JP2019108995A - Combustion condition determining device of combustion furnace, combustion condition determining method, and combustion system - Google Patents

Abstract

To provide a combustion condition determining device of a combustion furnace, capable of quickly and accurately determining combustion conditions.SOLUTION: The combustion condition determining device of a combustion furnace for determining combustion conditions including a plurality of combustion control parameters for controlling a combustion device of the combustion furnace, comprises: a prior-to-combustion fuel component acquiring part that acquires components of fuel prior to combustion including at least a part of measured values of the components of the fuel prior to combustion, i.e., fuel to be fed to a burner that feeds fuel and air into the combustion furnace; and a command value determining part that determines a command value of at least one of the combustion control parameters so that an index to be evaluated, i.e., at least one of unburned combustibles in ash generated by combusting the fuel prior to combustion and an NOx concentration in exhaust gas, satisfies prescribed criteria on the basis of the components of the fuel prior to combustion.SELECTED DRAWING: Figure 1

Description

  The present disclosure relates to a method of determining combustion conditions of a combustion furnace such as a boiler.

  Normally, in a coal-fired boiler, combustion adjustment is performed using several types of coal having different fuel properties at the time of trial operation. Then, through this combustion adjustment, the combustion control parameters are adjusted so that the reduction of NOx concentration in the exhaust gas and the reduction of unburned component (unburned component in ash) remaining in ash after combustion of coal are achieved, A determination of combustion conditions is made. For example, the combustion control parameters include the opening of the damper for adjusting the flow rate of combustion air supplied to the burner for combustion, the burner nozzle angle, the angle of the AA port for supplying additional air (AA) for NOx reduction inside the boiler, etc. It is. And the setting parameter (combustion control parameter) of equipment (combustion apparatus) for controlling such a combustion state is less than the guaranteed value where NOx concentration satisfies environmental standard, and the unburned part in ash is a material such as concrete material It is determined to be less than the specified value (such as 5% or less) required to be usable as a thing.

On the other hand, techniques for appropriately adjusting the combustion conditions during the operation of the boiler after the start of operation of the boiler have also been proposed (Patent Documents 1 to 3). For example, in Patent Document 1, complete combustion is achieved by measuring the O ₂ concentration and the CO ₂ concentration in the exhaust gas in real time by the semiconductor laser absorption method (TDLAS) and controlling the air ratio (ratio of fuel to air). It is disclosed to minimize excess air while maintaining the condition. Similarly, Patent Documents 2 and 3 disclose a system for measuring unburned carbon in ash that measures unburned carbon in ash in a boiler in real time using a laser induced breakdown method (LIBS method).

Patent No. 6135731 Japanese Patent Application Publication No. 2003-4634 Patent No. 4119624

  By the combustion adjustment at the time of the trial operation of the combustion furnace such as the boiler as described above, it is possible to operate the combustion device under the optimum combustion condition for the fuel to be used. However, since it is necessary to raise or lower those values for a plurality of combustion control parameters to search for an optimal operating condition, costs such as time for that are required. In addition, after the trial operation, NOx emissions may change by about 20 ppm even though coal of the same brand is used in the operation of 100 ppm of NOx. In addition, the combustion performance of the combustion furnace may change as ash generated by the combustion of the fuel adheres to the inside of the furnace such as the burner nozzle. It is difficult in operation to perform combustion adjustment every time such a case occurs, and it is difficult to maintain an optimal combustion state.

  In view of the above-described circumstances, at least one embodiment of the present invention aims to provide a combustion condition determination device for a combustion furnace that can determine combustion conditions quickly and accurately.

(1) A combustion condition determination device for a combustion furnace according to at least one embodiment of the present invention,
A combustion condition determination device for a combustion furnace that determines a combustion condition including a plurality of combustion control parameters for controlling a combustion apparatus of the combustion furnace, the combustion condition determining apparatus comprising:
A pre-combustion fuel component acquisition unit for acquiring a pre-combustion fuel component including a measurement value of at least a part of the fuel component of the pre-combustion fuel that is the fuel supplied to the burner that supplies the fuel and air into the furnace of the combustion furnace When,
Based on the pre-combustion fuel component, at least one evaluation target index, which is at least one of an unburned portion in ash of ash produced by the combustion of the pre-combustion fuel or NOx concentration in exhaust gas, satisfies a predetermined criterion. And a command value determination unit that determines a command value of the combustion control parameter.

  According to the configuration of the above (1), in the ashes of ash produced when the pre-combustion fuel is actually burned based on the component (pre-combustion fuel component) of the fuel (pre-combustion fuel) supplied to the burner For example, the opening degree of the damper that adjusts the combustion air flow rate in each burner, burner nozzle angle, classifier rotation speed when pulverizing solid fuel by the milling device, etc. such that unburned components and NOx concentration in exhaust gas satisfy the criteria The command value of the combustion control parameter of the combustion apparatus which controls the combustion of is determined. As described above, by focusing on the pre-combustion fuel component before being supplied to the burner, it is possible to quickly and accurately determine the combustion control parameter suitable for the fuel property.

(2) In some embodiments, in the configuration of (1) above,
The command value determination unit
A provisional setting value acquisition unit that acquires a provisional setting value of the combustion control parameter;
An evaluation target index prediction unit that calculates a predicted value of the evaluation target index based on the temporary setting value and the pre-combustion fuel component each time the temporary setting value acquisition unit acquires the temporary setting value;
A criteria determination unit that determines whether the predicted value satisfies the predetermined criteria;
The temporary setting value of the combustion control parameter acquired by the temporary setting value acquiring unit is changed and the temporary setting is changed until it is determined that the predicted value satisfies the predetermined criterion by the criteria determining unit. A parameter adjustment unit that causes the temporary setting value acquisition unit to acquire a value;
And a command value selection unit that uses the temporary set value as the command value when it is determined by the criteria determination unit that the predicted value satisfies the predetermined criteria.

  According to the configuration of the above (2), the unburned portion in the ashes of ash and NOx which will be produced when the fuel before combustion is burned based on the temporary setting value of the combustion control parameter and the fuel component before combustion While predicting the concentration and changing the temporary setting value of the combustion control parameter, a temporary setting value satisfying the criteria for the prediction result is found and used as the command value of the combustion control parameter. By this, it is possible to determine the command value of the combustion control parameter such that the unburned portion in the ash of the ash generated when the fuel before combustion is actually burned and the NOx concentration in the exhaust gas satisfy the criteria.

(3) In some embodiments, in the configuration of (2) above,
The evaluation target index prediction unit uses the flammability index and the temporary setting value using a flammability index indicating the combustibility of the pre-combustion fuel, and a prediction map indicating the relationship between the evaluation target index and the combustion control parameter. The predicted value is calculated from
According to the configuration of the above (3), the predicted value of the evaluation target indicator such as the unburned component in ash and the NOx concentration is calculated in advance based on the prediction map which can be acquired by experiment or the like. Thus, the command value of the combustion control parameter can be predicted quickly and accurately.

(4) In some embodiments, in the configuration of (2) above,
The command value determination unit may measure the measured value of the evaluation target index generated by the combustion of the pre-combustion fuel component and the pre-combustion fuel component including the pre-combustion fuel component, and the combustion when the pre-combustion fuel is burned The prediction value is calculated from the pre-combustion fuel component and the temporary setting value using a prediction model created by machine learning teacher data composed of a plurality of data associated with the command value of the control parameter. Do.
According to the configuration of the above (4), the prediction value of the command value of the combustion control parameter is generated by machine learning the relationship between the pre-combustion fuel component and the measured value of the evaluation target index at the time of the combustion Predict. This makes it possible to easily predict an appropriate command value of the combustion control parameter.

(5) In some embodiments, in the configuration of (4) above,
The information processing apparatus further includes a relearning determination unit that determines whether to perform relearning based on a difference between the predicted value of the evaluation target index and the measurement value of the evaluation target index.
According to the configuration of the above (5), when the predicted value of the fuel evaluation target indicator exceeds the allowable range and is different from the measured value, etc., it is determined that relearning is necessary. By performing relearning in accordance with this determination, it is possible to maintain prediction accuracy using a prediction model created based on machine learning.

(6) In some embodiments, in the configuration of (1) above,
The command value determination unit is a measurement value of the evaluation target index generated by the combustion of the pre-combustion fuel component including the pre-combustion fuel component and the pre-combustion fuel component, and the measurement value satisfying the predetermined criteria The combustion using a command value determination model created by machine learning of teacher data composed of a plurality of data associated with the command value of the combustion control parameter when the pre-combustion fuel is burned The command value is determined from the pre-fuel component.
According to the configuration of the above (6), the command value of the combustion control parameter is the measurement value of the pre-combustion fuel component and the evaluation target index at the time of the combustion, and the measurement value satisfying the predetermined criteria The relationship with the control parameter is determined using a command value determination model created by machine learning. This makes it possible to easily determine an appropriate command value of the combustion control parameter.

(7) In some embodiments, in the configurations of (1) to (6) above,
A driving state evaluation index acquisition unit that acquires measurement values of the driving state evaluation index including the evaluation target index;
An adjustment value determination unit that determines an adjustment value for adjusting the command value of the combustion control parameter determined by the command value determination unit based on the measured value of the driving state evaluation index and the pre-combustion fuel component; And further comprising
According to the configuration of the above (7), the command value of the combustion control parameter already determined is adjusted in accordance with the operating condition of the combustion furnace based on the measurement value of the operating condition evaluation indicator such as the evaluation object indicator. Thereby, the command value of the combustion control parameter can be adjusted to a more appropriate value according to the actual situation.

(8) In some embodiments, in the configuration of (7) above,
The driving state evaluation index includes the evaluation target index,
The driving state evaluation index acquisition unit acquires a measurement value of the evaluation target index,
The adjustment value determination unit has an evaluation object index adjustment value calculation unit that adjusts a command value of the combustion control parameter based on the measurement value of the evaluation object index and the pre-combustion fuel component.
According to the configuration of (8), the adjustment value of the combustion control parameter is calculated based on the measurement value of the evaluation target index. Thereby, the command value of the combustion control parameter can be adjusted to a more appropriate value according to the actual situation.

(9) In some embodiments, in the configurations of (7) to (8) above,
The operating condition evaluation index includes at least one of the unburned component in the ash or the NOx concentration,
The operation state evaluation index acquisition unit acquires at least one of the measurement value of the unburned ash content in the ash or the NOx concentration,
The adjustment value determination unit determines, when the measured value of at least one of the unburned ash content and the NOx concentration is larger than a planned value, between the tip of the burner and the ignition position of the fuel in the furnace. It has an ignition distance adjustment value calculation part which calculates the adjustment value which adjusts the said combustion control parameter so that an ignition distance may become short.
According to the configuration of the above (9), when the measured value is larger than the planned value, the measured value is made equal to or less than the planned value with respect to the unburned component in ash and NOx concentration of the ash generated after the combustion of the fuel. The adjustment value of the combustion control parameter is calculated such that the ignition distance between the burner and the ignition position is reduced. Thereby, the command value of the combustion control parameter can be adjusted to a more appropriate value according to the actual situation.

(10) In some embodiments, in the configurations of (7) to (9) above,
The operating condition evaluation index includes the unburned portion in the ash, and the amount of the ash attached to the height position of the burner on the wall in the furnace,
The operation state evaluation index acquisition unit acquires the measurement value of the unburned portion in ash and the measurement value of the adhesion amount of the ash,
If the measured value of the unburned part in ash is higher than the unburned part in ash planned value and the adhered amount of the ash is smaller than the adhered amount planned value, the adjustment value determination unit And a second ignition distance adjustment value calculation unit that calculates the adjustment value for adjusting the combustion control parameter so that the ignition distance between the tip of the burner and the ignition position of the fuel becomes short.
According to the structure of said (10), the adjustment value of a combustion control parameter is calculated in consideration of the adhesion condition of said ash further. By this, the command value of the combustion control parameter can be adjusted to a more appropriate value according to the actual situation.

(11) In some embodiments, in the configurations of (7) to (10) above,
The operating condition evaluation index includes the particle size of the pulverized fuel which is the pre-combustion fuel generated by a mill, or the fuel particle size of the pre-combustion fuel supplied from the burner into the furnace.
The operation state evaluation index acquisition unit acquires a measurement value of the fuel particle size,
The adjustment value determination unit has a particle size adjustment value calculation unit that calculates an adjustment value for adjusting the command value of the combustion control parameter based on the measurement value of the fuel particle size.
According to the configuration of (11), the measurement value of the fuel particle size of the pre-combustion fuel before being supplied into the furnace or the measurement of the fuel particle size of the unburned fuel of the pre-combustion fuel supplied into the furnace Based on the value, the adjustment value of the combustion control parameter is calculated. Thereby, the command value of the combustion control parameter can be adjusted to a more appropriate value according to the actual situation.

(12) A combustion system according to at least one embodiment of the present invention,
The combustion condition determination device for a combustion furnace according to any one of claims 1 to 11, which determines a combustion condition including a plurality of combustion control parameters for controlling a combustion apparatus of the combustion furnace.
A measuring device capable of measuring in real time an operation state evaluation index for evaluating the operation state of the combustion furnace;
And a combustion control device for transmitting the command value of the at least one combustion control parameter determined by the combustion condition determination device to the combustion device.
According to the structure of said (12), the same effect as said (1)-(11) is show | played.

(13) A method of determining a combustion condition of a combustion furnace according to at least one embodiment of the present invention,
A combustion condition determination method for a combustion furnace, comprising: determining a combustion condition including a plurality of combustion control parameters for controlling a combustion apparatus of the combustion furnace, the combustion condition determining method comprising:
A pre-combustion fuel component acquisition step of acquiring a pre-combustion fuel component including a measurement value of at least a part of a fuel component of the pre-combustion fuel which is the fuel supplied to a burner supplying the fuel and air into the furnace of the combustion furnace When,
Based on the pre-combustion fuel component, at least one evaluation target index, which is at least one of an unburned portion in ash of ash produced by the combustion of the pre-combustion fuel or NOx concentration in exhaust gas, satisfies a predetermined criterion. And a command value determining step of determining a command value of the combustion control parameter.

  According to the structure of said (13), the same effect as said (1) is show | played.

(14) In some embodiments, in the configuration of (13) above,
The command value determination step is
A provisional setting value acquisition step of acquiring a provisional setting value of the combustion control parameter;
An evaluation target index prediction step of calculating a predicted value of the evaluation target index based on the temporary setting value and the pre-combustion fuel component each time the temporary setting value is obtained by the temporary setting value obtaining step;
A criteria determination step of determining whether the predicted value satisfies the predetermined criteria;
The provisional setting value of the combustion control parameter acquired by the provisional setting value acquisition step is changed and the provisional setting is changed until it is determined that the predicted value satisfies the predetermined criterion in the criteria determination step. A parameter adjustment step of causing the temporary set value acquisition step to acquire a value;
A command value determination step of using the temporary set value as the command value when it is determined that the predicted value satisfies the predetermined criteria in the criteria determination step.

  According to the structure of said (14), there exists an effect similar to said (2).

(15) In some embodiments, in the configuration of (14),
The evaluation target index prediction step uses the combustion index and the temporary setting value by using a flammability index indicating the combustibility of the pre-combustion fuel, and a prediction map indicating the relationship between the evaluation target index and the combustion control parameter. The predicted value of the evaluation target index is calculated from
According to the structure of said (15), the same effect as said (3) is show | played.

(16) In some embodiments, in the configuration of (14),
In the command value determination step, the measured value of the evaluation target index generated by the combustion of the pre-combustion fuel component and the pre-combustion fuel component having the pre-combustion fuel component, and the above-described combustion component The command value is calculated from the pre-combustion fuel component and the temporary setting value using a prediction model created by machine learning teacher data composed of a plurality of data associated with the command value of the combustion control parameter. decide.
According to the configuration of the above (16), the same effect as the above (4) can be obtained.

(17) In some embodiments, in the configuration of (16),
The information processing apparatus further includes a relearning determination step of determining whether relearning is to be performed based on a difference between the predicted value of the evaluation target index and the measurement value of the evaluation target index.
According to the structure of said (17), the same effect as said (5) is show | played.

(18) In some embodiments, in the configuration of (13) above,
The command value determining step is a measurement value of the evaluation target index that is generated by the combustion of the pre-combustion fuel component having the pre-combustion fuel component and the pre-combustion fuel component, and the measurement value satisfying the predetermined criteria Using a command value determination model created by machine learning of teacher data composed of a plurality of data associated with the command value of the combustion control parameter when the pre-combustion fuel component is burned; The command value is determined from the pre-combustion fuel component.
According to the configuration of the above (18), the same effect as the above (6) is exerted.

(19) In some embodiments, in the above configurations (13) to (18),
A driving state evaluation index acquisition step of acquiring a measurement value of the driving state evaluation index including the evaluation target index;
An adjustment value determination step of determining an adjustment value for adjusting the command value of the combustion control parameter determined by the command value determination step based on the measured value of the operating condition evaluation index and the pre-combustion fuel component; And further comprising
According to the configuration of the above (19), the same effect as the above (7) is exerted.

(20) In some embodiments, in the configuration of (19) above,
The driving state evaluation index includes the evaluation target index,
The driving state evaluation index acquisition step acquires a measurement value of the evaluation target index,
The adjustment value determination step includes an evaluation object index adjustment value calculation step of adjusting a command value of the combustion control parameter based on the measurement value of the evaluation object index and the pre-combustion fuel component.
According to the configuration of the above (20), the same effect as the above (8) is exerted.

(21) In some embodiments, in the above configurations (19) to (20),
The operating condition evaluation index includes at least one of the unburned component in the ash or the NOx concentration,
The operating state evaluation index acquiring step acquires at least one of measured values of the unburned component in the ash or the NOx concentration,
The adjustment value determination step is performed between the tip of the burner and the ignition position of the fuel in the furnace if the measurement value of at least one of the unburned ash content and the NOx concentration is larger than a planned value. An ignition distance adjustment value calculation step of calculating an adjustment value for adjusting the combustion control parameter so as to shorten the ignition distance.
According to the configuration of the above (21), the same effect as the above (9) is exerted.

(22) In some embodiments, in the above configurations (19) to (21),
The operating condition evaluation index includes the unburned portion in the ash, and the amount of the ash attached to the height position of the burner on the wall in the furnace,
The operation state evaluation index acquiring step acquires a measurement value of the unburned portion in the ash and a measurement value of the adhesion amount of the ash;
In the adjustment value determination step, when the measured value of the unburned part in ash is higher than the unburned part in ash planned value and the adhered amount of the ash is smaller than the planned amount of adhered amount, the inside of the furnace A second ignition distance adjustment value calculation step of calculating the adjustment value for adjusting the combustion control parameter so that the ignition distance between the tip of the burner and the ignition position of the fuel in the case becomes short.
According to the configuration of the above (22), the same effect as the above (10) can be obtained.

(23) In some embodiments, in the configurations of (19) to (22),
The operating condition evaluation index includes the particle size of the pulverized fuel which is the pre-combustion fuel generated by a mill, or the fuel particle size of the pre-combustion fuel supplied from the burner into the furnace.
The operation state evaluation index acquisition step acquires a measurement value of the fuel particle size,
The adjustment value determination step has a particle size adjustment value calculation step of calculating an adjustment value for adjusting the command value of the combustion control parameter based on the measurement value of the fuel particle size.
According to the configuration of the above (23), the same effect as the above (11) can be obtained.

  According to at least one embodiment of the present invention, a combustion condition determination device capable of determining combustion conditions quickly and accurately is provided.

It is a figure showing roughly a combustion system provided with a combustion condition determination device of a combustion furnace concerning one embodiment of the present invention. It is a figure which refines and shows the command value determination part concerning one embodiment of the present invention, and determines a command value of a combustion control parameter based on a predicted value of an evaluation object index. It is a figure which refines and shows the command value determination part concerning one embodiment of the present invention, and determines a command value of a combustion control parameter using a command value determination model. It is a figure which refines and shows the adjustment value determination part which concerns on one Embodiment of this invention. It is a figure which refines and shows the evaluation object parameter | index adjustment value calculation part which concerns on one Embodiment of this invention. It is a flowchart which shows the combustion condition determination method of the combustion furnace which concerns on one Embodiment of this invention, and determines the command value of a combustion control parameter based on the predicted value of evaluation object parameter | index. It is a flowchart which shows the combustion condition determination method of the combustion furnace which concerns on one Embodiment of this invention, and determines the command value of a combustion control parameter using a prediction model. It is a flowchart which shows the relearning determination step of the prediction model which concerns on one Embodiment of this invention. It is a flowchart which shows the relearning determination step of the command value determination model which concerns on one Embodiment of this invention.

Hereinafter, some embodiments of the present invention will be described with reference to the accompanying drawings. However, the dimensions, materials, shapes, relative arrangements, etc. of the components described as the embodiments or shown in the drawings are not intended to limit the scope of the present invention to this, but are merely illustrative. Absent.
For example, a representation representing a relative or absolute arrangement such as “in a direction”, “along a direction”, “parallel”, “orthogonal”, “center”, “concentric” or “coaxial” is strictly Not only does it represent such an arrangement, but also represents a state of relative displacement with an angle or distance that allows the same function to be obtained.
For example, expressions that indicate that things such as "identical", "equal" and "homogeneous" are equal states not only represent strictly equal states, but also have tolerances or differences with which the same function can be obtained. It also represents the existing state.
For example, expressions representing shapes such as quadrilateral shapes and cylindrical shapes not only represent shapes such as rectangular shapes and cylindrical shapes in a geometrically strict sense, but also uneven portions and chamfers within the range where the same effect can be obtained. The shape including parts etc. shall also be expressed.
On the other hand, the expressions "comprising", "having", "having", "including" or "having" one component are not exclusive expressions excluding the presence of other components.

  FIG. 1 is a view schematically showing a combustion system 7 provided with a combustion condition determination device 1 of a combustion furnace 8 according to an embodiment of the present invention. The combustion system 7 shown in FIG. 1 is a pulverized coal-fired boiler system. More specifically, as shown in FIG. 1, the combustion system 7 includes a burner 91 (generally a plurality of burners) in a combustion chamber 8 f (in the furnace, the same shall apply hereinafter) formed inside a boiler which is the combustion furnace 8. By supplying pulverized coal fuel and air through the fuel chamber and burning the fuel (pre-combustion fuel F) in the combustion chamber 8f, the fluid such as water flowing in the heat transfer tube group 82 of the heat exchanger is Configured to heat. Also, in general, the boiler has a rectangular cross section in the horizontal direction in the installed state, and the burner 91 has a predetermined portion (corner portion) or four wall surface portions respectively including each of the four corners , And a plurality of stages are arranged above and below the boiler (in a direction perpendicular to the horizontal direction). For example, the combustion system 7 may be a hot water supply system that heats the fluid of the heat transfer tube group 82 and supplies hot water, or a turbine (not shown) is generated by steam generated by the heating of the fluid of the heat transfer tube group 82. It may be a power generation system that drives to generate power.

  In the embodiment shown in FIG. 1, the coal fuel stored in the coal storage facility 71 is supplied to the milling apparatus 92 through a coal hopper 72 and a coal supply apparatus 73 (such as a screw feeder). The milling apparatus 92 is adapted to grind the supplied coal fuel to a desired particle size (for example, about several μm to about several hundred μm). The mill apparatus 92 is also connected to the burner 91 installed in the boiler via the pulverized fuel pipe Lf, and the pulverized fuel (pulverized coal) is milled by the force of the primary air A1 (air for transfer). It is supplied from the device 92 to the burner 91. Then, the fuel is supplied (inputted) from the burner 91 to the combustion chamber 8f of the boiler together with the air to be burned. At this time, additional air (AA) preheated by a normal temperature or an air preheater 77 (described later) is supplied (charged) to a gas (fuel area) generated at the time of combustion by the burner 91 from the AA port 93. Stage combustion is to be performed. The AA port 93 is provided above the burner 91 (see FIG. 1), and supplies the amount of AA determined by the two-stage combustion rate into the furnace under flow control by the AA amount adjustment valve 93a. As a result, the NOx generated at the burner 91 is reduced. The two-stage combustion rate is calculated by the amount of air for combustion (the amount of AA) supplied from the AA port 93 / the amount of air for total combustion. The amount obtained by subtracting the AA amount from the total combustion air amount is supplied from the burner 91 side. For example, if the two-stage combustion rate rises, the amount of combustion air supplied from the AA port 93 increases and the amount of combustion air supplied from the burner 91 decreases, so the space up to AA becomes insufficient for air NOx is reduced.

  On the other hand, the exhaust gas G generated by the combustion of the fuel in the boiler is discharged to the outside through the exhaust gas pipe 75 connected to the flue 8p of the above-described boiler. In the exhaust gas pipe 75, the temperature of the outside air A passing through the air supply pipe L (described later) is raised to, for example, 200 ° C. to 300 ° C. by the heat of the exhaust gas G and the denitration device 76 for removing nitrogen oxides from the exhaust gas G. Air preheater 77 (described later), an electric precipitator 78 for removing dust contained in the exhaust gas G after heat recovery, a desulfurizer for removing sulfur oxides in the exhaust gas G after dust removal (not shown), etc. The apparatus which processes the waste gas G is installed, and the waste gas G is discharged | emitted from the chimney (not shown) to the exterior through the process by these apparatuses. Further, the air supply pipe L described above is branched downstream of the air preheater 77 into a transfer air supply pipe L1 connected to the mill device 92 and a combustion air supply pipe L2 connected to the boiler. As a result, the outside air A is supplied from the conveying air supply pipe L1 as the primary air A1 to the mill device 92 and as the secondary air A2 (combustion air) under the flow adjustment by the air box damper 94. After being supplied to the air box 83 from the supply pipe L2, the air is supplied to the inside of the boiler (in the furnace) via the burner 91.

  The combustion system 7 is a boiler system that can operate using two or more types of fuel, such as a biomass / coal mixed combustion system that operates using recycled fuel such as biomass fuel and fossil fuel such as coal fuel. It may be. A biomass fuel is a fuel which uses organic resources derived from renewable organisms, such as woody biomass such as wood chips, excluding fossil resources, as a raw material. The recycled fuel is a fuel that uses the above-mentioned woody biomass, waste tires, sludge, Refuse Paper and Plastic Fuel (RPF), and the like as raw materials. The plurality of types of fuel may include at least one of the above-described recycled fuels, or may include a plurality of types of fuel related to coal such as high grade coal and low grade coal.

  And as shown in FIG. 1, the combustion system 7 is provided with the combustion condition determination apparatus 1 (only henceforth the combustion condition determination apparatus 1) of the combustion furnace 8 which is a boiler etc. FIG. The combustion condition determination device 1 is a device for determining a combustion condition including a plurality of combustion control parameters P for controlling the combustion device 9 of the combustion furnace 8, and at least one of the plurality of combustion control parameters P among them. One command value Pv (hereinafter referred to simply as the command value Pv) is determined. More specifically, the combustion condition determination device 1 determines the unburned portion (unburned portion in the ash) remaining in the ash produced by the combustion of the fuel in the combustion furnace 8 (the combustion chamber 8f) and the exhaust gas G produced by the combustion thereof. The command value Pv of the combustion control parameter P is determined such that the concentration of NOx contained in the engine satisfies the criteria Tc (reference value) defined for each of them.

  The above-mentioned combustion device 9 is a device (equipment and equipment) for controlling the combustion (combustion state) of the combustion furnace 8. For example, the burner 91, the mill device 92, the AA port 93, and the AA amount adjustment valve described above 93a, wind box damper 94 and the like. The combustion control parameter P is a settable parameter that can be adjusted (changed) for controlling the combustion device 9. For example, a parameter for setting an injection angle of fuel such as a burner nozzle angle of the burner 91, an angle of the AA port 93, etc. These parameters include a parameter for setting the input angle of AA (AA angle), a two-stage combustion ratio (opening of the AA amount adjustment valve 93a), and a parameter for setting the opening of the air box damper 94. In addition, the criteria Tc for the unburned component in ash may be, for example, that the unburned component in ash becomes equal to or less than a specified value (for example, 5% or less). The criterion Tc for the NOx concentration may be that the NOx concentration is less than or equal to the guaranteed value satisfying the environmental standard.

Hereinafter, the above-described combustion condition determination device 1 will be described in detail.
As shown in FIG. 1, the combustion condition determination device 1 of the combustion furnace 8 includes a pre-combustion fuel component acquisition unit 2 and a command value determination unit 3. The combustion condition determination device 1 is configured by a computer, and includes a CPU (processor) (not shown), a memory such as a ROM or RAM, a storage device m such as an external storage device, a communication interface, and the like. Then, the CPU operates (eg, calculates data) according to an instruction of a program (combustion condition determination program) loaded to the main storage device, thereby realizing each of the above functional units. Each of the above-described functional units included in the combustion condition determination device 1 will be described.

  The pre-combustion fuel component acquisition unit 2 is a pre-combustion fuel F that is a fuel (coal in FIG. 1) supplied to the burner 91 that supplies (inputs) fuel and air into the furnace of the combustion furnace 8 (boiler in FIG. 1). The pre-combustion fuel component Cp including at least a part of the measurement value of the fuel component C of The pre-combustion fuel F is a fuel before the combustion by the combustion furnace 8 is performed. Specifically, the pre-combustion fuel F is a fuel located upstream of the burner 91, such as the coal storage facility 71, the coal hopper 72, the coal supply device 73, the milling device 92, and the pulverized fuel pipe Lf. The pre-combustion fuel F may be a fuel such as a sample product obtained from another place other than the above-mentioned equipment (equipment). The fuel component is the amount and concentration (hereinafter referred to as the amount) of each element contained in the fuel, and the pre-combustion fuel component Cp is carbon (C), nitrogen (N), hydrogen (H), oxygen ( O), measurement values of at least one amount of Ca (calcime), K (potassium), Fe (iron), Na (sodium) and the like which constitute ash after combustion are included.

  As described above, the reason for acquiring the pre-combustion fuel component Cp is that the fuel component affects the unburned component in the ash and the NOx concentration. More specifically, the fuel component C of the fuel provides information on the flammability of the fuel (flammable, not flammable). In particular, the carbon content (C content), the molar ratio of hydrogen to carbon (H / C), the molar ratio of oxygen to carbon (O / C), and the nitrogen content (N content) It becomes an indicator (flammability index Ib) to show. Specifically, when the amount of C is large, the fuel tends to be hard to burn. If the amount of H or O is larger than the amount of C (H / C, O / C is large), the fuel is easy to burn, and conversely, the amount of H or O is small (H / C, O / C is small) And the fuel tends to be hard to burn. Also, the larger the amount of N, the less likely the fuel will burn. In addition, ash components such as Ca, K, Fe, and Na are components that promote the combustion reaction, and when the ash components such as Ca, K, Fe, and Na are large, the fuel tends to burn easily, etc. In addition to O / C and H / C, the flammability index Ib is obtained.

  Also, the difference in the combustibility of such fuels affects the unburned matter in ash and the concentration of NOx. Specifically, in the case of a flammable fuel, the unburned component in ash and the concentration of NOx tend to be small. On the contrary, in the case of a non-burnable fuel, the unburned component in ash tends to increase. Similarly, in the case of a non-burnable fuel, the amount of N released to the side of the flue 8p also increases, so the concentration of NOx tends to increase. Normally, in the boiler, air is divided and introduced by the burner 91 and the AA port 93, and NOx generated at the burner 91 side is reduced until AA, but if fuel is difficult to burn, NOx generation is delayed Since it flows toward the flue 8p, the reduction time to AA becomes short, and as a result, the NOx concentration on the flue 8p side increases. Thus, the pre-combustion fuel component Cp affects the unburned component in the ash and the NOx concentration.

  The command value determination unit 3 determines, based on the pre-combustion fuel component Cp, an evaluation target index T which is at least one of the above-described unburned in-ash content or NOx concentration in the flue 8p downstream of the combustion chamber 8f. The command value Pv of at least one of the above-described combustion control parameters P is determined so as to satisfy the criteria Tc. Although a specific method will be described later, the command value Pv may be determined from the pre-combustion fuel component Cp using a prediction map, machine learning, or the like.

  In the embodiment shown in FIG. 1, the command value determination unit 3 is connected to the pre-combustion fuel component acquisition unit 2 described above, and the pre-combustion fuel component Cp acquired by the pre-combustion fuel component acquisition unit 2 is input. It is supposed to be. Further, the command value determination unit 3 (the combustion condition determination device 1) is connected to the combustion control device 7c included in the combustion system 7, and inputs the determined command value Pv to the combustion control device 7c. There is. The combustion control device 7c is a device configured to perform combustion control by transmitting (commanding) the command value Pv of the combustion control parameter P determined by the combustion condition determination device 1 to the combustion device 9.

  Further, in the embodiment shown in FIG. 1, the pre-combustion fuel component acquisition unit 2 is connected to the measurement device 12 capable of measuring the operation state evaluation index R (described later) including the pre-combustion fuel component Cp in real time. The pre-combustion fuel component Cp measured at a predetermined cycle or the like by the measuring device 12 is acquired. More specifically, in the embodiment shown in FIG. 1, the measuring device 12 is a LIBS device, and at least the coal stored in the coal hopper 72, the pulverized coal at the outlet of the milling device 92 (pulverized fuel pipe Lf), etc. The measurement is performed on one pre-combustion fuel F at one place. Then, each time the measurement device 12 obtains a measurement value, the measured pre-combustion fuel component Cp is input to the pre-combustion fuel component acquisition unit 2.

  In addition, the LIBS apparatus measures in real time the measurement of the fuel component Cp before combustion and the measurement of the ash component of ash contained in the exhaust gas G as described later by a laser induced breakdown method (LIBS method). A device that can In the LIBS method, laser light is irradiated on a measurement target to be converted to plasma, plasma light generated from the plasma is incident on a spectroscope, and the component is identified from the difference in emission wavelength of spectral light separated by the spectroscope From the strength, the amount of each element (molecule) constituting the ash component is determined. For example, measurement windows are provided in various facilities such as the pulverized fuel pipe Lf, the flue 8p, and the exhaust gas pipe 75, and the LIBS apparatus may be configured to emit laser light and receive plasma light via the measurement window. good.

  However, the present invention is not limited to the present embodiment. In some other embodiments, the pre-combustion fuel component acquisition unit 2 can input data of the pre-combustion fuel component Cp measured in advance manually from a user such as an operator of the combustion furnace 8 or can be carried It may be acquired by communication with a portable storage medium such as a USB memory, another computer, or the like.

  The combustion condition determination device 1 having the above-described configuration determines the command value Pv of the combustion device 9 based on the pre-combustion fuel component Cp, whereby the evaluation target index T (unburned in ash, NOx concentration) is predetermined. An appropriate command value Pv expected to satisfy the criteria Tc can be determined quickly. That is, for example, in the combustion adjustment at the time of trial operation, after the combustion furnace 8 is subjected to trial operation using one or more (usually a plurality of) fuels having different combustibility, the evaluation object index T obtained as a result of the trial operation is measured. While confirming, for example, a combustion condition which can make the above-mentioned evaluation target index T satisfy a predetermined criterion Tc is searched. That is, since the value of the combustion control parameter P suitable for the fuel property of the used fuel is determined for each fuel while repeating trial and error in which the combustion conditions are changed, the cost of time, labor, funds, etc. I need it. Further, the fuel actually used at the time of operation after the start of operation of the combustion furnace 8 may not be the same as the property of the fuel used at the trial operation. In this case, normally, the operation is performed using the combustion condition set for the fuel that is considered to be close to the fuel property among the plurality of fuels used in the trial operation, but the combustion condition is It is not always optimal for fuel. Depending on the operation period, the combustion performance of the combustion furnace 8 may change due to ash deposition in the furnace or the like.

  However, according to the present invention, the command value Pv of the combustion control parameter P (combustion condition) based on the pre-combustion fuel component Cp for the fuel actually used at the time of trial operation of the combustion furnace 8 and operation after start of operation. If the combustion control parameter P is to be adjusted according to the pre-combustion fuel component Cp that affects the evaluation target T, or if the value of the combustion control parameter P is increased or decreased It is possible to more quickly and easily determine the command value Pv because it is possible to more quickly and easily determine the amount of change.

  According to the above configuration, the ash of ash produced when the pre-combustion fuel F is actually burned based on the component (pre-combustion fuel component Cp) of the fuel (pre-combustion fuel F) supplied to the burner 91 For example, the opening degree of the damper (air box damper 94) for adjusting the flow rate of combustion air in each burner 91, the burner nozzle angle, the solid fuel by the mill device 92, etc. The command value Pv of the combustion control parameter P of the combustion apparatus 9 that controls the combustion such as the classifier rotation speed at the time of pulverizing is determined. Thus, by focusing on the fuel component before the combustion before being supplied to the burner 91, the combustion control parameter P suitable for the fuel property can be determined quickly and accurately.

  Next, some embodiments of the method of determining the command value Pv of the combustion control parameter P executed by the above-described command value determination unit 3 will be described in detail with reference to FIGS. 2A to 2B. FIG. 2A is a diagram illustrating in detail the command value determination unit 3 according to an embodiment of the present invention, and determines the command value Pv of the combustion control parameter P based on the predicted value Te of the evaluation target index T. FIG. 2B is a diagram showing the command value determination unit 3 according to an embodiment of the present invention in detail, and the command value determination model Mv is used to determine the command value Pv of the combustion control parameter P.

  In some embodiments, as shown in FIG. 2A, the determination of the command value Pv of the combustion control parameter P is based on the value of the pre-combustion fuel component Cp and the value of the combustion control parameter P. It may be carried out by searching a combustion control parameter P which is a predicted value Te of the unburned part, NOx concentration) and which can meet the above-mentioned predetermined criteria Tc. Specifically, in some embodiments, as shown in FIG. 2A, the command value determination unit 3 includes a temporary setting value acquisition unit 31, an evaluation target index prediction unit 32, a criteria determination unit 33, and parameter adjustment. A unit 34 and a command value selection unit 35 are provided. Each of the above functions will be described.

  The temporary setting value acquisition unit 31 acquires a temporary setting value Pp of the combustion control parameter P. The temporary setting value Pp is a value that can be arbitrarily set as a candidate of the above-described command value Pv, and is given as an initial value or given from the parameter adjustment unit 34 described later. For example, it may be a set value of the combustion control parameter P defined for the fuel used during the trial operation of the combustion furnace 8 (the combustion mode of FIG. 5A described later). It may be a set value of the combustion control parameter P defined for the fuel whose property is considered to be most similar. Alternatively, the temporary set value Pp may be a standard value defined as a standard set value of the combustion control parameter P (standard combustion condition of FIG. 5B described later). This standard value may be created based on the set value determined at the time of combustion adjustment of the combustion furnace 8 for which the command value Pv is to be determined, or based on the set value of the combustion control parameter P of another combustion furnace 8 It may be created based on both of these setting values. Usually, since there are a plurality of combustion control parameters P, the temporary set value Pp may be acquired for each of them. The temporary setting value Pp is stored in the storage device m or the like of the combustion condition determination device 1 by being input by a user such as an operator of the combustion furnace 8 or obtained from another device through communication or the like. Thus, the temporary setting value acquisition unit 31 may acquire the temporary setting value Pp from the storage device m.

  The evaluation target index prediction unit 32 evaluates the evaluation target index based on the temporary setting value Pp of the combustion control parameter P and the pre-combustion fuel component Cp each time the temporary setting value acquisition unit 31 acquires the temporary setting value Pp. The predicted value Te of T is calculated. More specifically, in some embodiments, the evaluation object index prediction unit 32 includes a combustibility index Ib (described above) indicating the combustibility of the fuel F before combustion, an evaluation object index T, and a combustion control parameter P (combustion condition) The predicted value Te of the evaluation target index T may be calculated from the flammability index Ib and the temporary setting value Pp using a prediction map Mf indicating the relationship with the above. The prediction map Mf may be a functionalization of the relationship between the flammability index Ib, the evaluation target index T, and the combustion control parameter P obtained through experiments and the like. By using such a prediction map Mf, it is possible to predict the command value Pv of the combustion control parameter P quickly and accurately.

  Alternatively, in some other embodiments, the pre-combustion fuel component Cp, the measurement value of the evaluation target index T generated by the combustion of the pre-combustion fuel F having the pre-combustion fuel component Cp, and the pre-combustion fuel F Pre-combustion fuel component using prediction model Mm created by machine learning teacher data composed of a plurality of data associated with command value Pv (combustion condition) of combustion control parameter P at the time of combustion The predicted value Te of the evaluation target index T may be calculated from the temporary setting value Pp of Cp and the combustion control parameter P. At this time, the prediction model Mm may be created by a known machine learning method (algorithm) such as, for example, a neural network. By using such a prediction model Mm, it becomes possible to easily predict the appropriate command value Pv of the combustion control parameter P and the appropriate command value Pv of the combustion control parameter P.

  The criteria determination unit 33 determines whether the predicted value Te of the evaluation target index T satisfies a predetermined criteria Tc. Specifically, it is determined whether or not the predicted value Te of the unburned portion in ash is less than or equal to a specified value (such as 5% or less). Alternatively, it is determined whether the predicted value Te of the NOx concentration is less than or equal to the guaranteed value. In the embodiment shown in FIG. 2A, it is determined whether the predicted value Te of the unburned part in ash is less than the specified value and the predicted value Te of the NOx concentration is less than the guaranteed value. Then, if both conditions are satisfied, it is determined that the criteria Tc is satisfied, and if at least one of the conditions is not satisfied, it is determined that the criteria Tc is not satisfied.

  The parameter adjustment unit 34 determines that the provisional control value acquisition unit 31 acquires the provisional value of the combustion control parameter P acquired by the temporary setting value acquisition unit 31 until the criterion determination unit 33 determines that the predicted value Te of the evaluation target index T satisfies the predetermined criterion Tc. The setting value Pp is changed, and the temporary setting value acquiring unit 31 acquires the changed temporary setting value Pp. In the embodiment shown in FIG. 2A, when the criteria determination unit 33 determines that the predicted value Te of the evaluation target index T does not satisfy the predetermined criteria Tc, the parameter adjustment unit 34 receives an input from the criteria determination unit 33. The temporary set value Pp thus obtained is changed (corrected), and the temporary set value Pp ′ after the change is input to the temporary set value acquiring unit 31. As a result, the temporary setting value acquiring unit 31 acquires a new temporary setting value Pp ′, and at the same time, the new temporary setting value Pp ′ is input to the criteria determining unit 33. Such a series of flows are repeated until the criteria determination unit 33 determines that the predicted value Te of the evaluation target index T satisfies the predetermined criteria Tc.

  Then, the command value selection unit 35 sets the temporary setting value Pp (Pp ') of the combustion control parameter P when the criteria determination unit 33 determines that the predicted value Te of the evaluation target index T satisfies the predetermined criteria Tc. As a command value Pv. In the embodiment shown in FIG. 2A, the command value selection unit 35 is connected to the criteria determination unit 33, and the criteria determination unit 33 inputs a temporary setting value Pp determined to satisfy the predetermined criteria Tc.

  According to the above-described configuration, the unburned portion in the ash of the ash which is generated when the pre-combustion fuel F is burned, based on the temporarily set value Pp of the combustion control parameter P and the pre-combustion fuel component Cp While predicting the NOx concentration, while changing the temporary setting value Pp of the combustion control parameter P, the temporary setting value Pp whose prediction result satisfies the criteria Tc is found, and is used as the command value Pv of the combustion control parameter P. Thus, the command value Pv of the combustion control parameter P is determined such that the unburned portion in ash of the ash produced when the fuel F before combustion is actually burned and the NOx concentration in the exhaust gas satisfy the criteria Tc. it can.

  Further, in some other embodiments, the evaluation object index T is a criterion Tc using the command value determination model Mv obtained by machine learning without calculating the prediction value Te of the evaluation object index T as described above. The command value Pv may be obtained directly so as to satisfy Specifically, in some embodiments, the command value determination unit 3 measures the measurement target index T generated by the combustion of the pre-combustion fuel component Cp and the pre-combustion fuel F having the pre-combustion fuel component Cp. Machine learning teacher data composed of a plurality of data in which a measured value satisfying a predetermined criterion Tc is associated with a command value Pv of a combustion control parameter P when the pre-combustion fuel F is burned The command value Pv of the combustion control parameter P is determined from the pre-combustion fuel component Cp using the command value determination model Mv created by the above. As described above, the appropriate command value Pv of the combustion control parameter P can be easily obtained by using the command value determination model Mv created on the basis of machine learning with the above data in the case of satisfying the predetermined criteria Tc as teacher data. It can be decided.

  In the two types of machine learning described above, the teacher data includes the amounts of C, H, O, N, Ca, K, Na, Fe, etc. in the fuel F before combustion, the unburned content in ash, the NOx concentration, etc. Composed of multiple data linked with measurement results. Under the present circumstances, when the unburned matter in ash can not be sorted out by C, H, O, N (when these correlations are low) or when the concentration of C, H, O, N and NO can not be sorted out, Ca, K , Na, Fe, or Na may be more. That is, even if the amount of C is large relative to the amount of H or O, if there is a large amount of Ca, K, Na, Fe, Na, the reaction may be promoted by these catalytic actions. Therefore, as described above, if teacher data is created and machine learning is performed, optimum combustion conditions according to the pre-combustion fuel component Cp should be determined promptly and appropriately even if fuel such as coal type changes. It is possible to easily determine the combustion conditions. If teacher data, a prediction model Mm, and a command value determination model Mv described later are stored on a cloud or the like, deployment to other combustion furnaces (plants) is facilitated. Further, more accurate prediction is possible if teacher data are also associated with the result of industrial analysis (fixed carbon amount, volatile matter amount, ash amount, water amount) by hand analysis or the like. That is, as the amount of fixed carbon (amount of C) in the fuel F before combustion is larger, it is more difficult to burn, and as the amount of ash component is lower, relatively more unburned in ash is calculated even with the same combustion rate (two-stage combustion rate) It will be In addition, when the water content is evaporated, the coal structure may be more easily burned by becoming porous or activated as the water content is increased.

  Further, in the embodiment in which the machine learning described above is performed, in some embodiments, the combustion condition determination device 1 determines whether to perform relearning of the prediction model Mm or the command value determination model Mv. The unit 6 may be further provided. Specifically, for the prediction model Mm, the relearning determination unit 6 relearns the prediction model Mm based on the difference between the prediction value Te of the evaluation target index T and the measurement value Rm of the evaluation target index T. It is determined whether or not it is (see FIG. 6 described later). For example, when the predicted value Te of the evaluation object index T exceeds ± x% (x is an integer greater than or equal to 0) or the measured value Rm of the evaluation object index T, the relearning determination unit 6 If the absolute value of the difference (subtraction) between the predicted value Te of and the measured value Rm of the evaluation target index T is larger than a predetermined value, it may be determined that relearning is necessary. As described above, when the predicted value of the evaluation target index T exceeds the allowable range and differs from the measured value, etc., it is determined that relearning is necessary, and relearning is performed according to this determination, based on machine learning. The prediction accuracy using the prediction model to be created can be maintained.

  On the other hand, for the command value determination model Mv, the relearning determination unit 6 determines that relearning of the command value determination model Mv is necessary when the measurement value Rm of the evaluation target index T does not satisfy the predetermined criteria Tc. Is also good (see FIG. 7 described later). Thus, the determination of the command value Pv of the combustion control parameter P using the command value determination model Mv can be appropriately maintained.

Next, an embodiment will be described in which the command value Pv of the combustion control parameter P is adjusted in accordance with the actual operating state during operation of the combustion furnace 8.
As described above, by operating the combustion apparatus 9 with the command value Pv of the combustion control parameter P determined by the command value determination unit 3, it is expected that the evaluation target index T satisfies the criteria Tc. In the actual operation of the case, the evaluation object index T may not satisfy the criteria Tc, or a more appropriate combustion control parameter P may exist. Therefore, the command value Pv of the combustion control parameter P, which has already been determined, is adjusted during the operation of the combustion furnace 8 in accordance with the actual operating state of the combustion furnace 8.

  For this reason, in some embodiments, as shown in FIG. 1, the combustion condition determination device 1 acquires the driving state evaluation index acquisition unit 4 that acquires the measured value Rm of the driving state evaluation index R, and acquires the driving state evaluation index. Adjustment value Pa for adjusting the command value Pv of the combustion control parameter P determined by the command value determination unit 3 based on the measured value Rm of the driving state evaluation index R and the pre-combustion fuel component Cp acquired by the unit 4 And an adjustment value determination unit 5 that determines the In the embodiment shown in FIG. 1, the adjustment value determination unit 5 is connected to the driving condition evaluation index acquisition unit 4 and the command value determination unit 3 respectively, and the measurement value Rm from the driving condition evaluation index acquisition unit 4 is The command value Pv is updated by determining a new command value Pv as an adjustment value Pa and transmitting the command value to the command value determination unit 3 as well as being input. That is, the command value determination unit 3 transmits the new command value Pv received from the adjustment value determination unit 5 to the combustion control device 7c, whereby the command value Pv is updated.

  Here, as the operating condition evaluation index R, there are various indexes such as the adhesion amount B of ash, the fuel particle size S, etc., in addition to the evaluation object index T (unburned content in ash, NOx concentration). The driving state evaluation index acquisition unit 4 acquires at least one of these indices. A representative driving condition evaluation index R will be described using FIGS. 3 to 4. FIG. 3 is a diagram showing in detail the adjustment value determination unit 5 according to the embodiment of the present invention. Moreover, FIG. 4 is a figure which shows the processing content of the evaluation object parameter | index adjustment value calculation part 51 which concerns on one Embodiment of this invention.

[1] When the operating state evaluation index R is the evaluation target index T (unburned in ash, NOx concentration),
In some embodiments, the driving condition evaluation index R includes an evaluation target index T. Further, the driving state evaluation index acquisition unit 4 acquires the measurement value Rm of the evaluation target index T. Then, the adjustment value determination unit 5 calculates an adjustment value for adjusting the command value Pv of the combustion control parameter P based on the measured value Rm of the evaluation object index T and the pre-combustion fuel component Cp. It has 51. In this case, as shown in FIG. 1, the measured value Rm of the operating condition evaluation index R may be measured on the upstream side of the denitration device 76 in the flue 8p. In the embodiment shown in FIG. 1, the unburned portion in ash and the NOx concentration are measured between the economizer 82e (carbon economizer) and the NOx removal apparatus 76 (economizer outlet). In addition, the unburned content in ash is measured by the measuring device 12 which is a LIBS device, and the NOx concentration is measured by a NOx concentration meter (not shown) capable of measuring the NOx concentration. Rm is input to the driving state evaluation index acquisition unit 4 in real time. However, the present invention is not limited to the present embodiment. In some other embodiments, the measuring device 12 may be an imaging means such as a camera capable of acquiring an image (still image, moving image) in the flue 8p. In this case, the driving state evaluation index acquisition unit 4 can acquire the measurement value Rm of the unburned portion in ash obtained by the image diagnosis in real time. In some other embodiments, the measuring device 12 may be a device capable of measuring unburned ash in ash in real time using a microwave or the like.

  When the operating state evaluation index R is the evaluation target index T, as shown in FIG. 4, the pre-combustion fuel component Cp may be the above-described combustibility index Ib. For example, in some embodiments, the pre-combustion fuel component Cp includes the concentration of each of hydrogen, carbon and oxygen in the pre-combustion fuel F, and the measurement value Rm of the operating condition evaluation index R described above is not yet in ash. It is the measured value Rm of the fuel component. For example, if the amount of C in the fuel F before combustion is smaller than the amount of H or O and the H / C and O / C are large (when the fuel F before combustion is easily burned), temporary measurement of the unburned component in ash If the value Rm is larger than the assumed value which is a value within the range satisfying the predetermined criteria Tc such as the predicted value Te (see step S43 in FIG. 4), the combustion state is not so good, and it is used at that time It is determined that the command value Pv of the combustion control parameter P being set is inappropriate. Therefore, adjustment (control) on the side to lower the concentration of the unburned component in ash is performed (see step S44 in FIG. 4). Specifically, adjustment may be made to lower the charging angle so as to direct AA to the lower part of the combustion furnace 8 so as to maintain the residence time of AA in the furnace. By increasing the opening degree of the air box damper 94, the amount of air in the vicinity of the burner 91 may be increased to promote combustion. Alternatively, control may be performed to make the particles of the pre-combustion fuel F smaller. That is, at least one adjustment such as decreasing the two-stage combustion rate as the combustion control parameter P, decreasing the AA angle, increasing the classifier rotation speed of the pre-combustion fuel F or the like by the mill device 92 is performed.

  In some other embodiments, the pre-combustion fuel component Cp includes the concentrations of hydrogen, carbon, oxygen, and nitrogen in the pre-combustion fuel F, and the measured value Rm of the operating condition evaluation index R described above is the NOx concentration The measured value Rm of For example, in the case where the amount of C in the fuel F before combustion is smaller than the amount of H or O and the H / C and O / C are large (when the fuel F before combustion is easy to burn) and the N amount is small. If the measured value Rm of the NOx concentration is larger than the above assumed value (see step S45 in FIG. 4), the ignition condition (combustion condition) of the burner 91 is bad, so the combustion control parameters used at that time The command value Pv of P is determined to be inappropriate. Therefore, the adjustment (control) on the side of lowering the concentration of NOx is performed (see step S46 in FIG. 4). Specifically, at least one adjustment is performed such as decreasing the fuel injection angle of the burner 91, which is the combustion control parameter P, increasing the two-burning rate, and increasing the AA angle.

  Also, as described above, the control for reducing the NOx concentration and the control for reducing the unburned component in the ash may be contradictory events, so the NOx concentration and the measured value Rm of the unburned component in the ash may be The adjustment value Pa may be calculated by looking at the balance. For example, if the unburned content in ash can be treated as a valuable material below the specified value, if the unburned content in ash exceeds the specified value, the unburned content in ash can be increased even if the NOx concentration is increased within the guaranteed value range. You may give priority to control that lowers

  In some other embodiments, the pre-combustion fuel F comprises an ash component that produces ash upon combustion of the pre-combustion fuel F, and the pre-combustion fuel component Cp comprises the amount of ash component in the pre-combustion fuel F, The measured value Rm of the above-mentioned operating condition evaluation index R is a measured value Rm of the unburned component in ash. In the case where the amount of Ca, K, Fe, Na in the pre-combustion fuel F is large, the case where the measured value Rm of the unburned portion in ash is larger than the above assumed value (see step S47 in FIG. 4) Such adjustment (control) is performed on the side to lower the concentration of unburned matter in ash (see step S48 in FIG. 4).

  In the embodiment shown in FIG. 4, in step S41, the fuel component of the fuel (pre-combustion fuel F) at the outlet of the milling apparatus 92 is obtained, and in step S42 the operating condition evaluation index (unburned in ash, NOx concentration) Obtain the measured value Rm at the exit of the economizer of Then, in steps S43 to S48, calculation of the adjustment value as described above is performed. The order of steps S43 to S44, S45 to S46, and S47 to S48 may be arbitrarily changed.

[2] When the operating condition evaluation index R is the unburned component in NOx or the concentration of NOx, and the ignition distance D is adjusted,
In some embodiments, the operating condition evaluation index R includes at least one of the above-described unburned component in the ash or the concentration of NOx. Further, the operating condition evaluation index acquisition unit 4 acquires at least one of the measurement value Rm of the unburned component in the ash or the NOx concentration. Then, the adjustment value determination unit 5 determines the burner in the furnace when the measured value Rm of at least one of the unburned ash content or the NOx concentration is larger than the planned value of the unburned content in ash or the planned value of the NOx concentration. The ignition distance adjustment value calculation unit 52 calculates the adjustment value Pa for adjusting the combustion control parameter P so that the distance (ignition distance D) between the tip of the fuel injection valve 91 and the fuel ignition position 91 becomes short. The planned value of the unburned part in the above-mentioned ash is an arbitrary value (value range) less than the specified value of the unburned part in the above-mentioned ash, and the planned value of the NOx concentration is smaller than the guaranteed value of the NOx concentration It may be any value (value range), for example, the above-described predicted value Te. As described above, the unburned component in the ash and the concentration of NOx may be measured on the upstream side of the NOx removal apparatus 76 such as the economizer outlet. Further, the ignition position is a position where the fuel injected from the burner 91 into the furnace is ignited, and is at a position separated from the tip of the burner 91 by the ignition distance D. In the embodiment shown in FIG. 1, the operating condition evaluation index acquiring unit 4 is connected to the measuring device 12 capable of real time measurement as described above, and measures the unburned component in ash or NOx concentration by the measuring device 12. Each time, the measured value Rm is input to the operation state evaluation index acquisition unit 4 to acquire the measured value Rm of the unburned component in the ash or the NOx concentration in real time.

  And when measured value Rm of unburned part in ash is higher than plan value, optimization of the ignition position for reducing unburned part in ash is performed. The flow rate of the secondary air A2 supplied from the burner 91 into the furnace may be reduced so that the ignition distance D becomes short. The classifier rotation speed of the milling apparatus 92 may be increased. Both of these may be combined. By this, the ignition distance D can be controlled to be short (the ignition position approaches the burner side), and the measurement value Rm of the unburned portion in the ash can be further reduced. Similarly, in the case of performing two-stage combustion, if the measured value Rm of NOx concentration is higher than the planned value, the shorter the ignition distance D, the faster the combustion occurs, and NOx is released earlier, so reduction to AA As the time is increased, the NOx concentration can be reduced. At this time, the ignition distance D obtained by measuring the ignition distance D may be further considered. Specifically, in the case where the measured value Rm of the unburned component in the ash and the NOx concentration is higher than the planned value, the ignition property is deteriorated such that the ignition distance D becomes longer than the appropriate value (the ignition position is separated). In this case, control may be performed such that the above-mentioned ignition distance D can be shortened. The ignition position can be measured by a measuring device 12 such as a two-color thermometer, an imaging diagnosis, a radiation thermometer, etc. The ignition distance D is measured through individual measurement.

  According to the above configuration, when the measured value Rm of the ash unburned component generated after combustion of the fuel is larger than the planned value, the burner 91 and the burner 91 are used to make the measured value Rm equal to or less than the planned value. The adjustment value Pa of the combustion control parameter P is calculated such that the ignition distance D to the ignition position becomes smaller. Thereby, the command value Pv of the combustion control parameter P can be adjusted to a more appropriate value according to the actual situation.

  In some other embodiments, the operating condition evaluation index R includes the above-described unburned ash content in the ash and the deposited amount B of ash deposited at the height position of the burner 91 on the wall in the furnace. In addition, the driving state evaluation index acquisition unit 4 acquires the measurement value Rm of the unburned portion in ash and the measurement value Rm of the adhesion amount B of ash. And adjustment value determination part 5 is in the furnace, when measured value Rm of unburned part in ash is higher than the unburned part planned value in ash, and adhesion amount B of ash is smaller than the adhesion amount planned value. The second ignition distance adjustment value calculation unit 53 calculates the adjustment value Pa for adjusting the combustion control parameter P so that the above-mentioned ignition distance D between the tip of the burner 91 and the ignition position of the fuel becomes short. The in-ash unburned part planned value is an arbitrary value (value range) equal to or less than the specified value of the unburned in-ash portion such as the above-described predicted value Te in the unburned in-ash part. As described above, the unburned component in ash may be measured on the upstream side of the NOx removal apparatus 76 such as the economizer outlet.

  Moreover, said adhesion amount plan value is a standard amount of the ash adhering to a furnace wall. Normally, the combustion conditions are determined to be optimum with a standard amount of ash adhering to the furnace wall, but the ash adhesion amount B reaches a standard amount immediately after the start of operation of the combustion furnace 8 etc. However, the release of heat through the furnace wall to the outside of the road is larger than when ash adheres to the standard amount, and the unburned content in the ash tends to be large. Therefore, the combustion conditions of the combustion furnace 8 are optimized according to the amount B of attached ash measured using the measuring device 12 such as an infrared camera. For example, the amount of attached ash may be estimated from the ratio of heat absorption of the furnace wall and the upper heat transfer surface (82). If a large amount of ash adheres to the furnace wall, the heat absorption amount at the furnace wall decreases and the heat absorption amount at the upper heat transfer surface (82) increases, so that it is possible to evaluate at this ratio.

Specifically, when the amount of deposited ash B is small and the unburned content in ash is high, the amount of fuel burned in the burner area (burning area) of the burner 91 is increased by reducing the two-stage burning rate. By setting the temperature, the burner area may be heated to increase the ash deposition amount B in the burner area. Alternatively, by increasing the two-stage combustion rate, the amount of oxygen (amount of O ₂ ) in the burner area may be reduced to lower the melting temperature of ash and increase the amount B of deposited ash.

  The driving state evaluation index acquisition unit 4 may further acquire the ignition part temperature as the driving state evaluation index R. The ignition part temperature may be a temperature inside the air box 83 or in the vicinity of the burner 91 in the furnace. In this case, the second ignition distance adjustment value calculation unit 53 calculates the adjustment value Pa for adjusting the combustion control parameter P by adding the condition that the in-furnace temperature is lower than the in-furnace temperature planned value. Also good.

  According to the above configuration, the adjustment value Pa of the combustion control parameter is calculated in consideration of the above-described amount of deposited ash B. By this, the command value of the combustion control parameter can be adjusted to a more appropriate value according to the actual situation.

[3] When the operating condition evaluation index R is the fuel particle size S,
In some embodiments, the operating condition evaluation index R is the fuel particle size S of the pulverized fuel which is the pre-combustion fuel F generated by the mill device 92 or the pre-combustion fuel F supplied into the furnace from the burner 91. Fuel particle size S is included. In addition, the driving state evaluation index acquisition unit 4 acquires the measured value Rm of the fuel particle size S. The adjustment value determination unit 5 has a fuel particle size adjustment value calculation unit 54 that calculates an adjustment value Pa for adjusting the command value Pv of the combustion control parameter P based on the measurement value Rm of the fuel particle size S. The fuel particle size S can be measured, for example, by a laser measurement device that measures using a laser. The driving state evaluation index acquiring unit 4 is connected to the measuring device 12 capable of real-time measurement such as a laser measuring device, and the driving state evaluation indicator acquiring portion 4 in accordance with the degree of real time measurement by the measuring device 12 or the like. The measured value Rm of the fuel particle size S may be acquired in real time by inputting the measured value Rm into 4.

  More specifically, when the fuel particle size S is that of the pulverized fuel which is the pre-combustion fuel F generated by the mill device 92, for example, the burner 91 at each corner in the furnace of the combustion furnace 8 is used. The fuel particle size S of the pulverized fuel to be supplied may be measured in units of burners 91 or in units of corner portions, and the command value Pv of the combustion control parameter P may be adjusted according to the measured value Rm. Specifically, when the fuel particle size S supplied to a specific burner 91 or a corner is large, the opening degree of the wind box damper 94 is controlled to set the specific burner 91 or the corner. The amount of secondary air (the amount of air for combustion) is reduced, and the amount of secondary air to other burners 91 or other corners is increased. Alternatively, the amount of secondary air to that particular burner 91 or corner may be increased and the amount of secondary air to another burner 91 or corner may be decreased.

  On the other hand, when the fuel particle size S is that of the fuel supplied from the burner 91 into the furnace (pre-combustion fuel F supplied to the furnace), a fuel containing coarse particles such as biomass is used in particular When the combustion furnace 8 is operated, the fuel particle size S of the unburned fuel falling to the bottom 84 b (furnace bottom) of the bottom hopper portion 84 is measured, and the command value of the combustion control parameter P according to the measured value Rm. You may adjust Pv. Specifically, when the fuel particle size S is larger than the assumed value according to the classifier rotation speed of the mill device 92, the fuel rotation speed of the pre-combustion fuel F is increased by increasing the classifier rotation speed of the mill device 92. By promoting the pulverization, the size of the coarsely falling particles may be reduced. By adjusting the injection angle of the burner 91 more vertically upward, the residence time of the falling particles in the furnace may be increased to reduce the coarse falling particles. Further, by increasing the secondary air amount of the burner 91 on the lower side of the plurality of burners 91 installed along the vertical direction of the furnace wall by adjusting the opening degree of the air box damper 94, the empty tower in the lower part of the furnace By increasing the speed, the coarse falling particles may be reduced. At least two of these adjustments may be combined.

  In the above embodiment, although measured at the furnace bottom, the fuel particle size S of floating unburned particles between the AA port 93 and the furnace outlet is measured, and a command for the combustion control parameter P is issued according to this measured value Rm. The value Pv may be adjusted. Specifically, when the fuel particle size S of the floating unburned particles is larger than the expected value, the pulverization of the pre-combustion fuel F is promoted by increasing the classifier rotation speed of the milling apparatus 92, The size of the coarsely falling particles may be reduced. By adjusting the injection angle of the burner 91 downward in the vertical direction, the residence time of the combustion completion zone may be increased to reduce floating unburned particles. By reducing the secondary air volume of the burner 91 on the lower side of the burners 91 installed in multiple stages along the vertical direction of the furnace wall by adjusting the opening degree of the air box damper 94, the sky velocity of the lower part in the furnace is By reducing the amount, coarse floating unburned particles may be reduced. At least two of these adjustments may be combined.

  According to the above configuration, the measured value Rm of the fuel particle size S of the pre-combustion fuel F before being supplied into the furnace, or the fuel particle size of the unburned fuel of the pre-combustion fuel F supplied into the furnace Based on the measurement value Rm of S, the adjustment value Pa of the combustion control parameter P is calculated. Thereby, the command value Pv of the combustion control parameter P can be adjusted to a more appropriate value according to the actual situation.

  Hereinafter, the combustion condition determination method of the combustion furnace 8 corresponding to the process which the combustion condition determination apparatus 1 mentioned above performs is demonstrated using FIG. 5A-FIG. FIG. 5A is a flowchart showing a method of determining the combustion condition of the combustion furnace 8 according to an embodiment of the present invention, in which the command value Pv of the combustion control parameter P is determined based on the predicted value Te of the evaluation target index T. FIG. 5B is a flowchart showing the method of determining the combustion condition of the combustion furnace according to the embodiment of the present invention, in which the command value Pv of the combustion control parameter P is determined using the prediction model Mm. FIG. 6 is a flow chart showing the relearning determination step according to an embodiment of the present invention. FIG. 7 is a flow chart showing the relearning determination step of the command value determination model Mv according to an embodiment of the present invention.

The combustion condition determination method of the combustion furnace 8 (hereinafter simply referred to as the combustion condition determination method) is a method of determining a combustion condition including a plurality of combustion control parameters P for controlling the combustion device 9 of the combustion furnace 8. As shown in FIGS. 5A to 5B, the combustion condition determination method includes a pre-combustion fuel component acquisition step (S2) and a command value determination step (S3). The combustion condition determination method may be performed by the above-described combustion condition determination device, or may be performed manually while acquiring measurement results of the measurement device 12 or the like. The execution timing of this method may be at the time of fuel change such as coal change at the time of operation other than at the time of test operation of the combustion furnace 8, at regular intervals, or in real time.
The combustion condition determination method shown in FIGS. 5A to 5B will be described in the order of steps shown in the drawing, taking coal change at the time of operation of the pulverized coal cooking boiler as an example.

  In step S1 of FIGS. 5A-5B, preparations are made prior to execution of the method. In the embodiment shown in FIG. 5A, at the time of combustion adjustment (at the time of test operation), combustion conditions for each of a plurality of types of fuel with different fuel properties are determined, and a plurality of combustion modes are prepared (step S1a). On the other hand, in the embodiment shown in FIG. 5B, the standard combustion conditions are determined and prepared based on the combustion conditions for each fuel type obtained through the combustion adjustment for each of the target combustion furnace 8 and the other combustion furnaces 8. Step S1 b. Thereafter, at the time of operation of the combustion furnace 8, an execution timing (coal change in this embodiment) arrives, and the subsequent step S2 and subsequent steps are executed.

  In step S2, a pre-combustion fuel component acquisition step is performed. The pre-combustion fuel component acquisition step (S2) includes the measurement including at least a part of the measurement value of the fuel component C of the pre-combustion fuel F which is the fuel (coal) supplied to the above-described burner 91 of the combustion furnace 8 (boiler). This is a step of acquiring the pre-fuel component Cp. Since the pre-combustion fuel component acquisition step (S2) is the same as the processing content executed by the pre-combustion fuel component acquisition unit 2 described above, the details will be omitted. In the embodiment shown in FIGS. 5A to 5B, the post-combustion fuel component Cp of coal after change is automatically read through the measurement in real time by the above-described LIBS device.

  In step S3, a command value determination step is performed. In the command value determination step (S3), at least one evaluation target T mentioned above in the flue 8p downstream of the combustion chamber 8f satisfies a predetermined criterion Tc based on the pre-combustion fuel component Cp. This is a step of determining the command value Pv of the combustion control parameter P described above. The command value determination step (S3) is the same as the processing content executed by the command value determination unit 3 described above, and thus details thereof will be omitted. However, in some embodiments, the above-described command value determination step (S3) The command value Pv may be determined by steps as shown in FIGS. 5A to 5B. Alternatively, in some other embodiments, the command value determination step (S3) described above is the evaluation target index T generated by the combustion of the pre-combustion fuel component Cp and the pre-combustion fuel component Cp. Teaching data consisting of a plurality of data in which the measured values satisfying the predetermined criteria Tc are associated with the command value Pv of the combustion control parameter P when the pre-combustion fuel component Cp is burned. The command value Pv may be determined from the pre-combustion fuel component Cp using a command value determination model Mv created by machine learning.

  In the embodiment shown in FIGS. 5A to 5B, the command value determining step (S3) includes a temporary setting value acquiring step (S31) for acquiring the temporary setting value Pp of the combustion control parameter P, and the above-mentioned temporary setting value acquiring step ( Evaluation target index prediction step (step S31) of calculating the predicted value Te of the evaluation target index T based on the temporary setting value Pp of the combustion control parameter P and the pre-combustion fuel component Cp each time the temporary setting value Pp is acquired. S32), the criteria determination step (S33) of determining whether the predicted value Te of the evaluation target index T satisfies the predetermined criteria Tc, and the predicted value Te of the evaluation target index T by the criteria determination step (S33) The provisional setting value Pp of the combustion control parameter P acquired by the provisional setting value acquiring unit 31 until it is determined that the predetermined criteria Tc are satisfied. The parameter adjustment step (S34) which causes the temporary setting value acquisition step (S31) to acquire the changed temporary setting value Pp while changing (loop from S31 with the changed temporary setting value Pp), and the above-mentioned criteria determination unit 33 A command value selection step (S35) in which the temporary setting value Pp (Pp ') of the combustion control parameter P is determined as the command value Pv when it is determined that the predicted value Te of the evaluation target index T satisfies the predetermined criteria Tc Have.

  Under the present circumstances, in embodiment shown to FIG. 5A, evaluation object index prediction step (S32) is combustibility index Ib (described above) which shows the combustibility of the fuel F before combustion, evaluation object index T, and combustion control parameter P (combustion condition The predicted value Te of the evaluation target index T is calculated from the flammability index Ib and the temporary setting value Pp using the prediction map Mf indicating the relationship with (1) (S32a). On the other hand, in the embodiment shown in FIG. 5B, the measured values of the evaluation target index T generated by the combustion of the pre-combustion fuel component Cp and the pre-combustion fuel F having the pre-combustion fuel component Cp and the pre-combustion fuel component Cp Pre-combustion fuel component using prediction model Mm created by machine learning teacher data composed of a plurality of data associated with command value Pv (combustion condition) of combustion control parameter P at the time of combustion The predicted value Te of the evaluation target index T is calculated from Cp and the temporary setting value Pp (S32b).

  The temporary set value acquiring step (S31), the evaluation target index predicting step (S32), the criteria determination step (S33), the parameter adjusting step (S34), and the command value selecting step (S35) mentioned above Since the process contents performed by the set value acquisition unit 31, the evaluation target index prediction unit 32, the criteria determination unit 33, the parameter adjustment unit 34, and the command value selection unit 35 are the same as each other, the details will be omitted.

  According to the above configuration, the command value Pv of the combustion control parameter P is determined such that the unburned portion in ash of the ash produced when the fuel F before combustion is actually burned and the NOx concentration in the exhaust gas satisfy the criteria Tc. can do.

  In some embodiments, as shown in FIGS. 5A to 5B (steps S4 to S6), the command value Pv of the combustion control parameter P is adjusted in accordance with the operating state of the combustion furnace 8 during operation. Also good. That is, the combustion condition determination method includes a driving state evaluation index acquisition step (S4) of acquiring the measured value Rm of the driving state evaluation index R, and a measured value of the driving state evaluation index R acquired by the driving state evaluation index acquisition unit 4. Adjustment value determination step (S5 to S6) for determining an adjustment value Pa for adjusting the command value Pv of the combustion control parameter P determined at the command value determination step (S3) based on Rm and the pre-combustion fuel component Cp And further comprising Since the driving state evaluation index acquisition step (S4) and the adjustment value determination step (S5 to S6) are respectively the same as the processing contents executed by the driving state evaluation index acquisition unit 4 and the adjustment value determination unit 5 described above, Details are omitted.

  In the embodiment shown in FIGS. 5A to 5B, in step S4 executed after step S3, the measurement value Rm of the evaluation target index T measured by LIBS or the like is acquired. Then, in step S5, it is determined whether the measurement value Rm of the evaluation target index T satisfies a predetermined criterion Tc. When it is determined in step S5 that the measured value Rm of the evaluation object index T does not satisfy the predetermined criteria Tc, based on the measured value Rm of the evaluation object index T and the pre-combustion fuel component Cp in step S6. At least one combustion control parameter P (combustion condition) is changed by calculating an adjustment value for adjusting the command value Pv of the combustion control parameter P (executing the evaluation object index adjustment value calculating step), and step S4 again. Return to Conversely, if it is determined in step S5 that the measurement value Rm of the evaluation target index T satisfies the predetermined criteria Tc, control based on the combustion condition is continued. Since the evaluation target index adjustment value calculation step described above is the same as the processing content executed by the evaluation target index adjustment value calculation unit 51 described above, the details will be omitted.

  In some other embodiments, the operating condition evaluation index R may include at least one of the above-described ash unburned component or NOx concentration. In this case, the operating state evaluation index acquiring step (S4) acquires at least one of the measurement values Rm of the unburned component in the ash or the NOx concentration. In the adjustment value determination step (S5), the combustion control parameter P is set such that the above-described ignition distance D becomes short when at least one of the unburned ash in the ash and the measured value Rm of NOx concentration is larger than the planned value. To calculate the adjustment value Pa for adjusting the (ignition distance adjustment value calculation step). Since this ignition distance adjustment value calculation step is the same as the processing content executed by the ignition distance adjustment value calculation unit 52 described above, the details will be omitted.

  In some other embodiments, the operating condition evaluation index R may include the above-described unburned ash content in the ash and the deposited amount B of ash deposited at the height position of the burner 91 on the wall in the furnace. . In this case, the operating state evaluation index acquiring step (S4) acquires the measured value of the unburned part in ash and the measured value of the attached amount B of ash. Then, in the adjustment value determination step (S5 to S6), when the measured value of the unburned component in ash is higher than the unburned component planned value in ash and the attached amount of ash is smaller than the attached value planned value, The adjustment value Pa for adjusting the combustion control parameter P is calculated so as to shorten the ignition distance D (second ignition distance adjustment value calculation step). Since this second ignition distance adjustment value calculation step is the same as the processing content executed by the second ignition distance adjustment value calculation unit 53 described above, the details will be omitted.

  In some other embodiments, the operating condition evaluation index R is the fuel particle size S of the pulverized fuel which is the pre-combustion fuel F generated by the mill device 92, or the pre-combustion fuel supplied into the furnace from the burner 91. The fuel particle size S of F may be included. In this case, the operation state evaluation index acquisition step (S4) acquires the measured value Rm of the fuel particle size S. Further, the adjustment value determination step (S4 to S5) calculates an adjustment value Pa for adjusting the command value Pv of the combustion control parameter P based on the measurement value Rm of the fuel particle size S (fuel particle size adjustment value calculation step ). Since this fuel particle size adjustment value calculation step is the same as the processing content executed by the fuel particle size adjustment value calculation unit 54 already described, the details will be omitted.

  In some embodiments, as shown in FIG. 6, the combustion condition determination method re-learns the prediction model Mm based on the difference between the prediction value Te of the evaluation target index T and the measurement value Rm thereof. It may further include a relearning determination step (predictive model relearning determination step) for determining whether or not to be. In the embodiment shown in FIG. 6, in step S61, the predicted value Te of the evaluation object index T (the unburned in ash, NOx concentration) predicted using the above-described prediction model Mm is acquired. In step S62, the measurement value Rm of the evaluation target index T measured by the LIBS device or the like is acquired. Then, in steps S63 to S64, when the difference between the predicted value Te for the NOx concentration and the measured value Rm exceeds a predetermined value, or the difference between the predicted value Te for the unburned component in ash and the measured value Rm exceeds the predetermined value In the case, in step S65, it is determined to execute relearning of the prediction model Mm. Conversely, when the difference between the predicted value Te for the NOx concentration and the measured value Rm is within the predetermined value in steps S63 to S64, the difference between the predicted value Te for the unburned portion in ash and the measured value Rm is a predetermined value If not, the flow ends without executing step S65.

  Further, in some embodiments, as shown in FIG. 7, in the combustion condition determination method, the relearning of the command value determination model Mv is performed when the measurement value Rm of the evaluation target index T does not satisfy the predetermined criteria Tc. It may further include a relearning determination step (instruction value determination model relearning determination step) that determines that it is necessary. In the embodiment shown in FIG. 7, in step S71, the measured value Rm of the evaluation object index T (NOx concentration, unburned component in ash) measured by the LIBS device or the like is acquired. Then, in steps S72 to S73, when the measured value Rm of the NOx concentration does not satisfy the predetermined criteria Tc (NOx concentration> guaranteed value), or the measured value Rm of the unburned portion in ash does not satisfy the predetermined criteria Tc In the case (unburned in ash> prescribed value), in step S74, it is determined to execute relearning of the command value determination model Mv. Conversely, in S72 to S73, when the measured value Rm of the NOx concentration satisfies the predetermined criteria Tc (NOx concentration 保証 guaranteed value), and when the measured value Rm of the unburned portion in ash satisfies the predetermined criteria Tc ( The flow is ended without executing step S74 for unburned carbon in ash ≦ prescribed value).

  The above-described relearning determination step (FIGS. 6 to 7) may be performed, for example, periodically, separately from the flow of FIG. 5B. For example, in this case, steps S61 to S62 of FIG. 6 may be more specifically steps S2 to S4 of FIG. 5B. Or you may perform in combination with the flow of FIG. 5B. Specifically, the relearning determination step (FIGS. 6 to 7) may be performed before and after steps S5 and S6 after step S4 in the flow of FIG. 5B, or may be performed in parallel with step S5. It is good. In this case, steps S61 and S62 of FIG. 6 may be performed by steps S2 to S4 of FIG. 5B, and steps S63 to S65 of FIG. 6 may be performed after step S4 of FIG. 5B. As a result, it is possible to maintain the prediction accuracy by the prediction model Mm and the accuracy (reliability) by the command value determination model Mv at the execution timing of the next coal change or the like.

  The present invention is not limited to the above-described embodiments, and includes the embodiments in which the above-described embodiments are modified, and the embodiments in which these embodiments are appropriately combined.

1 combustion condition determination device m storage device 12 measurement device 2 front fuel component acquisition unit 3 command value determination unit 31 temporary setting value acquisition unit 32 evaluation target index prediction unit 33 criteria determination unit 34 parameter adjustment unit 35 command value selection unit 4 operation state Evaluation index acquisition unit 5 Adjustment value determination unit 51 Evaluation target index adjustment value calculation unit 52 Ignition distance adjustment value calculation unit 53 Second ignition distance adjustment value calculation unit 54 Fuel particle size adjustment value calculation unit 6 Relearning determination unit 7 Combustion system 7c Combustion control device 71 Coal storage facility 72 Coal hopper 73 Coal supply device 75 Exhaust gas piping 76 Denitrification device 77 Air preheater 78 Electric precipitator 8 Combustion furnace 8f Combustion chamber 8p Flue 82 Heat transfer tube group 83 Air box 84 Heart bottom hopper section 84b Bottom 9 Combustion device 91 Burner 92 Mill device 93 Port 93a Amount adjustment valve 94 Wind box damper

L Air supply pipe L1 Transport air supply pipe L2 Combustion air supply pipe Lf Fine-powdered fuel pipe F Front fuel G Exhaust gas A Outside air A1 Primary air A2 Secondary air

C Fuel component Cp Pre-fuel component T Evaluation target index (unburned content in ash, NOx concentration)
Tc Criteria to be evaluated indicator Te Predicted value to be assessed indicator P Combustion control parameter Pv Command value Pp Temporary set value

Mf Prediction map Mm Prediction model Mv Command value determination model R Operating condition evaluation index Rm Measured value of driving condition evaluation index Ib Flammability index S Fuel particle size B Ash adhesion amount D Ignition distance

Claims (23)

A combustion condition determination device for a combustion furnace that determines a combustion condition including a plurality of combustion control parameters for controlling a combustion apparatus of the combustion furnace, the combustion condition determining apparatus comprising:
A pre-combustion fuel component acquisition unit for acquiring a pre-combustion fuel component including a measurement value of at least a part of the fuel component of the pre-combustion fuel that is the fuel supplied to the burner that supplies the fuel and air into the furnace of the combustion furnace When,
Based on the pre-combustion fuel component, at least one evaluation target index, which is at least one of an unburned portion in ash of ash produced by the combustion of the pre-combustion fuel or NOx concentration in exhaust gas, satisfies a predetermined criterion. And a command value determination unit that determines a command value of the combustion control parameter. The command value determination unit
A provisional setting value acquisition unit that acquires a provisional setting value of the combustion control parameter;
An evaluation target index prediction unit that calculates a predicted value of the evaluation target index based on the temporary setting value and the pre-combustion fuel component each time the temporary setting value acquisition unit acquires the temporary setting value;
A criteria determination unit that determines whether the predicted value satisfies the predetermined criteria;
The temporary setting value of the combustion control parameter acquired by the temporary setting value acquiring unit is changed and the temporary setting is changed until it is determined that the predicted value satisfies the predetermined criterion by the criteria determining unit. A parameter adjustment unit that causes the temporary setting value acquisition unit to acquire a value;
The command value selection unit having the temporary set value as the command value when the criterion determination unit determines that the predicted value satisfies the predetermined criteria. Combustion condition determination device for combustion furnace.   The evaluation target index prediction unit uses the flammability index and the temporary setting value using a flammability index indicating the combustibility of the pre-combustion fuel, and a prediction map indicating the relationship between the evaluation target index and the combustion control parameter. The apparatus according to claim 2, wherein the predicted value is calculated from   The command value determination unit may measure the measured value of the evaluation target index generated by the combustion of the pre-combustion fuel component and the pre-combustion fuel component including the pre-combustion fuel component, and the combustion when the pre-combustion fuel is burned The prediction value is calculated from the pre-combustion fuel component and the temporary setting value using a prediction model created by machine learning teacher data composed of a plurality of data associated with the command value of the control parameter. The apparatus for determining the combustion conditions of a combustion furnace according to claim 2, wherein:   It further comprises a relearning determination unit that determines whether or not to perform relearning based on the difference between the predicted value of the evaluation target index and the measurement value of the evaluation target index. The combustion condition determination apparatus of the combustion furnace as described in.   The command value determination unit is a measurement value of the evaluation target index generated by the combustion of the pre-combustion fuel component including the pre-combustion fuel component and the pre-combustion fuel component, and the measurement value satisfying the predetermined criteria The combustion using a command value determination model created by machine learning of teacher data composed of a plurality of data associated with the command value of the combustion control parameter when the pre-combustion fuel is burned The apparatus according to claim 1, wherein the command value is determined from a pre-fuel component. A driving state evaluation index acquisition unit that acquires measurement values of the driving state evaluation index including the evaluation target index;
An adjustment value determination unit that determines an adjustment value for adjusting the command value of the combustion control parameter determined by the command value determination unit based on the measured value of the driving state evaluation index and the pre-combustion fuel component; The combustion condition determination device of the combustion furnace according to any one of claims 1 to 6, further comprising The driving state evaluation index includes the evaluation target index,
The driving state evaluation index acquisition unit acquires a measurement value of the evaluation target index,
The said adjustment value determination part is characterized by having an evaluation object index adjustment value calculation part which adjusts the command value of the said combustion control parameter based on the measured value of the said evaluation object parameter | index, and the said pre-combustion fuel component. The combustion condition determination apparatus of the combustion furnace as described in 7. The operating condition evaluation index includes at least one of the unburned component in the ash or the NOx concentration,
The operation state evaluation index acquisition unit acquires at least one of the measurement value of the unburned ash content in the ash or the NOx concentration,
The adjustment value determination unit determines, when the measured value of at least one of the unburned ash content and the NOx concentration is larger than a planned value, between the tip of the burner and the ignition position of the fuel in the furnace. The combustion condition determination device for a combustion furnace according to claim 7 or 8, further comprising an ignition distance adjustment value calculation unit that calculates an adjustment value for adjusting the combustion control parameter so as to shorten an ignition distance. The operating condition evaluation index includes the unburned portion in the ash, and the amount of the ash attached to the height position of the burner on the wall in the furnace,
The operation state evaluation index acquisition unit acquires the measurement value of the unburned portion in ash and the measurement value of the adhesion amount of the ash,
If the measured value of the unburned part in ash is higher than the unburned part in ash planned value and the adhered amount of the ash is smaller than the adhered amount planned value, the adjustment value determination unit A second ignition distance adjustment value calculation unit for calculating the adjustment value for adjusting the combustion control parameter so that the ignition distance between the tip of the burner and the ignition position of the fuel becomes short in the The combustion condition determination apparatus of the combustion furnace of any one of Claims 7-9. The operating condition evaluation index includes the particle size of the pulverized fuel which is the pre-combustion fuel generated by a mill, or the fuel particle size of the pre-combustion fuel supplied from the burner into the furnace.
The operation state evaluation index acquisition unit acquires a measurement value of the fuel particle size,
The said adjustment value determination part has a particle size adjustment value calculation part which calculates the adjustment value which adjusts the command value of the said combustion control parameter based on the measured value of the said fuel particle size, It is characterized by the above-mentioned. The combustion condition determination apparatus of the combustion furnace of any one of 10. The combustion condition determination device for a combustion furnace according to any one of claims 1 to 11, which determines a combustion condition including a plurality of combustion control parameters for controlling a combustion apparatus of the combustion furnace.
A measuring device capable of measuring in real time an operation state evaluation index for evaluating the operation state of the combustion furnace;
A combustion control device for transmitting the command value of the at least one combustion control parameter determined by the combustion condition determination device to the combustion device. A combustion condition determination method for a combustion furnace, comprising: determining a combustion condition including a plurality of combustion control parameters for controlling a combustion apparatus of the combustion furnace, the combustion condition determining method comprising:
A pre-combustion fuel component acquisition step of acquiring a pre-combustion fuel component including a measurement value of at least a part of a fuel component of the pre-combustion fuel which is the fuel supplied to a burner supplying the fuel and air into the furnace of the combustion furnace When,
Based on the pre-combustion fuel component, at least one evaluation target index, which is at least one of an unburned portion in ash of ash produced by the combustion of the pre-combustion fuel or NOx concentration in exhaust gas, satisfies a predetermined criterion. And a command value determination step of determining a command value of the combustion control parameter. The command value determination step is
A provisional setting value acquisition step of acquiring a provisional setting value of the combustion control parameter;
An evaluation target index prediction step of calculating a predicted value of the evaluation target index based on the temporary setting value and the pre-combustion fuel component each time the temporary setting value is obtained by the temporary setting value obtaining step;
A criteria determination step of determining whether the predicted value satisfies the predetermined criteria;
The provisional setting value of the combustion control parameter acquired by the provisional setting value acquisition step is changed and the provisional setting is changed until it is determined that the predicted value satisfies the predetermined criterion in the criteria determination step. A parameter adjustment step of causing the temporary set value acquisition step to acquire a value;
14. A command value determination step of using the temporary set value as the command value when it is determined by the criteria determination step that the predicted value satisfies the predetermined criteria. How to determine the combustion conditions of the combustion furnace.   The evaluation target index prediction step uses the combustion index and the temporary setting value by using a flammability index indicating the combustibility of the pre-combustion fuel, and a prediction map indicating the relationship between the evaluation target index and the combustion control parameter. The combustion condition determination method of a combustion furnace according to claim 14, wherein a predicted value of the evaluation target index is calculated from the above.   In the command value determination step, the measured value of the evaluation target index generated by the combustion of the pre-combustion fuel component and the pre-combustion fuel component having the pre-combustion fuel component, and the above-described combustion component The command value is calculated from the pre-combustion fuel component and the temporary setting value using a prediction model created by machine learning teacher data composed of a plurality of data associated with the command value of the combustion control parameter. The method for determining the combustion condition of a combustion furnace according to claim 14, wherein the method is determined.   17. The method according to claim 16, further comprising a relearning determination step of determining whether or not relearning is to be performed based on a difference between the predicted value of the evaluation target indicator and the measurement value of the evaluation target indicator. The combustion condition determination method of the combustion furnace as described in.   The command value determining step is a measurement value of the evaluation target index that is generated by the combustion of the pre-combustion fuel component having the pre-combustion fuel component and the pre-combustion fuel component, and the measurement value satisfying the predetermined criteria Using a command value determination model created by machine learning of teacher data composed of a plurality of data associated with the command value of the combustion control parameter when the pre-combustion fuel component is burned; The method according to claim 13, wherein the command value is determined from the pre-combustion fuel component. A driving state evaluation index acquisition step of acquiring a measurement value of the driving state evaluation index including the evaluation target index;
An adjustment value determination step of determining an adjustment value for adjusting the command value of the combustion control parameter determined by the command value determination step based on the measured value of the operating condition evaluation index and the pre-combustion fuel component; The combustion condition determination method of the combustion furnace according to any one of claims 13 to 18, further comprising The driving state evaluation index includes the evaluation target index,
The driving state evaluation index acquisition step acquires a measurement value of the evaluation target index,
The adjustment value determining step has an evaluation object index adjusting value calculating step of adjusting a command value of the combustion control parameter based on the measured value of the evaluation object index and the pre-combustion fuel component. The combustion condition determination method of the combustion furnace as described in 19. The operating condition evaluation index includes at least one of the unburned component in the ash or the NOx concentration,
The operating state evaluation index acquiring step acquires at least one of measured values of the unburned component in the ash or the NOx concentration,
The adjustment value determination step is performed between the tip of the burner and the ignition position of the fuel in the furnace if the measurement value of at least one of the unburned ash content and the NOx concentration is larger than a planned value. 21. The combustion condition determination method for a combustion furnace according to claim 19, further comprising: an ignition distance adjustment value calculating step of calculating an adjustment value for adjusting the combustion control parameter so as to shorten an ignition distance. The operating condition evaluation index includes the unburned portion in the ash, and the amount of the ash attached to the height position of the burner on the wall in the furnace,
The operation state evaluation index acquiring step acquires a measurement value of the unburned portion in the ash and a measurement value of the adhesion amount of the ash;
In the adjustment value determination step, when the measured value of the unburned part in ash is higher than the unburned part in ash planned value and the adhered amount of the ash is smaller than the planned amount of adhered amount, the inside of the furnace A second ignition distance adjustment value calculating step of calculating the adjustment value for adjusting the combustion control parameter so that the ignition distance between the tip of the burner and the ignition position of the fuel becomes short in The combustion condition determination method of the combustion furnace according to any one of claims 19 to 21. The operating condition evaluation index includes the particle size of the pulverized fuel which is the pre-combustion fuel generated by a mill, or the fuel particle size of the pre-combustion fuel supplied from the burner into the furnace.
The operation state evaluation index acquisition step acquires a measurement value of the fuel particle size,
The adjustment value determining step has a particle size adjustment value calculating step of calculating an adjustment value for adjusting the command value of the combustion control parameter based on the measurement value of the fuel particle size. 22. The combustion condition determination method of the combustion furnace as described in any one of 22.

Images