<SB>p</SB>by a state feed back gain -K<SB>S</SB>to an integral feed back control variable u<SB>I</SB>obtained by multiplying a value v obtained by integrating a deviation e of an exhausted contaminating substance amount y from a target exhausted contaminating substance amount r by an integral gain -K<SB>I</SB>. An ammonium injecting amount u is, for example, obtained by adding the feed back control variable u<SB>B</SB>to a feed forward control variable u<SB>W</SB>. <P>COPYRIGHT: (C)2009,JPO&amp;INPIT

Description

  The present invention relates to an exhaust gas treatment device for reducing pollutants such as nitrogen oxides contained in exhaust gas such as a gas turbine of a power plant, a control device therefor, a plant with the exhaust gas treatment device, and a control method for the exhaust gas treatment device. And control program.

The response to the increase in energy demand tends to rely on fossil fuels, and the amount of energy supplied by fossil fuels has increased, and accordingly, the amount of carbon dioxide (CO ₂ ) emissions has also increased. For this reason, the crisis of global warming has been screamed, and there is a movement to regulate CO ₂ emissions on a global scale. From such a background, a power plant equipped with a gas turbine is expected to be a combined power plant combining a gas turbine cycle and a steam turbine cycle, which can be expected to be highly efficient and consequently reduce CO ₂ . . Such a combined power plant includes, for example, a gas turbine, an exhaust heat recovery boiler that generates steam using the exhaust gas as a heat source, a steam turbine that uses the generated steam as driving steam, and a chimney that exhausts the exhaust gas that has been recovered heat. ing.

In such a combined power plant or a power plant that generates power with a gas turbine alone, nitrogen oxides (NOx) and sulfur oxides (SOx) are discharged in addition to CO ₂ . Since NOx and SOx are air pollutants, their release into the atmosphere is strictly regulated. For this reason, a denitration device or a desulfurization device may be installed in the exhaust gas flow path.

  Among them, as one of the denitration methods in the denitration apparatus, there is a method in which ammonia is blown to the catalyst portion from the ammonia injection system to activate ammonia and react with NOx to reduce and decompose NOx into harmless nitrogen gas and water vapor. . This method is generally installed between an evaporator and a economizer in a combined power plant because the reaction efficiency at 300 to 400 ° C. is good due to the temperature characteristics of the catalyst. Since ammonia is also a harmful component, it is necessary to suppress unreacted ammonia from being discharged as much as possible. Further, sulfur oxides contained in the exhaust gas can be reduced by injecting ammonia into the exhaust gas, as in the case of such a denitration apparatus.

In such a denitration reaction process using ammonia of exhaust gas, the following points should be considered.
(1) The reaction between ammonia and NOx is a chemical reaction through a catalyst, and the reaction characteristics change nonlinearly due to fluctuations in exhaust gas temperature, flow rate, pressure, ammonia / NOx concentration, and the like. Therefore, in the operating state of different plants, even if the injected ammonia flow rate is changed by the same amount, the change in the amount of NOx discharged in the denitration reaction process is not equal.
(2) Although the NOx concentration at the inlet / outlet of the catalyst unit is constantly measured by a gas analyzer, the dead time required for the measurement for gas analysis is very large (several tens of seconds to several minutes).
(3) The amount of NOx discharged to the atmosphere and the amount of ammonia discharged have a concentration restriction or a flow restriction. For example, in the case of concentration regulation, the following restrictions must be considered.
a. The NOx emission amount at a certain time does not exceed the instantaneous emission amount regulation value.
b. The average exhaust NOx amount per hour should not exceed the average travel time regulation value.
c. The amount of ammonia discharged at a certain time does not exceed the instantaneous emission control value.
d. The average amount of ammonia discharged per hour does not exceed the average regulation value for travel time.
In the case of flow regulation, the above a. -D. The unit of regulation value is set to flow rate instead of concentration.

  As the denitration control of the gas turbine exhaust gas, there is a method of adjusting the ammonia flow rate by PID control or the like based on the deviation between the exhaust NOx set value and the exhaust NOx measurement value. However, such feedback control by PID control or the like makes it difficult to stably and responsively control a system having a large dead time for NOx measurement. Also, even if there is a change in the amount of NOx generated in the gas turbine exhaust gas, the change in the waste time cannot be detected, so even if the ammonia flow rate is manipulated after the change in the exhaust NOx amount, only the waste time Cannot change the amount of exhausted NOx. For this reason, attempts have been made to utilize feedforward signals.

For example, as described in Patent Document 1, the amount of NOx generated from the gas turbine can be estimated from the fuel flow rate of the gas turbine, the flow rate of the intake air, the temperature, and the like. Moreover, the ammonia flow rate injected with respect to the generated NOx amount can be calculated from the reaction characteristics of ammonia and NOx in the catalyst so that the exhausted NOx amount becomes a certain value. Therefore, the NOx control performance can be improved by changing the ammonia injection amount in advance using the estimated generated NOx amount as a feedforward signal. Further, for example, Patent Document 2 and Patent Document 3 also disclose techniques for improving the control performance.
Japanese Patent Publication No. 3-42930 Japanese Patent No. 3410823 Japanese Patent No. 3500208 Karl J. Astrom, Bjorn Wittenmark, "Computer-Controlled Systems", 3rd Edition, USA, Prentice Hall, 1997, p. 413-418

Since sufficient controllability cannot be obtained by feedback control by PID control or the like, the amount of generated NOx is obtained by estimation calculation, and the ammonia flow rate to be injected is obtained by calculation from the reaction characteristics by using this generated NOx estimated value. It has been changed to improve controllability. However, such a control method has the following problems.
(1) The amount of NOx generated is estimated using the process amount measured for each control cycle, and is used for calculating the ammonia flow rate to be injected. Ammonia is actually injected into the process using the calculated ammonia flow control system. The ammonia flow rate control system has a time delay, and the ammonia injection amount changes with a time delay with respect to the change in the generated NOx amount. For this reason, when the amount of generated NOx changes, the ammonia injection amount becomes unbalanced, and the exhausted NOx amount may fluctuate.
(2) When the accuracy of the generated NOx estimated value is not good, a deviation occurs in the ammonia flow rate injected in advance. For this reason, fluctuations in the amount of exhausted NOx may occur.
(3) Since the reaction rate of ammonia and NOx in the catalyst changes depending on the operating state and aging, etc., the calculation of the ammonia injection amount with the exhausted NOx amount as a target value based on the generated NOx amount also depends on the operating state and aging. Deviation may occur. For this reason, fluctuations in the amount of exhausted NOx may occur.
(4) When the exhausted NOx amount deviates from the regulation value, feedback does not work for the dead time. For this reason, it takes time to return the exhausted NOx amount to the regulation value.
(5) Since the control target has a dead time, feedback cannot be applied strongly in a control method such as PID control. For this reason, even if the amount of exhausted NOx deviates from the target value, it may take a long time to return.

  Such a problem is caused by constructing a control system that relies heavily on the static feedforward control using the estimated calculated NOx amount. That is, in the case of estimating and calculating the generated NOx amount as a static characteristic for each control cycle and estimating and calculating the ammonia flow rate to be injected as the value of the static characteristic, the time of reaction of the injected ammonia and the generated NOx with the catalyst Does not take into account misalignment. For this reason, when the generated NOx amount varies, the exhausted NOx amount deviates from the target value. Further, since the control function by feedback is weak, it may take a long time to return to the target value even if a deviation from the target value occurs in the exhausted NOx amount.

  Moreover, when reducing the sulfur oxide in exhaust gas, there exists a problem similar to a denitration apparatus.

  Therefore, the present invention provides a dead time when ammonia is injected into a gas to be treated containing at least one contaminant of nitrogen oxide and sulfur oxide to reduce the content of the pollutant in the gas to be treated. The purpose is to enable strong feedback control.

In order to achieve the above-mentioned object, the present invention is configured to inject ammonia into a treatment target gas containing at least one of nitrogen oxide and sulfur oxide to reduce the content of the contaminant in the treatment target gas. In the exhaust gas treatment device control apparatus for controlling the exhaust gas treatment device to be reduced, the process amount of the exhaust gas treatment device, the amount of exhaust pollutant contained in the exhaust gas discharged from the exhaust gas treatment device, and the target value of the amount of exhaust pollutant Means for receiving the target exhaust pollutant amount, means for transmitting the ammonia injection amount injected into the gas to be treated to the exhaust gas treatment device, the measured value of the exhaust pollutant amount , the process amount of the exhaust gas treatment device and using said ammonia injection amount, and calculates the internal state quantity estimation values estimated internal state quantity of the exhaust gas treatment device for determining the exhaust pollutant amount A state feedback control amount obtained by multiplying the internal state amount estimated value by a state feedback gain, and a value obtained by integrating a deviation of the exhausted pollutant amount from the target discharged pollutant amount by an integral gain. Feedback control amount calculation means for calculating a feedback control amount based on the integral feedback control amount; and ammonia injection amount calculation means for calculating the ammonia injection amount based on the feedback control amount.

In the exhaust gas treatment system, the present invention reduces the content of the pollutant in the gas to be treated by injecting ammonia into the gas to be treated containing at least one of nitrogen oxide and sulfur oxide. Receiving an exhaust gas treatment device, a process amount of the exhaust gas treatment device, an amount of exhaust pollutant contained in exhaust gas discharged from the exhaust gas treatment device, and a target exhaust pollutant amount that is a target value of the exhaust pollutant amount Means for transmitting to the exhaust gas treatment device, an ammonia injection amount to be injected into the gas to be treated, a measured value of the exhaust pollutant amount , a process amount of the exhaust gas treatment device, and an ammonia injection amount, an internal state estimating means for calculating the internal state quantity estimation values estimated internal state quantity of the exhaust gas treatment device for determining the exhaust pollutant amount, the internal Based on a state feedback control amount obtained by multiplying a state feedback gain by a state feedback gain, and an integral feedback control amount obtained by multiplying a value obtained by integrating the deviation of the exhaust pollutant amount from the target exhaust pollutant amount by an integral gain. An exhaust gas treatment device control device comprising: a feedback control amount calculation unit that calculates a feedback control amount; and an ammonia injection amount calculation unit that calculates the ammonia injection amount based on the feedback control amount. To do.

Further, the present invention provides a plant with an exhaust gas treatment system, injecting ammonia into a combustor and a gas to be treated containing at least one pollutant of nitrogen oxide and sulfur oxide discharged from the combustor. Exhaust gas treatment device for reducing the content of the pollutant in the gas to be treated, the process amount of the exhaust gas treatment device, the amount of pollutant contained in the exhaust gas discharged from the exhaust gas treatment device, and the exhausted pollutant Means for receiving a target exhaust pollutant amount that is a target value of the amount; means for transmitting an ammonia injection amount to be injected into the gas to be treated to the exhaust gas treatment device; a measured value of the exhaust pollutant amount ; using the process amount and the ammonia injection amount of vessels, the internal state quantity estimating the internal state quantity of the exhaust gas treatment device for determining the exhaust pollutant amount An internal state estimating means for calculating a constant value, a state feedback control amount obtained by multiplying the internal state amount estimated value by a state feedback gain, and a value obtained by integrating a deviation of the discharged pollutant amount from the target discharged pollutant amount. Exhaust gas comprising: a feedback control amount calculation unit that calculates a feedback control amount based on an integral feedback control amount multiplied by a gain; and an ammonia injection amount calculation unit that calculates the ammonia injection amount based on the feedback control amount And a processor control device.

Further, the present invention provides an exhaust gas treatment device for injecting ammonia into a gas to be treated containing at least one pollutant of nitrogen oxide and sulfur oxide to reduce the content of the pollutant in the gas to be treated. In the control method of the exhaust gas treatment device to be controlled, the process amount of the exhaust gas treatment device, the amount of exhaust pollutant contained in the exhaust gas discharged from the exhaust gas treatment device, and the target exhaust pollution which is a target value of the amount of exhaust pollutant A step of receiving a substance amount; a step of transmitting an ammonia injection amount to be injected into the gas to be treated to the exhaust gas treatment device; a measured value of the exhaust pollutant amount ; a process amount of the exhaust gas treatment device; and an ammonia injection amount. with, the internal state estimating step of calculating the internal state quantity estimation values estimated internal state quantity of the exhaust gas treatment device for determining the exhaust pollutant amount A state feedback control amount obtained by multiplying the internal state amount estimated value by a state feedback gain, and an integral feedback control amount obtained by multiplying a value obtained by integrating a deviation of the exhaust pollutant amount from the target exhaust pollutant amount by an integral gain. A feedback control amount calculating step for calculating a feedback control amount based on the feedback control amount; and an ammonia injection amount calculating step for calculating the ammonia injection amount based on the feedback control amount.

Further, the present invention provides an exhaust gas treatment device for injecting ammonia into a gas to be treated containing at least one pollutant of nitrogen oxide and sulfur oxide to reduce the content of the pollutant in the gas to be treated. In the control program of the exhaust gas treatment device to be controlled, the computer has the process amount of the exhaust gas treatment device, the amount of exhaust pollutant contained in the exhaust gas discharged from the exhaust gas treatment device, and the target value of the amount of exhaust pollutant A function of receiving a target exhaust pollutant amount; a function of transmitting an ammonia injection amount injected into the processing target gas to the exhaust gas treatment device; a measured value of the exhaust pollutant amount ; a process amount of the exhaust gas treatment device; using ammonia injection volume, the internal state quantity estimation values estimated internal state quantity of the exhaust gas treatment device for determining the exhaust pollutant amount An integrated function obtained by multiplying a value obtained by integrating a deviation from the target discharged pollutant amount with a state feedback control amount obtained by multiplying the estimated value of the internal state amount by a state feedback gain, and the target discharged pollutant amount. a function of calculating a feedback control amount based on the feedback control amount, characterized that you to realize a function that calculates the ammonia injection amount based on the feedback control amount.

  According to the present invention, when injecting ammonia into a gas to be treated containing at least one pollutant of nitrogen oxide and sulfur oxide to reduce the content of the pollutant in the gas to be treated, dead time is consumed. Strong feedback control.

  An embodiment of an exhaust gas treatment device control apparatus according to the present invention will be described with reference to the drawings. In addition, the same code | symbol is attached | subjected to the same or similar structure, and the overlapping description is abbreviate | omitted.

[First Embodiment]
FIG. 2 is a block diagram of a combined power plant as an example of a plant including an exhaust gas treatment device to be controlled in the first embodiment of the exhaust gas treatment device control device according to the present invention.

  This combined power plant includes a gas turbine 1, an exhaust heat recovery boiler 3 that generates steam using the exhaust gas 2 as a heat source, a steam turbine 4 that uses the generated steam as driving steam, and a chimney that exhausts the exhaust gas that has been recovered heat. And 5. The gas turbine 1 includes an air compressor 7 that pressurizes the introduced air 6, a combustor 9 that combusts the pressurized air together with fuel supplied from the fuel system 8, and a turbine 10 that is operated by the combustion gas generated by the combustion, And a generator 11 that takes a load.

  The exhaust heat recovery boiler 3 includes a superheater 13, an evaporator 14, an exhaust gas treatment device 15, and a economizer 16 along the upstream side of the exhaust gas duct 12 through which the exhaust gas 2 flows. Steam is supplied to the steam turbine 4 through the steam pipe 17. The steam turbine 4 includes a turbine 18 that is operated by steam generated in the exhaust heat recovery boiler 3, a generator 19 that takes a load, and a condenser 20 that condenses the steam after working in the turbine 18. ing.

  Condensate from the condenser 20 is guided to the economizer 16 by the water supply pipe 21 and heated, and then evaporated by the evaporator 14, and this steam is further heated by the superheater 13. In the evaporator 14, the feed water is heated and evaporated while being forcedly circulated or naturally circulated by a temperature difference.

  Here, the turbine 10 of the gas turbine 1 and the turbine 18 of the steam turbine 4 are connected to different generators 11 and 19, respectively, but may be connected to the same generator.

  The exhaust gas treatment device 15 is, for example, a denitration device having an ammonia injection system including an ammonia injector 23, a catalyst portion 22, and an ammonia blowing portion 24. Ammonia supplied from the ammonia injector 23 is blown from the ammonia blowing section 24 to the catalyst section 22 and is activated in contact with the catalyst. The activated ammonia reacts with nitrogen oxide (NOx) contained in the exhaust gas 2, and a part of NOx is reduced and decomposed into harmless nitrogen gas and water vapor. Since this reaction has good efficiency at 300 to 400 ° C. due to the temperature characteristics of the catalyst, the exhaust gas treatment device 15 is installed between the evaporator 14 and the economizer 16. An exhaust gas treatment device controller 25 that controls the exhaust gas treatment device 15 is connected to the exhaust gas treatment device 15. In this embodiment, nitrogen oxides are mainly described as pollutants. However, even if sulfur oxide (SOx) contained in the exhaust gas 2 is used, for example, a desulfurization device having the same configuration as the exhaust gas treatment device 15 is used. Is used. In addition, in FIG. 2, although the combined power generation plant was illustrated as a plant to which this invention is applied, in the case of the plant provided with the desulfurization apparatus which removes sulfur oxide, it is equipped with a coal or heavy oil fired boiler instead of a combined power generation plant. Often a plant.

  That is, the exhaust gas treatment device control device 25 of the present embodiment injects ammonia into a treatment target gas containing at least one of the pollutants of nitrogen oxides and sulfur oxides to remove the contaminants in the treatment target gas. The exhaust gas treatment device for reducing the content is controlled.

  The exhaust gas treatment device 15 injects ammonia into the exhaust gas 2 that is the gas to be treated to reduce the amount of pollutants contained in the exhaust gas 2 and discharge it. The amount of pollutant discharged from the exhaust gas treatment device 15 (discharge pollutant amount) varies depending on the amount of ammonia injected into the exhaust gas 2 (ammonia injection amount), the process amount, and the like. Here, the process amount is an amount involved in the exhaust gas treatment process of the exhaust gas treatment device 15, and is an amount that the exhaust gas treatment device 15 does not control or cannot control. Examples of the process amount include the amount of pollutants contained in the exhaust gas 2 and the combustion temperature.

  In the present embodiment, a dynamic characteristic model representing the dynamic characteristics of the exhaust gas treatment process of the exhaust gas treatment device 15 and an internal state quantity indicating the internal state of the exhaust gas treatment device 15 are introduced. This dynamic characteristic model is a model that describes the relationship between the ammonia injection amount, the process amount, the internal state amount, and the amount of discharged pollutants.

Exhaust gas treatment processes such as a denitration process actually have nonlinear characteristics, but can be approximately represented by a linear time-invariant dynamic characteristic model. Therefore, in the present embodiment, it is assumed that the dynamic characteristics of the exhaust gas treatment process are expressed by the following state equation of discrete time t.
x (t + 1) = Ax (t) + B _u u (t) + B _w w (t) (1)
y (t) = Cx (t) (2)
here,
u (t) is the ammonia injection amount at time t,
w (t) is the process amount at time t,
x (t) is an internal state quantity indicating the internal state of the exhaust gas treatment device at time t,
y (t) is the amount of pollutant discharged at time t,
It is.

In the present embodiment, u (t) and y (t) are scalar quantities. x (t) is an n-order vector quantity, and w (t) is an m-order vector quantity. A is a matrix of n rows and n columns, B _u is a matrix of n rows and 1 column, B _w is a matrix of n rows and m columns, and C is a matrix of 1 row and n columns. x (t) does not have to have a physical meaning in particular. In addition, as the process amount w (t), an estimated amount of generated pollutant, which is an estimated value of the amount of contaminant such as NOx contained in the exhaust gas 2, is used. Therefore, m is 1.

A, B _u and B _w are constants determined so as to better express the dynamic characteristics of the actual exhaust gas treatment process of the exhaust gas treatment device 15. These constants can be determined from the physical causal relationship based on the characteristics of the measuring instrument, the length of the pipe, the system design, and the like. It can also be determined based on the response to the test signal. Furthermore, the constant determined from the physical causal relationship may be corrected based on the actual response.

  Next, a method for determining the ammonia injection amount u (t) using this dynamic characteristic model will be described.

  FIG. 1 is a block diagram of an exhaust gas treatment device control apparatus according to the present embodiment.

  The exhaust gas treatment device control device 25 includes a target exhausted pollutant amount receiving unit 31, an exhausted pollutant amount receiving unit 32, and a process amount receiving unit 33. The target exhausted pollutant amount receiving means 31 has a target exhausted pollutant amount r given by some method in order to keep the amount of pollutant contained in the exhaust gas discharged from the exhaust gas treatment device 15 below a predetermined value. (T) is received. The target exhausted pollutant amount r (t) may be a value that changes over time or may be a constant value. The discharged pollutant amount receiving means 32 receives the measured value y (t) of the pollutant amount contained in the exhaust gas discharged from the exhaust gas processor 15. The process quantity receiving means 33 receives the process quantity w (t).

In the present embodiment, the ammonia injection amount u (t), which is an input of the dynamic characteristic model of the exhaust gas treatment process, is a feedforward control amount u _w (t) and a feedback control amount u _B (t) as follows: And the sum of
u (t) = u _w (t) + u _B (t) (3)
Further, the feedforward control amount u _{w (t)} is assumed to process the amount of w (t) multiplied by the feed forward gain K _w. That is,
u _w (t) = K _w w (t) (4)
It is.

The feedback control amount u _B (t) is given as the sum of the state feedback control amount u _S (t) and the integral control amount u _I (t). That is,
u _B (t) = u _S (t) + u _I (t) (5)
It is.

The state feedback control amount u _S (t) is given by multiplying the internal state amount x (t) by the state feedback gain −K _S. That is,
u _S (t) = − K _S x (t) (6)
It is.

The integral control amount u _I (t) is obtained by integrating the integral gain −K _I into a value v (t) obtained by integrating the deviation e (t) of the discharged pollutant amount y (t) from the target discharged pollutant amount r (t). It shall be given by multiplying by That is,
e (t) = r (t) -y (t) (7)
v (t + 1) = e (t) + v (t)
= R (t) -y (t) + v (t)
= R (t) -Cx (t) + v (t) (8)
u _I (t) = − K _I v (t) (9)
It is.

ここで、Ｉは単位行列を表す。
また、

とおくと、次のように書ける。
ｘ'（ｔ＋１）＝Ａ'ｘ（ｔ）＋Ｂ'ｕ_Ｂ（ｔ）＋Ｆ'ｗ'（ｔ） …（１２）
ｙ'（ｔ）＝Ｃ'ｘ'（ｔ） …（１３） When the equations (1) to (8) are used, the entire system of the exhaust gas treatment device 15 and the exhaust gas treatment device control device 25 is expressed by the following equation.

Here, I represents a unit matrix.
Also,

Then, it can be written as follows.
x ′ (t + 1) = A′x (t) + B′u _B (t) + F′w ′ (t) (12)
y ′ (t) = C′x ′ (t) (13)

ここで、ＱおよびＲは調整パラメータであり、Ｑはｎ行ｎ列の正定行列、Ｒは正の値を持つスカラー量である。 Next, a method for calculating the gain appearing in this system will be described. The state feedback gain −K _S and the integral gain −K _I can be obtained by introducing an evaluation function and using, for example, an optimal regulator theory in an infinite interval. The evaluation function is expressed using at least one of the deviation e (t), the internal state variable x (t), and the ammonia injection amount u (t), and is given by the following equation, for example.

Here, Q and R are adjustment parameters, Q is a positive definite matrix of n rows and n columns, and R is a scalar quantity having a positive value.

In order to reduce the evaluation function J1, the internal state quantity x ′ and the feedback control quantity u _B (t) may be small values. The feedback gain (−K _S and −K _I ) that minimizes the evaluation function J is given by Expression (16) using the solution S of the discrete time Riccati equation expressed by Expression (15). .
A ′ ^T SA′-SA ′ ^T SB ′ (B ′ ^T SB ′ + R) ⁻¹ B ′ ^T SA ′ + Q = 0
... (15)
K = [K _S K _I ] = (B ′ ^T SB ′ + R) ⁻¹ B ′ ^T SA ′ (16)

The feedback control amount u _B (t) is determined using the state feedback gain −K _S and the integral gain −K _I obtained in this way. Such feedback control can be said to be an optimal control law in the sense that the evaluation function J1 is minimized by the constant gains −K _S and −K _I.

Actually, since the internal state quantity x (t) cannot be measured directly, it is necessary to estimate its value. Accordingly, a method for obtaining the internal state quantity estimated value x _p (t) obtained by estimating the integral gain x (t) will be described next.

The adjustment parameter L is introduced, and the dynamic characteristic model of the exhaust gas treatment process is expressed as follows using the internal state quantity estimated value x _p (t).
x _p (t + 1) = Ax _p (t) + B _u u (t) + B _w w (t)
_{+ L {y (t) -y} p (t)} ... (17)
y _p (t) = Cx _p (t) (18)
Here, y _p (t) is an estimated value of the discharged pollutant amount y (t) calculated using the internal state quantity estimated value x _p (t).

The adjustment parameter L is determined so that the A-LC is stabilized. For example, with respect to (A ^T , C ^T ), using the solution S ₀ obtained by solving the Riccati equation represented by Equation (19), the adjustment parameter L is determined as in Equation (20). LC becomes stable.
AS ₀ ^AT −S ₀ −AS ₀ C ^T (CS ₀ C ^T + R ₀ ) ⁻¹ CS ₀ A ^T + Q ₀ = 0
... (19)
L ^T = K ₀ = (CS ₀ C ^T + R ₀ ) ⁻¹ CS ₀ A ^T (20)
Here, Q ₀ and R ₀ are adjustment parameters, Q ₀ is an n-by-n positive definite matrix, and R ₀ is a scalar quantity having a positive value.

Using the adjustment parameter L thus determined, a dynamic characteristic model of the exhaust gas treatment process is expressed as in Expression (17) and Expression (18). At this time, since the internal state quantity estimated value x _p (t) approaches the true internal state x (t), the exhaust gas treatment process is controlled using the internal state quantity estimated value x _p (t). Can do. That is, inside the exhaust gas treatment device controller 25, for example, the state feedback control amount u _S (t) is obtained using the following equation instead of the equation (6).
u _S (t) = − K _S x (t) (21)

  In order to realize such control, the exhaust gas processor control device 25 includes an internal state estimation means 41, a state feedback gain multiplier 42, a subtractor 43, an integrator 44, an integral gain multiplier 47, and a feedback control amount adder. 48, a feedforward gain multiplier 49, and a feedforward control amount adder 50.

  FIG. 3 is a flowchart showing a control flow of the exhaust gas treatment device control apparatus in the present embodiment.

When using the exhaust gas treatment unit controller 25 first obtains the various gains, such as state feedback gain -K _S as described above, is set to the exhaust gas treatment device controller 25 (step S1). Next, the time is advanced by 1 (step S2), and the measured value y (t) of the discharged pollutant amount, the target discharged pollutant amount r (t), and the process amount w (t) at that time are set to the discharged pollutant Received by the quantity receiving means 32, the target discharged pollutant quantity receiving means 31 and the process quantity receiving means 33 (step S3).

Next, the internal state estimation means 41 uses the measured value y (t), the process amount w (t), and the ammonia injection amount u (t) of the discharged pollutant amount to estimate the internal state amount x _p (t). Is calculated. In addition, the state feedback gain multiplier 42 calculates the state feedback control amount u _S (t) by multiplying the internal state amount estimated value x _p (t) by the state feedback gain −K _S (step S4).

On the other hand, the subtractor 43 calculates the deviation e (t) by subtracting the discharged pollutant amount y (t) from the target discharged pollutant amount r (t). The integrator 44 integrates this deviation e (t) in the time direction to calculate an integral value v (t). The integrator 44 is realized by, for example, a shift register (z ⁻¹ ) 45 and an adder 46. The integral gain multiplier 47 multiplies the integral value v (t) by the integral gain −K _I to calculate the integral control amount u _I (t) (step S5). In this way, until the deviation e (t) between the discharged pollutant amount y (t) and the target discharged pollutant amount r (t) becomes 0, that is, the discharged pollutant amount y (t) is the target discharged pollutant. An integral control amount u _I (t) is provided that reduces the deviation e (t) until it becomes equal to the substance amount r (t). Note that since the true value of v (t), which is the state variable of the integrator, is known, it is not necessary to perform state estimation like the internal state quantity x (t).

Thereafter, the feedback control amount adder 48 adds the state feedback control amount u _S (t) to the integral control amount u _I (t) to calculate the feedback control amount u _B (t) (step S6).

Feed forward gain multiplier 49 multiplies the feedforward gain _{K w} to process variable w (t), and calculates the feedforward control amount _u w (t) (step S7). The feedforward control amount adder 50 calculates the ammonia injection amount u (t) by adding the feedforward control amount u _w (t) to the feedback control amount u _B (t) (step S8).

Even when the feedforward control amount u _w (t) is set to 0, the exhaust gas treatment process can be controlled. Such control can be realized, for example, by setting the feedforward gain _Kw to 0 or deleting the feedforward gain multiplier 49 and the feedforward control amount adder 50 from the exhaust gas treatment device controller 25. it can.

The ammonia injection amount u (t) obtained in this way is transmitted to the internal state estimating means 41 and used for calculating the internal state amount estimated value x _p (t). It is transmitted to the processor 15 and used for controlling the exhaust gas treatment process.

  Next, it is determined whether or not the control of the exhaust gas treatment device 15 is ended by the end of the plant operation or the like (step S9), and the steps S2 to S8 are repeated until the control is ended.

  The exhaust gas treatment device control device 25 can also be realized by a program that causes one or more computers to realize the above-described functions.

The exhaust gas treatment device control device 25 according to the present embodiment includes an internal state estimation unit 41 and a feedback control amount calculation unit that calculates a feedback control amount u _I using an integrator 44 or the like. For this reason, powerful feedback control in consideration of dynamic characteristics can be performed, and feedback control that is strong in dead time can be performed. In addition, the feedback control uses a feedback gain in consideration of dynamic characteristics, so that the stability and reliability of the control of the exhaust gas treatment process is improved.

Furthermore, since the feed forward control using the process amount of the exhaust gas treatment process can be effectively combined and used, it is possible to control the exhaust gas treatment process that is highly responsive. Further, when the feedforward control amount u _w (t) is used in the internal state estimation means 41, the reliability of the internal state amount estimated value x _p (t) is increased, and more accurate control is possible.

  In this embodiment, the sum of squares in the infinite interval is used for both the deviation e (t) and the ammonia injection amount u (t) as the evaluation function. This makes it possible to apply the optimal regulator theory for an infinite interval and calculate the feedback gain. Since the control gain obtained by the optimal regulator theory in an infinite interval is a constant value, it can be used endlessly. Further, by calculating the control gain using the optimal regulator theory in an infinite interval, it is possible to realize the control stability and quick control response of the exhaust gas treatment process.

  In this embodiment, a dynamic characteristic model extended to a system that takes into account the presence of a feedforward control amount and an integrator is used as a dynamic characteristic model used when applying the optimal regulator theory. As a result, in addition to the state feedback gain, an integral gain can be obtained at the same time, so there is no flaw in both parameters, and highly reliable denitration control can be realized.

The feedback gain (-K _S and -K _I ) can also be obtained by other methods. In the above description, the feedback gain is determined using the evaluation function J1 with an evaluation interval up to an infinite step ahead. For example, the feedback gain is determined using the next evaluation function J2 with an evaluation interval up to a finite N step ahead. You can also.

The feedback gain (−K _S and −K _I ) that minimizes the evaluation function J2 is given by Equation (19) using the solution S (t) of the discrete time Riccati equation of Equation (18).
S (t) = A ′ ^T S (t + 1) A ′ + Q
^{-A 'T S (t + 1} ) B' (B 'T S (t + 1) B' + R) -1 B 'T S (t + 1) A'
... (23)
K (t) = [K _S (t) K _I (t)]
^{= (B 'T S (t} + 1) B' + R) -1 B 'T S (t + 1) A' ... (24)
The discrete-time Riccati equation can be solved, for example, by dynamic optimization (for example, see Non-Patent Document 1).

In this way, a feedback gain from time t to time t + N is obtained by the optimal regulator theory in a finite interval. Using these feedback gains, the state feedback amount u _S (t) from time t to time t + N is given as follows.
u _S (t) = − K _S (t) x (t)
u _S (t + 1) = − K _S (t + 1) × (t + 1)
u _S (t + 2) = − K _S (t + 2) × (t + 2) (25)
:
u _S (t + N) = − K _S (t + N) × (t + N)
Similarly, the integral control amount u _I (t) from time t to time t + N is given as follows.
u _I (t) = − K _I (t) v (t)
u _I (t + 1) = − K _I (t + 1) v (t + 1)
u _I (t + 2) = − K _I (t + 2) v (t + 2) (26)
:
u _I (t + N) = − K _I (t + N) v (t + N)

In the equations (25) and (26), in order to calculate the state feedback control amount u _S (t) and the integral control amount u _I (t) from time t to time t + N, the internal values from time t to time t + N are calculated. A value v (t) obtained by integrating the state quantity x (t) and the deviation e (t) is used. By the way, u _B (t) = u _S (t) + u _I (t) as in Expression (1), and x ′ (t + k) = [x (t + k) v (t + k)] ^{T as in} Expression (11). Therefore, Expression (25) and Expression (26) can be written as follows.
u _B (t) = − K (t) x ′ (t)
u _B (t + 1) = − K (t + 1) x ′ (t + 1)
= −K (t + 1) {A′x ′ (t) + B′u _B (t) + F′w ′ (t)}
= −K (t + 1) {A′x ′ (t) −B′K (t) x ′ (t) + F′w ′ (t)}
= −K (t + 1) {A′−B′K (t)} x ′ (t) −K (t + 1) F′w ′ (t)
u _B (t + 2) =
−K (t + 2) {A′−B′K (t + 1)} {A′−B′K (t)} x ′ (t)
−K (t + 2) F′w ′ (t + 1)
−K (t + 2) {A′−B′K (t + 1)} F′w ′ (t)
:
u _B (t + N) =
−K (t + N) {A′−B′K (t + N−1)} {A′−B′K (t + N−2)}
... {A'-B'K (t)} x '(t)
−K (t + N) F′w ′ (t + N−1)
−K (t + N) {A′−B′K (t + N−2)} F′w ′ (t + N−2)
-...
-K (t + N) {A'-B'K (t + N-2)} ...
{A'-B'K (t)} F'w '(t)
... (27)
In the equation (27), future signals w ′ (t + k), k = 1, 2,..., N−1 are required. If the future w ′ (t) is known, use it. When the future w ′ (t) is unknown, by setting w ′ (t + k) = w ′ (t), k = 1, 2,..., N, the current value will continue in the future. become.

From the above, the feedback control amount u _B (t) from time t to time t + N is calculated from the state variable x ′ (t) at time t and K (t), K (t + 1),. It can be calculated using K (t + N). Therefore, at time t, the feedback gain and control input from time t to time t + N can be determined at a time. Therefore, a cycle of time elapsed between the time t + N from the time t, for each this period, it is possible to continue using by switching the state feedback gain -K _S and the integral gain -K _I.

  In this way, the feedback gain can be obtained not only by the infinite interval but also by the optimal regulator theory of the finite interval, and optimal control can be performed in the sense of minimizing the evaluation function J2 of the finite interval.

Further, the feedback gain (−K _S and −K _I ) can be calculated by the least square method.

である。また、式（１２）より、時刻ｔ＋１から時刻ｔ＋Ｎまでの状態変数ｘ'（ｔ）は、次のように表される。

Here, the evaluation function J2 (see Expression (22)) is expressed as follows using a vector and a matrix.
_{^{J2 = X T Q n X +}} U T RU ... (28)
here,

It is. Further, from the equation (12), the state variable x ′ (t) from the time t + 1 to the time t + N is expressed as follows.

This can be expressed in matrix form as follows.

This equation can be written as:
X = A _n x ′ (t) + A _b U + F _w (29)
_{Here, A n,} A _b and Fw are as defined by the following equation.

Substituting equation (29) into equation (28) yields:
J2 = X ^T Q _n X + U ^T RU
= (A _n x ′ (t) + A _b U + F _w ) ^T Q _n (A _n x ′ (t) + A _b U + F _w )
+ U ^T RU
_{^{= U T (A b T Q}} n A b + R) U
_{^{+ (X '(t) T}} A n T Q n A b + F w T Q n A b) U
_{^{+ U T (A b T Q}} n F w + A b T Q n A n x '(t))
+ X ′ (t) ^TA _n ^T Q _n A _b x ′ (t)
_{^{+ X '(t) T A}} n T Q n F w + F w T Q n A n x' (t)
+ _{F ^w} T _Q n _F ^w
U that minimizes J2 can be calculated as a value obtained by partial differentiation of J by U. By solving ∂J2 / ∂U = 0, the optimal solution U = [u _B (t), u _B (t + 1),..., U _B (t + N−1)] ^T is obtained as follows.
U = − (A _b ^T Q _n A _b + R) ⁻¹ (A _b Q _n F _w + A _b Q _n A _n x ′ (t))
... (30)
Using u _B (t), u _B (t + 1),..., U _B (t + N−1) obtained by the equation (30), the ammonia injection amounts u (t), u (t + 1),. -1) is obtained. This equation shows that u _B (t), u _B (t + 1),..., U _B (t + N) can be calculated by the internal state quantity x ′ (t) at time t and the feedback gain K (t), K It can be calculated by (t + 1),..., K (t + N). That is, at time t, the feedback gain and the control input from time t to time t + N can be determined at a time. Therefore, a cycle of time elapsed between the time t + N from the time t, for each this period, it is possible to continue using by switching the state feedback gain -K _S and the integral gain -K _I.

  Further, as described above, the feedback gain calculated every predetermined time can be handled as a constant gain.

For example, the feedback control amount u _B (t) is obtained at a time from time t to time t + N at time t using equation (27) or equation (30). Therefore, at time t, only the first u _B (t) is used for control, and at the next time t + 1, t + 1 is replaced with t, and only the first u _B (t) is used again for control. I will do it. By performing this continuously, only the portion for calculating u _B (t) is used in the obtained feedback gain sequence.

Specifically, the following formula (31) or formula (32) is used instead of formula (25) and formula (26) or formula (30).
u (t) =-K (t) x (t) (31)
U = − [1 0... 0] (A _b ^T Q _n A _b + R) A _b ^T Q _n A _n x (t)
... (32)

  In this way, a constant value can be used as the feedback gain. That is, only one type of optimum control gain needs to be mounted. In other words, a control algorithm using a feedback gain calculated every time a predetermined time elapses can be a constant gain control rule, and handling is easy.

[Second Embodiment]
FIG. 4 is a block diagram of the second embodiment of the exhaust gas treatment device control apparatus according to the present invention.

  As in the first embodiment, the exhaust gas treatment device controller 25 of the present embodiment controls the exhaust gas treatment device 15 provided in, for example, a combined power plant. The exhaust gas treatment device control device 25 includes a target exhausted pollutant amount receiving unit 31, an exhausted pollutant amount receiving unit 32, and a process amount receiving unit 33.

The exhaust gas treatment device control device 25 of the present embodiment further includes a target ammonia discharge amount receiving means 36 and an exhaust ammonia amount reception means 35. The target ammonia discharge amount receiving means 36 has a target ammonia discharge amount r _A (t given by some method in order to keep the ammonia amount contained in the exhaust gas discharged from the exhaust gas treatment device 15 to be a predetermined value or less. ). The target ammonia discharge amount r _A (t) may be a value that changes with time or may be a constant value. The discharged ammonia amount receiving means 35 receives the measured value y _A (t) of the discharged ammonia amount contained in the exhaust gas discharged from the exhaust gas treatment device 15.

The exhaust gas treatment device controller 25 also uses the target ammonia discharge amount r _A (t) and the discharge ammonia amount y _A (t) for control. Therefore, the exhaust gas treatment device controller 25, by subtracting the target ammonia emissions r (t) from the target ammonia emissions r _{A (t),} and a subtractor 52 for calculating a deviation e _{A (t).} Further, an ammonia integral gain multiplier 53 for calculating Δu _A by multiplying the deviation e _A (t) by the ammonia integral gain −K _IA is provided.

Further, the integral gain multiplier 47, before the deviation from the target discharge pollutant quantity r (t) obtained by subtracting the discharge pollutant quantity y (t) e (t) is integrated, the integral gain -K _I Multiply to calculate Δu _c (t). Δu _A and Δu _c are added by the adder 54 to become Δu _I , and integrated by the integrator 44 to become u _I (t).

The internal state estimation means 41 also sequentially calculates an internal state amount estimated value x _p (t), which is an estimated value of the internal state x (t), using the exhausted ammonia amount y _A (t).

In the present embodiment, it is assumed that the exhaust gas treatment process has linear time-invariant dynamic characteristics represented by the following state equation of discrete time t.
x (t + 1) = Ax (t) + B _u u (t) + B _w w (t) (33)
[Y (t) y _A (t)] ^T = Cx (t) (34)
Here, y _A (t) is the amount of ammonia discharged at time t. The symbols used in Expression (33) and Expression (34) represent a vector quantity, a scalar quantity, or a matrix, as in the first embodiment. The order of may vary.

For an expanded system in which the integrator 44 inside the exhaust gas treatment device controller 25 is incorporated in the dynamic characteristic model of the exhaust gas treatment device 15 represented by the equations (33) and (34), the first Similar to the embodiment, the evaluation function J1 represented by the equation (14) is introduced. The optimal control law that minimizes the evaluation function J1 can be realized by using the feedback gain K given by the following equation using the solution S of the discrete time Riccati equation expressed by the equation (15).
_{K = [K S K I K} IA] = (B 'T SB' + R) -1 B 'T SA' ... (35)
Here, using the obtained state feedback gain −K _S , integral gain −K _I and integral gain for ammonia −K _IA , and the state variable x (t), the state feedback control amount u _S (t) and the integral are integrated. The control amount u _I (t) is given as follows.
u _S (t) = − K _S x (t) (36)
Δu _c (t) = − K _I {r (t) −y (t)} (37)
Δu _A (t) = − K _IA {r _A (t) −y _A (t)} (38)
u _I (t + 1) = u _I (t) + Δu _c (t) + Δu _A (t) (39)
Using such a model, the internal state estimation means 41 estimates the internal state quantity estimated value x _p (t) and uses it for control. Such control can be said to be an optimal control law in the sense of minimizing the evaluation function J1.

  In this way, it is possible to perform feedback control of the exhaust gas treatment device controller 25 in consideration of the dynamic characteristics related to the amount of discharged ammonia in addition to the amount of discharged pollutants. As a result, stable and highly reliable denitration control can be performed not only for the amount of pollutant discharged, but also for the amount of ammonia discharged as close as possible to each target value.

[Third Embodiment]
FIG. 5 is a block diagram of the third embodiment of the exhaust gas treatment device control apparatus according to the present invention.

In the exhaust gas treatment device control apparatus 25 of the present embodiment, the exhaust gas treatment device control apparatus of the first embodiment has a speed configuration, and the integrator 44 is moved at the end of the feedback control. Accordingly, before the internal state quantity estimated value x _p (t) estimated by the internal state estimating means 41 is sent to the state feedback gain multiplier 42, the difference is calculated by the differentiator (1-z ⁻¹ ) 56. Then, Δu _S (t) is calculated by multiplying by the state feedback gain −K _S.

Similarly to the fifth embodiment, before the deviation e (t) obtained by subtracting the discharged pollutant amount y (t) from the target discharged pollutant amount r (t) is integrated, the integral gain − is multiplied by K _I is calculated Δu _I (t). Δu _S (t) and Δu _I (t) are added by the adder 57 to be Δu _B (t). If there is no signal selector 55, the exhaust gas treatment device control device 25 realizes a control calculation equivalent to the exhaust gas treatment device control device of the first embodiment.

  The signal selector 55 is, for example, a signal changeover switch, and the ammonia injection amount can be set from the outside by switching this switch. This makes it possible for the operator to continuously switch between forcibly setting the ammonia injection amount and setting the ammonia injection amount by another control law. Further, as the signal selector 55, an upper / lower limiter or an intermediate amount selector can be used.

  At this time, the feedback control amount determined based on the discharged pollutant amount y (t) using a feedback gain or the like is not integrated. For this reason, when it is switched to the automatic control by the exhaust gas treatment device control measure 25 after forcibly setting the ammonia injection amount from the outside, it can be switched to bumpless. Therefore, an accurate control operation is realized immediately after switching.

[Fourth Embodiment]
FIG. 6 is a partially enlarged block diagram of the fourth embodiment of the exhaust gas treatment device control apparatus according to the present invention.

  The exhaust gas treatment device control apparatus of the present embodiment includes a plurality of feedback control differential calculators 61 and a signal selector 62.

  The gas turbine 1 that generates exhaust gas containing pollutants may vary greatly in pollutant emissions depending on the operating state. Therefore, the exhaust gas treatment device may have different dynamic characteristics according to the operating state of the gas turbine 1. Alternatively, the dynamic characteristics may differ depending on the operating state of the gas turbine 1. In such a case, it may be difficult to model the dynamic characteristics of the exhaust gas treatment device 15 with one model.

In the present embodiment, the plurality of feedback control differential calculators 61 are, for example, the internal state estimation means 41, the difference unit 56, the state feedback gain multiplier 42, the subtractor 43, and the integral gain multiplier in the third embodiment. 47 and an adder 57, respectively. The feedback control differential calculator 61 is configured to obtain an appropriate control law for each different operating state of the gas turbine 1. This appropriate control law uses a dynamic characteristic model of the exhaust gas treatment device 15 modeled in accordance with the operating state of the gas turbine 1 and an appropriate evaluation function. The state feedback gain −K _S and the integral gain −K _I are obtained using a dynamic characteristic model and an evaluation function for each operation state of the gas turbine 1.

Each of the feedback control differential calculators 61 calculates Δu _B (t). Among these Δu _B (t), an appropriate Δu _B (t) is selected for the operating state of the gas turbine 1 by the signal selector 62, and the Δu _B (t) is integrated using the shift register 45. Thus, the feedback control amount u _B (t) is calculated.

  As described above, the exhaust gas treatment device control apparatus according to the present embodiment performs control while switching so that the most suitable control law can be selected at any time according to the operating state of the equipment that generates pollutants such as boilers and gas turbines. Is. Therefore, it is possible to perform exhaust gas treatment control such as accurate denitration control with little deviation corresponding to the operating state. Thereby, control of the exhaust gas treatment device with good response characteristics can be realized even in a plant or the like whose operating state greatly changes.

[Other embodiments]
The above-described embodiments are merely examples, and the present invention is not limited to these. Moreover, you may implement combining the characteristic of each embodiment.

It is a block diagram in a 1st embodiment of an exhaust gas treatment device control device concerning the present invention. 1 is a block diagram of a combined power plant as an example of a plant including a denitrator to be controlled in a first embodiment of an exhaust gas treatment device control apparatus according to the present invention. It is a flowchart which shows the flow of control in 1st Embodiment of the exhaust gas processor control apparatus which concerns on this invention. It is a block diagram in 2nd Embodiment of the exhaust gas processor control apparatus which concerns on this invention. It is a block diagram in 3rd Embodiment of the exhaust gas processor control apparatus which concerns on this invention. It is a block diagram in 4th Embodiment of the exhaust gas processor control apparatus which concerns on this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Gas turbine, 2 ... Exhaust gas, 3 ... Waste heat recovery boiler, 4 ... Steam turbine, 5 ... Chimney, 6 ... Introduction air, 7 ... Air compressor, 8 ... Fuel system, 9 ... Combustor, 10 ... Turbine, DESCRIPTION OF SYMBOLS 11 ... Generator, 12 ... Exhaust gas duct, 13 ... Superheater, 14 ... Evaporator, 15 ... Exhaust gas processor, 16 ... Carbon-saving device, 17 ... Steam piping, 18 ... Turbine, 19 ... Generator, 20 ... Condensate , 22 ... Catalyst unit, 23 ... Ammonia injector, 24 ... Ammonia spraying unit, 25 ... Exhaust gas treatment device control device, 31 ... Target exhausted pollutant amount receiving means, 32 ... Exhaust pollutant amount receiving means, 33 ... Process Amount receiving means 34 ... Ammonia injection amount transmitting means 35 ... Exhaust ammonia amount receiving means 36 ... Target exhaust ammonia quantity receiving means 41 ... Internal state estimating means 42 ... State feedback gain multiplier 43 ... Subtractor 4 ... integrator, 45 ... shift register, 46 ... adder, 47 ... integral gain multiplier, 48 ... feedback control amount adder, 49 ... feed forward gain multiplier, 50 ... feed forward control amount adder, 52 ... Subtractor, 53 ... Integral gain multiplier for ammonia, 54 ... Adder, 55 ... Signal selector, 56 ... Differentiator, 57 ... Adder, 61 ... Feedback control differential calculator, 62 ... Signal selector

Claims (14)

Exhaust gas treatment device control for controlling an exhaust gas treatment device for injecting ammonia into a gas to be treated containing at least one pollutant of nitrogen oxide and sulfur oxide to reduce the content of the pollutant in the gas to be treated In the device
Means for receiving the process amount of the exhaust gas treatment device, the amount of exhaust pollutant contained in the exhaust gas discharged from the exhaust gas treatment device, and the target exhaust pollutant amount which is a target value of the amount of exhaust pollutant;
Means for transmitting an ammonia injection amount to be injected into the gas to be treated to the exhaust gas treatment device;
Using the measured value of the exhaust pollutant amount , the process amount of the exhaust gas treatment device, and the ammonia injection amount, an internal state amount estimated value obtained by estimating the internal state amount of the exhaust gas treatment device that determines the exhaust pollutant amount Internal state estimating means for calculating;
A state feedback control amount obtained by multiplying the internal state amount estimated value by a state feedback gain, and an integral feedback control amount obtained by multiplying a value obtained by integrating a deviation of the exhaust pollutant amount from the target exhaust pollutant amount by an integral gain. A feedback control amount calculating means for calculating a feedback control amount based on;
An ammonia injection amount calculating means for calculating the ammonia injection amount based on the feedback control amount;
An exhaust gas treatment device control apparatus comprising:   A means for inputting the ammonia injection amount from the outside, and the feedback control amount calculating means is configured to deviate the discharge pollutant amount from the target discharge pollutant amount while the ammonia injection amount is input from the outside. The exhaust gas treatment device control device according to claim 1, wherein the exhaust gas treatment device control device is not integrated.   3. The exhaust gas treatment device control device according to claim 1, wherein a plurality of the feedback control amount calculation means are provided according to an operation state of the exhaust gas treatment device.   The state feedback gain and the integral gain maximize and minimize an evaluation function using at least one of the deviation of the exhaust pollutant amount from the target exhaust pollutant amount, the internal state amount, and the ammonia injection amount. The exhaust gas treatment device control device according to claim 1, wherein the exhaust gas treatment device control device is calculated by using any one of the following.   The evaluation function is a sum of a square sum of deviations of the exhaust pollutant amount from the target exhaust pollutant amount in an infinite interval and a square sum of the ammonia injection amount in an infinite interval, and the state feedback gain and the The exhaust gas processor control device according to claim 4, wherein the integral gain is calculated by an optimal regulator theory for an infinite interval using the evaluation function.   The evaluation function is a sum of a square sum of deviations of the exhaust pollutant amount from the target exhaust pollutant amount in a finite interval and a square sum of the ammonia injection amount in a finite interval, and the state feedback gain and the The exhaust gas processor control device according to claim 4, wherein the integral gain is calculated by an optimal regulator theory of a finite interval using the evaluation function.   The exhaust gas treatment device control apparatus according to claim 6, wherein control is performed using only the value of the first step among the state feedback gain and the integral gain calculated by the optimal regulator theory of a finite interval.   Means for receiving an exhaust ammonia amount contained in the exhaust gas exhausted from the exhaust gas treatment device and a target exhaust ammonia amount which is a target value of the exhaust ammonia amount, the feedback control amount being a state feedback control amount and an integral feedback 2. A control amount and an ammonia deviation integral feedback control amount obtained by multiplying a value obtained by integrating a deviation of the exhaust ammonia amount from the target exhaust ammonia amount by an integral gain for ammonia. The exhaust gas treatment device control device according to any one of claims 7 to 9.   The state feedback gain, the integral gain, and the integral gain for ammonia are the deviation of the exhaust pollutant amount from the target exhaust pollutant amount, the deviation of the exhaust ammonia amount from the target exhaust ammonia amount, the internal state amount and The exhaust gas treatment device control apparatus according to claim 8, wherein the exhaust gas treatment device control device is calculated by maximizing or minimizing an evaluation function using at least one of the ammonia injection amounts.   Feed forward control amount calculation means for calculating a feed forward control amount obtained by multiplying a process amount of the exhaust gas treatment device by a feed forward gain, and the ammonia injection amount calculation means includes the feedback control amount and the feed forward control amount. The exhaust gas treatment device control apparatus according to any one of claims 1 to 9, wherein a value obtained by adding a control amount is used as the ammonia injection amount. An exhaust gas treatment device for injecting ammonia into a gas to be treated containing at least one contaminant of nitrogen oxides and sulfur oxides to reduce the content of the pollutant in the gas to be treated;
Means for receiving the process amount of the exhaust gas treatment device, the amount of exhaust pollutant contained in the exhaust gas discharged from the exhaust gas treatment device, and the target exhaust pollutant amount which is a target value of the exhaust pollutant amount; Using the means for transmitting the ammonia injection amount injected into the target gas to the exhaust gas treatment device, the measured value of the exhaust pollutant amount , the process amount of the exhaust gas treatment device, and the ammonia injection amount, the amount of exhaust pollutant is determined. An internal state estimating means for calculating an internal state quantity estimated value obtained by estimating an internal state quantity of the exhaust gas processor to be determined; a state feedback control amount obtained by multiplying the internal state quantity estimated value by a state feedback gain; and the exhausted pollutant Feedback based on the integral feedback control amount obtained by multiplying the value obtained by integrating the deviation of the amount from the target emission pollutant amount by the integral gain A feedback control amount calculating means for calculating a control amount, and the exhaust gas treatment device control apparatus and a ammonia injection amount calculating means for calculating the ammonia injection amount based on the feedback control amount,
An exhaust gas treatment system characterized by comprising: A combustor,
An exhaust gas treatment device for injecting ammonia into a gas to be treated containing at least one pollutant of nitrogen oxide and sulfur oxide discharged from the combustor to reduce the content of the pollutant in the gas to be treated When,
Means for receiving the process amount of the exhaust gas treatment device, the amount of exhaust pollutant contained in the exhaust gas discharged from the exhaust gas treatment device, and the target exhaust pollutant amount which is a target value of the exhaust pollutant amount; Using the means for transmitting the ammonia injection amount injected into the target gas to the exhaust gas treatment device, the measured value of the exhaust pollutant amount , the process amount of the exhaust gas treatment device, and the ammonia injection amount, the amount of exhaust pollutant is determined. An internal state estimating means for calculating an internal state quantity estimated value obtained by estimating an internal state quantity of the exhaust gas processor to be determined; a state feedback control amount obtained by multiplying the internal state quantity estimated value by a state feedback gain; and the exhausted pollutant Feedback based on the integral feedback control amount obtained by multiplying the value obtained by integrating the deviation of the amount from the target emission pollutant amount by the integral gain A feedback control amount calculating means for calculating a control amount, and the exhaust gas treatment device control apparatus and a ammonia injection amount calculating means for calculating the ammonia injection amount based on the feedback control amount,
A plant with an exhaust gas treatment system characterized by comprising: An exhaust gas treatment device for controlling an exhaust gas treatment device for injecting ammonia into a treatment target gas containing at least one pollutant of nitrogen oxide and sulfur oxide to reduce the content of the pollutant in the treatment target gas. In the control method,
Receiving the process amount of the exhaust gas treatment device, the amount of exhaust pollutant contained in the exhaust gas discharged from the exhaust gas treatment device, and the target exhaust pollutant amount that is a target value of the amount of exhaust pollutant;
Transmitting an ammonia injection amount to be injected into the gas to be processed to the exhaust gas treatment device;
Using the measured value of the exhaust pollutant amount , the process amount of the exhaust gas treatment device, and the ammonia injection amount, an internal state amount estimated value obtained by estimating the internal state amount of the exhaust gas treatment device that determines the exhaust pollutant amount An internal state estimation step to be calculated;
A state feedback control amount obtained by multiplying the internal state amount estimated value by a state feedback gain, and an integral feedback control amount obtained by multiplying a value obtained by integrating a deviation of the exhaust pollutant amount from the target exhaust pollutant amount by an integral gain. A feedback control amount calculation step for calculating a feedback control amount based on the
An ammonia injection amount calculation step of calculating the ammonia injection amount based on the feedback control amount;
An exhaust gas treatment device control method comprising: An exhaust gas treatment device for controlling an exhaust gas treatment device for injecting ammonia into a treatment target gas containing at least one pollutant of nitrogen oxide and sulfur oxide to reduce the content of the pollutant in the treatment target gas. In the control program, on the computer,
A function of receiving the process amount of the exhaust gas treatment device, the amount of exhaust pollutant contained in the exhaust gas discharged from the exhaust gas treatment device, and the target exhaust pollutant amount which is a target value of the amount of exhaust pollutant;
A function of transmitting an ammonia injection amount to be injected into the processing target gas to the exhaust gas treatment device;
Using the measured value of the exhaust pollutant amount , the process amount of the exhaust gas treatment device, and the ammonia injection amount, an internal state amount estimated value obtained by estimating the internal state amount of the exhaust gas treatment device that determines the exhaust pollutant amount A function to calculate,
A state feedback control amount obtained by multiplying the internal state amount estimated value by a state feedback gain, and an integral feedback control amount obtained by multiplying a value obtained by integrating a deviation of the exhaust pollutant amount from the target exhaust pollutant amount by an integral gain. A function for calculating a feedback control amount based on
A function of calculating the ammonia injection amount based on the feedback control amount;
An exhaust gas treatment device control program characterized by realizing the above.

Images