JPH11294741A - Combustion control method in incineration plant with boiler - Google Patents

Abstract

PROBLEM TO BE SOLVED: To achieve a stabilization and an efficiency in the generated output of electric power by stabilizing the boiler outlet pressure and the flow rate of steam. SOLUTION: A boiler pressure process value PPV (t) at a boiler outlet is measured cyclically and a pressure set value is corrected according to the deviation between the pressure process value and a pressure set value PSV (t) to obtain a corrected pressure set value PSV (t)'. The corrected pressure set value is supplied to a boiler pressure adjusting meter 2 in stead of the pressure set value and based on the value, the boiler pressure adjusting meter generates an output for adjusting manipulated variables K (t) such as number of revolutions of a dust feeder to adjust the amount of a fuel to be supplied to an incinerator. The correction of the set value is executed by performing a model forecasting control based on a model such as ARX model previously generated using the pressure set value of a boiler at the boiler outlet as input and the pressure process value as output. A more proper control is possible by correcting the coefficient of the model using the corrected pressure set value and the pressure process value obtained by the measurement.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

TECHNICAL FIELD The present invention relates to general waste (municipal waste).
Control method in an incineration plant with a boiler for incineration of waste and industrial waste, etc., in particular, for stabilizing boiler pressure and boiler steam flow rate during steady operation, and thereby stabilizing power generation (power generation) The present invention relates to a combustion control method.

[0002]

2. Description of the Related Art In recent years, an incineration plant with a boiler for generating electricity by driving a steam turbine using incineration heat of municipal solid waste and industrial waste has been developed and operated. In such an incineration plant, the calorific value of municipal solid waste and industrial waste used as fuel fluctuates, and in order to control the amount of power generation to a desired value, the response delay characteristics of the plant itself are combined. And complicated control based on empirical rules and the like. In order to perform such control, in an incineration plant, a PID of one input and one output is basically used in a steady state by a distributed control system (DCS) or the like.
A controller is formed by combining a plurality of regulators, and automatic combustion control (ACC) is performed so that various process values such as a steam pressure and a steam flow rate at a boiler outlet fall within a set range (set value ± tolerance).

FIG. 5 is a block diagram for explaining a conventional PID automatic combustion control in a waste incineration plant with a boiler. In FIG. 5, reference numeral 1 denotes a boiler-equipped refuse combustion plant. The plant 1 includes a refuse incinerator, a waste heat boiler, a dust collector, and the like, and further includes a boiler outlet pressure sensor,
Various sensors such as a boiler outlet steam flow sensor and an incinerator outlet temperature sensor are included. Reference numeral 2 denotes a boiler outlet pressure (steam pressure) controller, and a boiler outlet pressure set value Psv (t) set by an operator or the like and a boiler outlet pressure process value Ppv from a boiler outlet pressure sensor in the plant 1.
(T) is input, and the amount of dust is adjusted by changing the operation amount K (t) such as the number of revolutions of the duster in a direction to make the difference between them zero. Reference numeral 3 denotes a boiler steam flow controller, which receives a boiler steam flow set value Qsv (t) set by an operator or the like and a steam flow process value Qpv (t) from a steam flow sensor at the boiler outlet. Of the steam flow control valve in the direction where
It is configured to adjust (t). Reference numeral 4 denotes another controller including a furnace top temperature control system, which is supplied with a process value from an appropriate sensor in the plant 1 and a set value corresponding to the process value. It is configured to adjust the operation amount.

[0004] These controllers 2 to 4 provide P
By performing the ID combustion control, as described above, the operation amount is automatically adjusted so that the process value falls within a target set range (set value ± permissible error), and further, control is performed so as to keep the plant 1 in an equilibrium state. doing. For example, when the boiler outlet pressure process value Ppv (t) is higher than the set value Psv (t), the boiler pressure controller 2 controls to reduce the number of rotations of the duster per unit time. As a result, the boiler heat input decreases, so that the boiler outlet pressure process value Ppv (t) decreases and the boiler outlet steam flow process value Qpv (t) tends to decrease. However, at this time, the opening of the steam flow control valve is increased by the action of the boiler steam flow controller 3 to promote a decrease in the boiler outlet pressure process value Ppv (t), and the boiler outlet steam flow process value Qpv (t). ), The plant 1 reaches an equilibrium state. When the boiler outlet pressure process value Ppv (t) falls below the set value, the operation opposite to the above is performed. In this way, adjustment is made so that each process value falls within the set range and the plant 1 is kept in an equilibrium state.

[0005]

However, even in the above-described conventional combustion control method, the boiler outlet pressure and the steam flow rate fluctuate at a period of about 10 minutes to several tens of minutes, and thus deviate from the target set range. Sometimes. As a main cause of such fluctuation, a change in waste quality such as a change in the density of the waste supplied to the incinerator can be raised. In other words, focusing on the dust pressure density at the bottom of the hopper,
The waste pressure density is highest when the refuse is put into the hopper by the crane, then gradually decreases, and then again increases at the next refuse input. Due to such a periodic change in the waste pressure density, the waste density supplied to the incinerator changes, and as a result, the heat input to the boiler per unit time is not proportional to the number of operations of the dust supply device. Become. Therefore, in the conventional combustion control method in which PID control of the boiler outlet pressure is performed while keeping the waste quality constant, the boiler outlet pressure and the steam flow rate may fluctuate, making it difficult to obtain a stable power generation amount. is there. The fluctuation of the boiler outlet pressure is also caused by a reason other than the fluctuation of the waste density, for example, a change in the water content based on the weather or the like at the time of collecting the waste, and a difference in the waste quality depending on the collection area. Furthermore, it is also caused by other disturbances. Fluctuations in boiler outlet pressure due to disturbances also occur in industrial waste incineration plants with boilers.

[0006] When the boiler outlet pressure fluctuates, the boiler steam flow rate fluctuates, and eventually, the power generation amount fluctuates. In order to suppress the fluctuation of the power generation amount due to the fluctuation of the boiler outlet pressure, in the conventional example shown in FIG. 5, the opening degree R (t) of the steam flow control valve is controlled by the boiler steam flow controller 3.
However, even if the opening of the control valve is simply adjusted, the change in heat input does not immediately appear as a change in pressure because the boiler outlet pressure is an integral system. For this reason, the fluctuation of the boiler outlet pressure is further increased, and eventually, the power generation amount may not be stabilized. Also, in a configuration in which the steam flow rate is controlled by the fuel supply amount and the boiler outlet pressure is adjusted by the control valve, for the same reason as above, the more the pressure is kept constant, the more the steam flow rate changes. In this case, the power generation amount may not be stabilized. An object of the present invention is to improve the safety and reliability of an incineration plant by reducing fluctuations in boiler outlet pressure due to such disturbances in an incineration plant with a boiler, thereby stabilizing the steam flow rate and thereby reducing the amount of power generation. An object of the present invention is to provide a combustion control method capable of stabilizing and improving efficiency.

[0007]

In order to achieve the above object, a method for controlling combustion in a boiler-mounted incineration plant according to the present invention comprises the steps of periodically measuring a boiler pressure process value at a boiler outlet; A set value correction step of obtaining a corrected pressure set value by correcting the pressure set value according to the deviation between the pressure process value and the pressure set value obtained by the pressure setting, and setting the obtained corrected pressure set value to the pressure setting The method is characterized by including a step of supplying to the boiler pressure controller instead of the value, and a fuel adjusting step of adjusting a supply amount of fuel to the incinerator based on an output of the boiler pressure controller.

In a preferred embodiment of the present invention, the set value correcting step includes a model predictive control based on a mathematical model created in advance using a boiler pressure set value at the boiler outlet as an input and a pressure process value as an output. , The pressure set value is corrected to obtain a corrected pressure set value. Further, in a preferred embodiment of the present invention, the set value correcting step further comprises using the corrected pressure set value and the pressure process value obtained by the measurement,
Correcting the coefficients of the mathematical model. Further, the fuel adjusting step is configured to adjust the number of times of driving of the dust supply means for supplying waste to the incinerator based on the output of the boiler pressure controller. Further, the mathematical model created in advance is an ARX model.

[0009]

FIG. 1 shows a functional block diagram of a system for executing a combustion control method according to first and second embodiments of the present invention. In FIG. 1, the components shown in FIG. The same components are denoted by the same reference numerals. In the combustion control method of the present invention, a set value correction amount calculating means 5 and an adding means 6 are added to the boiler pressure controller 2, and the set value correction amount calculating means 5 includes a boiler outlet pressure sensor provided at a boiler outlet. Boiler outlet pressure process value (hereinafter simply referred to as "pressure process value")
Based on Ppv (t) and a boiler outlet pressure set value (hereinafter simply referred to as “pressure set value”) Psv (t), a pressure correction value Paj (t) for the pressure set value Psv (t) is calculated,
The adding means 6 adds the pressure set value and the pressure correction value, and adds the obtained value to the corrected pressure set value Psv '.
As (t), it is supplied to the boiler pressure controller 2.

In the first embodiment of the present invention, in the set value correction amount calculating means 5, the pressure correction value Paj (t) and the corrected pressure set value Psv '(t) are calculated by the following equation (1) or It is calculated based on (2). Note that “t” represents the execution point (ie, the number of executions) of the PID control that is executed at a predetermined cycle after the correction control is started.

Paj (t) = − {Ppv (t) −Psv (t) −Paj (t−1)} (1) = −… (Ppv (t) −Psv (t)) (2) However, Σ is an addition from t = 1 to the current execution time.
(T)

Psv ′ (t) = Psv (t) + Paj (t) (3)

As is apparent from the equation (1), in the control method of the present invention, the deviation of the pressure process value Ppv (t) from the pressure set value Psv (t) is the current pressure correction value Paj.
The current pressure correction value Paj (t) is set so as to be the difference between (t) and the previous pressure correction value Paj (t-1).
The pressure correction amount is increased or decreased according to the deviation of Ppv (t) from Psv (t). Also, from the equations (2) and (3), in the control method of the present invention, the pressure process value Ppv (t) at each execution time from the start of the correction control to the present time is shown.
Of the pressure set value Psv (t) from the pressure set value Psv (t), and integrating the obtained value, and correcting the pressure set value Psv (t) at that point in the direction opposite to the integrated value by the obtained integrated value. It can be said that the completed pressure set value Psv '(t) has been obtained. Accordingly, since the pressure set value is corrected based on the history of the deviation between the pressure process value and the pressure set value, the fluctuation amplitude of the process value can be reduced.

The above-described combustion control method according to the first embodiment of the present invention intends to correct the current pressure set value based on the past history, but the second embodiment of the present invention. In the combustion control method described above, the set value correction amount calculating means 5 performs model prediction control to predict the future based on actual measurement data up to the present time, and to optimize the correction amount of the current set value. Things. Therefore, according to the second embodiment of the present invention, the ratio of the fluctuation period of the system including the plant 1 and the controllers 2 to 4 to the response time of the process value to the change of the pressure set value (variation period / response time) The effect can be exerted even when () is relatively small. Hereinafter, the model prediction control according to the second embodiment of the present invention will be described.

In the model predictive control, first,
Input u (t) = pressure set value Psv (t), output y
A system consisting of a plant and a conventional control system, where (t) = pressure process value Ppv (t) is identified, and an ARX model for predictive control represented by the following equation (4) is determined.

Equation 3] A (q) y (t) = B (q) u (t) + e (t) ... (4) However, A (q) = 1 + Σa n q -n ... (5) [Σ: n = 1 to n added for _{a] B (q) = q} -nk Σb n q -n + 1 ... (6) [Σ: n = 1~n addition of _b] e (t): disturbance t: time ( t =..., -2, -1, 0, 1, 2,...) n _a ≧ 1, n _b ≧ 1, _nk ≧ 1 q ⁻¹ : shift operator q ⁻ⁿ x (t) = x ( t−n) Here, assuming that the response time of the system is not zero and that real-time phase lag filtering is assumed, _{nk is} usually used.
Although it may be ≧ 1, it is also possible to formulate as _nk ≧ 0.

[0014] Next, performs predictive control, the current time and t _0, is assumed as the time required for one operation is negligible small, the input over a period pT _M in operation period T _M and t ₀ to the base point T _S ≧ _nk , T _E ≧ _nk , T _P = p
It is assumed that the input is not changed by means other than this control when t ≧ t ₀ as T _M −T _S + T _E. The relationship with respect to time is as shown in FIG. Then, the predicted value sequence Y _P of the output by the ARX model represented by the following equation (7) matches the predetermined reference trajectory Y _R represented by the equation (8) as much as possible, and the input change Is obtained as an input value sequence U _FM [Equation (9)] that is as small as possible. The reference trajectory Y _R is defined as a first-order lag curve having an appropriate time constant, for example.

Y _P = [y _P (t ₀ + T _S ) y _P (t ₀ + T _S +1)... Y _P (t ₀ + T _S + T _P )] ^T (7) Y _R = [y _R ( _{t 0 + T S) y R} (t 0 + T S +1) ...... y R (t 0 + T S + T P)] T ... (8) U FM = [u F (t 0) u F (t 0 + T M) _{...... u F (t 0 + pT} M)] T ... (9) and, actually input to process the first value u _F in the input value column U _FM (t _0). Thereafter, this operation is repeated for each operation cycle T _M.

[0015] Hereinafter, concrete shows how to determine the input value U _FM. First, if n _k ≧ 1, t ₀ + s based on the ARX model
Is represented as in the following equation (10).

Equation 5] _{y P (t 0 + s)} = -Σa s, n y O (t 0 -n + 1) + Σb s, n Δu F (t 0 + s-n + 1) + ΣG s, n Δu O (t 0 -n) _{+ d s u (t 0 -n} b ') ... (10) _{However, △ u F (t) =} u F (t) -u F (t-1) △ u O (t) = u O (t) - _{u O (t-1) Σ} : a s, with respect to the _n, n = 1 to n _a sum b _s for, for _{n, n = 1~ (s +} 1) adding G _s for, for _n is, n = 1 to (n _b '-1) Here, if the vectors Y ₀ , △ U _F , △ U _O and U _H are defined as in the following expressions (11) to (14), the expression ( 15)
Expression (19) holds for the matrices A to D expressed in the form of (18). Where U _H is the number of elements (T _P +
Let it be the column vector of 1).

[6] _{Y O = [y O (t} 0) y O (t 0 -1) ... y O (t 0 -n a +1)] T ... (11) △ U F = [△ u F (t 0 _{) △ u F (t 0 +1} ) ... △ u F (t 0 + pT M)] T ... (12) △ U O = [△ u O (t 0 -1) △ u O (t 0 -2) ... △ _{u O (t 0 -n b '} +1)] T ... (13) U H = [u O (t 0 -n b') u O (t 0 -n b ') ... u O (t 0 -n b ')] ^T ... (14)

(Equation 7)

Further, a column vector ｐ of the number of elements p + 1
U _FM, column vector g number of elements T _{M M} and (pT _M +1) x
The matrix G _F of (p + 1) is calculated by the following equations (20) and (21).
And (22), equation (23) holds.

(Equation 8) Thus, placing and B '= BG _F, formula (19) from equation (23) can be expressed as follows.

Y _P = −AY _O + B ′ △ U _FM + C △ U _O + DU _H (24)

Now, when the evaluation function J of the model predictive control is set as in the following equation (25), the partial differential ∂ △ J / ∂ △ U _FM of the evaluation function J is expressed by the equation (26).

Equation 10] J = ‖Y _R -Y _P ‖ ² + ^ξ‖ △ U _FM ‖ ^{_{2 = ‖Y R + AY O -B}} '△ U FM -C △ U O -DU H ‖ ² + ^ξ‖ △ U _FM ‖ ² ... (25) ^{However, ‖X‖ 2 = X T X ξ} > 0 ∂J / ∂ △ U FM = [△ ∂J / ∂ △ U 1 ∂J / ∂ △ U 2 ... ∂J / ∂ △ U p + _1] ^T = -2 However _{^{(B ') T (Y R}} + AY O -B' △ U FM -C △ U O -DU H) + 2 △ ξU FM ... (26), △ U i is the △ U _FM I-th component

Here, assuming that ∂J / ∂ △ U _FM = 0,

Equation 11] _{△ U FM = {(B '} ) T B' + ξE} -1 (B ') T (Y R + AY O -C △ U O -DU H) ... (27) However, E is a unit matrix is obtained, the output of the predicted value sequence Y _P reference trajectory Y _R close to, and the change in the input value column ‖ △ U _FM ‖ such decrease △ U _FM
Is obtained, and thus the input value sequence _UFM is obtained.
Note that the obtained input value sequence U _FM is a boiler pressure set value to be directly supplied to the boiler pressure control system 2 as described in relation to the equations (4) to (6). the value of each element of the input value column U _FM has becomes the corrected pressure setpoint Psv at the time the corresponding '(t). Therefore, the set value calculating means 5 shown in FIG. 1 calculates the pressure correction value Paj (t) by subtracting the set value Psv (t) at the corresponding time from the value of the first element of the obtained input value sequence. Is supplied to the adding means 6 at that time. Further, without using the addition means 6, the value itself of the first element of the obtained input value sequence U _FM may be fed directly to the boiler pressure control system 2.

Next, a description will be given of a combustion control method according to a third embodiment of the present invention. The third embodiment is based on Equations (4) to (4).
The pressure correction amount P based on the ARX model represented by (6)
When seeking aj (t), the coefficient of ARX model a _n and b _n
Can be adjusted. For this reason, as shown in FIG. 3, a prediction calculation model coefficient correction means 7 is added, and the correction value obtained in the operating state in which the set value correction amount calculation means 5 and the addition means 6 are operated in the coefficient correction means 7. based on the already pressure setpoint Psv '(t) and pressure process value Ppv (t), is configured to correct by learning the coefficients a _n and b _n of the ARX model. In the learning of the coefficient, the model prediction value calculated at the time point t ₀ is represented as y _p (t), and the pressure process value y (t) is used as a teacher signal within the range of t ₀ ≦ t ≦ t ₀ + T. Predicted value y _p
Correct the coefficients so that (t) is improved. Now, the evaluation function

Equation 12] _{J = Σ {y (t 0} + s) -y p (t 0 + s)} 2/2 ... (28) However, sigma: if sum of s = 0 to T, the coefficient a _n and b _n of correction fee △ a _n and △ b _n is,
It is expressed as follows.

Equation 13] _{△ a n = -ε∂J / ∂a n} = εΣ {y (t 0 + s) -y p (t 0 + s)} p n (t 0 + s) ... (29) △ b n = - _{ε∂J / ∂b n = εΣ {y} (t 0 + s) -y p (t 0 + s)} q n (t 0 + s) ... (30) However, epsilon: learning rate sigma: for s = 0 to t summing _{p n (t) = ∂y p} (t) / ∂a n (1 ≦ n ≦ n a) q n (t) = ∂y p (t) / ∂b n (1 ≦ n ≦ n b) Actually, one set is sequentially selected from a plurality of sets of teacher data, and learning of coefficients is repeated until a preset convergence criterion is satisfied. In the above model predictive control, AR
Although the X model has been described, it goes without saying that another mathematical model may be used.

FIG. 4 shows the result of executing the combustion control method according to the second embodiment of the present invention by an actual machine test (set value of pressure Psv (t) = constant). Fig. 3 is a graph showing the test results of the method in comparison with the test results. 4A shows the boiler pressure at the boiler outlet, FIG. 4B shows the boiler steam flow rate, and the time T
0 to T1 and T2 to T3 indicate pressure process values obtained by the conventional example, and times T1 to T2 indicate pressure process values obtained by the present invention. As is clear from these graphs, according to the present invention, the fluctuation from the set value of the pressure process value can be extremely reduced and the fluctuation of the steam flow rate of the boiler can be suppressed as compared with the conventional example. It has been demonstrated that

Since the present invention is configured as described above, it is possible to sufficiently reduce the fluctuation of the boiler pressure by correcting the set value according to the deviation between the boiler outlet pressure process value and the pressure set value. Therefore, the safety and reliability of the plant can be improved. In addition, this can stabilize the boiler steam flow,
The amount of power generation can be stabilized and efficiency can be improved. In addition, if model predictive control is adopted to determine the correction amount for correcting the boiler pressure set value, the ratio between the system fluctuation cycle and the response time of the process value to the change in the boiler pressure set value is relatively small. Can be dealt with, and the control performance such as stabilizing the boiler pressure and the steam flow rate is improved. Furthermore, if a function of correcting the model coefficient in the model predictive control is provided, the power generation amount can be stabilized without deteriorating the control performance irrespective of the aging of the incinerator.

[Brief description of the drawings]

FIG. 1 is a block diagram for explaining first and second combustion control methods of the present invention.

FIG. 2 is a diagram for explaining a time factor relationship in a combustion control method according to a second embodiment of the present invention.

FIG. 3 is a block diagram for explaining a combustion control method according to a third embodiment of the present invention.

FIG. 4 is a graph showing a result of an actual machine test performed using the combustion control method according to the second embodiment of the present invention in comparison with a result of a conventional example.

FIG. 5 is a block diagram for explaining a conventional combustion control method.

Claims (5)

[Claims] 1. A method for controlling combustion in a boiler-equipped incineration plant, comprising the steps of: periodically measuring a boiler pressure process value at a boiler outlet; A set value correction step of obtaining a corrected pressure set value by correcting the pressure set value; a step of supplying the obtained corrected pressure set value to the boiler pressure controller instead of the pressure set value; A fuel adjusting step of adjusting the amount of fuel supplied to the incinerator based on the output of the meter. 2. The combustion control method according to claim 1, wherein
The set value correction step corrects and corrects the pressure set value by performing model predictive control based on a mathematical model created in advance with the boiler pressure set value at the boiler outlet as input and the pressure process value as output. A method for controlling the combustion pressure, the method comprising: obtaining a set pressure value. 3. The combustion control method according to claim 2, wherein
The combustion control method, further comprising the step of correcting the coefficient of the mathematical model using the corrected pressure set value and the pressure process value obtained by the measurement. 4. The combustion control method according to claim 2, wherein the mathematical model created in advance is an ARX model. 5. The combustion control method according to claim 1, wherein the fuel adjusting step includes the step of driving the dust supply means for supplying waste to the incinerator based on the output of the boiler pressure controller. Controlling the combustion.