CN101713533B - Control device and method of thermal power generation plant - Google Patents

Abstract

The invention provides a control apparatus of boils which can reduce carbon monoxide in exhaust of a boiler when bias is existed in coal dust flow putted from a plurality of burners. The control apparatus measures the flow rate of fuel supplied to the burner to each burner, further comprising: a graphical unit which generates the flow pattern of fuel flow based on the measuring value measured by a measurer; an operation signal generating unit which calculates airflow supplied to the burner or an air opening based on pattern information generated by the graphical unit.

Description

The control device of steam power plant and control method

The present invention is that application number that applicant Hitachi Co., Ltd submitted to Patent Office of the People's Republic of China in 2008 in 01 month 30 days is 2008100053060, denomination of invention is divided an application for the invention application of " control device of boiler and control method ".

Technical field

The present invention relates to the control device and the control method of boiler.

Background technology

For using coal to act as a fuel for the steam power plant (plant) of generating, require to reduce carbon monoxide (CO), nitrogen oxide (NO _x) etc. the discharge rate of carrying capacity of environment material.

Based on this background, proposing has the CO of reduction and NO _xBurner (burner), air port (air port) structure.For example, NO is hanged down in the realization of in TOHKEMY 2005-273973 communique, putting down in writing _xThe burner of changing reduces NO when constructing and in TOHKEMY 2006-162185 communique, putting down in writing _xTuyere structure with CO.

Which kind of technology no matter in TOHKEMY 2005-273973 communique and the TOHKEMY 2006-162185 communique all adopts two grate firings to burn as method for burning coal.This combustion method is that a kind of make from burner supplied with the coal that comes out and under the state of lack of air, burnt the method for the air that supply completing combustion is used from the air port afterwards.

The control technology of the air mass flow of supplying with out from burner and air port as operation, the method in japanese kokai publication hei 5-33906 communique on the books.In japanese kokai publication hei 5-33906 communique, put down in writing according to the oxygen (O that makes in the boiler export exhaust ₂) setting value of concentration and the O that measures ₂The method of the air mass flow that the consistent mode of concentration value determines to supply with out from burner and air port.And also having put down in writing should technology through utilizing, unburnt ingredient and NO that can be in boiler exhaust _xConcentration surpass in the scope of limits value, reduce operating cost.

Patent documentation 1: TOHKEMY 2005-273973 communique

Patent documentation 2: TOHKEMY 2006-162185 communique

Patent documentation 3: japanese kokai publication hei 5-33906 communique

In general, burner is configured to a plurality of in boiler.In japanese kokai publication hei 5-33906 communique, record the technology of decision from the air total amount of burning mouth input.But,,, perhaps the total amount of air mass flow is pressed predetermined pro rate in each burner so adopt the air mass flow homogeneous that makes each burner owing to there be not the technology relevant with the air mass flow of individually setting a plurality of burners.

Yet even if be certain state in the total amount of pulverized coal flow, the pulverized coal flow that supplies to each burner still can't reach homogeneous in each burner, change in time sometimes.Produce deviation in case supply to the pulverized coal flow of burner, then will produce air mass flow many relatively burner and few relatively burner.If air mass flow is on the low side, coal dust can't completing combustion, owing to this reason possibly produce carbon monoxide.

And, when the pulverized coal flow of the burner that is configured in side wall portion increases, when the pulverized coal flow that perhaps is configured in the burner of central portion increases etc., the flow figure (pattern) of pulverized coal flow can have multiple supposition.The flow figure is different, and is also different in the carbonomonoxide concentration of boiler export.

Summary of the invention

The object of the present invention is to provide and a kind ofly consider that the flow figure of pulverized coal flow decides the air mass flow that from burner, drops into, and reduces the control device of carbonomonoxide concentration thus.

And, from another viewpoint, in the technology that japanese kokai publication hei 5-33906 communique is put down in writing, because O ₂The measurement of concetration device is the O that measures in exhaust to be contained through the part in the exhaust of boiler export stream through extracting out ₂Concentration is so though can pass through this O ₂The measurement of concetration device is grasped the O in 1 place in the boiler export stream of the exhaust of being extracted out ₂Concentration, but the O in other places in the unclear boiler export stream of extracting exhaust out ₂Concentration.

For this reason at the O of the exhaust of the boiler export stream of flowing through ₂When CONCENTRATION DISTRIBUTION has deviation, even if by O ₂The O in the place that the measurement of concetration device is measured ₂Concentration value is very high, the O in other places ₂Concentration also might be very low.In the boiler export stream, O ₂So the less zone of concentration since CO is not oxidized just maybe residual CO in exhaust.

Yet, in the above-mentioned technology of in japanese kokai publication hei 5-33906 communique, putting down in writing, owing to be with O ₂The typical value or the mean value of concentration measurement are set at desired value, control and supply with the air mass flow of coming out in burner and the air port from be arranged on boiler, so in the exhaust of flowing through in the boiler export stream O ₂When CONCENTRATION DISTRIBUTION has deviation, can't detect the O that makes the CO oxidation in the boiler export stream exactly ₂The zone that concentration is less, therefore uncontrollable to above-mentioned O ₂Necessary air mass flow is supplied with in concentration zone seldom, is difficult to reduce effectively CO amount in the boiler exhaust.

Another object of the present invention is to provide a kind of in the boiler that uses coal to act as a fuel, the O in the exhaust of boiler export ₂CONCENTRATION DISTRIBUTION exists under the situation of deviation, can be to O ₂The control device of the steam power plant of CO in the exhaust that necessary air mass flow reduces boiler and the control method of steam power plant are supplied with in the less zone of concentration.

To achieve these goals, the control device with boiler is constructed as follows.Have: a plurality of burners that fuel and air are supplied with in boiler; Be positioned at the burning gases that make the fuel that supplies to this burner and air burning and generate the flow direction the downstream and to the air fed air port of these burning gases; Adjustment is in the control device of the boiler of the operating side of the air mass flow of above-mentioned burner and air port supply; Has the measuring appliance that flow fuel that the above-mentioned burner of subtend supplies with is measured; The control device of above-mentioned boiler possesses: graphical unit; It generates the flow figure of fuel flow rate based on the measured value that is measured by above-mentioned measuring appliance; And the operation signal generation unit, it calculates the air mass flow to above-mentioned burner or the supply of above-mentioned air port based on by the graphical information that above-mentioned graphical unit generated.

In addition, the control device of steam power plant of the present invention possesses boiler, and this boiler has: with air and the burner of the coal that acts as a fuel to the boiler supply; With the downstream that is arranged in the flow direction that makes the burning gases that generate from this burner air of supplying with and the coal combustion that acts as a fuel and to the air fed air port of these burning gases; Wherein, In the boiler of steam power plant, possesses measuring appliance; It is measured oxygen concentration or carbonomonoxide concentration in the burning gases of this boiler export; In the controller of the control device that constitutes steam power plant, possess respectively: arrive the zone and infer the unit, it infers the zone that the air of from the burner of above-mentioned boiler or air port, being supplied with arrives boiler export; With the operation signal generation unit; Its based on the measured value of the measured value of carbonomonoxide concentration in the burning gases of the boiler export of measuring by above-mentioned measuring appliance or oxygen concentration and by above-mentioned arrival zone infer the air that is supplied to of inferring out the unit arrive boiler export the zone infer the result; Setting supplies to the air mass flow the boiler from burner or air port, increases so that arrive the air mass flow in the low zone of carbonomonoxide concentration is high in the burning gases of the boiler export of being measured by above-mentioned measuring appliance zone or oxygen concentration.

(invention effect)

According to the present invention, when there is deviation in the pulverized coal flow that in by a plurality of burners, drops into, can realize reducing effectively the control device of the boiler of carbon monoxide in the exhaust of boiler and the control method of boiler.

In addition, through the present invention, be implemented in the O in the exhaust of adopting the boiler export that coal acts as a fuel ₂Exist in the CONCENTRATION DISTRIBUTION under the situation of deviation, to O ₂Control device and the control method of steam power plant that necessary air mass flow reduces effectively the steam power plant of the CO in the exhaust of boiler supplied with in the few zone of concentration.

Description of drawings

Fig. 1 is the block diagram that the system of the boiler controlling device of expression embodiment constitutes.

Fig. 2 is the key diagram of the formation of steam power plant.

Fig. 3 is the key diagram in path that supplies to coal dust and the air of burner.

Fig. 4 is the block diagram that the system of expression operation signal generation unit constitutes.

Fig. 5 is the key diagram of the generation method of operation signal.

Fig. 6 is the action flow chart of control device of the boiler of embodiment.

Fig. 7 is the key diagram of numeric value analysis execution in step.

Fig. 8 is that model is set up, the key diagram of the action of unit.

Fig. 9 is the key diagram that is kept at the data in learning information database and the graphic data base.

Figure 10 is the key diagram that figure appends step.

Figure 11 is the time dependent figure of explanation measured value.

Figure 12 is the embodiment that is presented at the picture of image display device, will be presented at the figure on the same picture to pulverized coal flow and the air mass flow that burner is supplied with.

Figure 13 is the embodiment that is presented at the picture of image display device, is that shape and the characteristics/properties with figure is presented at the figure on the picture.

Figure 14 is the block diagram that the integral body of control device of the steam power plant of expression one embodiment of the invention constitutes.

Figure 15 is the summary pie graph of the steam power plant of the control object of the control device of steam power plant among the expression embodiment shown in Figure 14.

Figure 16 is the flow chart of the control sequence of the control device of steam power plant among the expression embodiment shown in Figure 14.

Figure 17 is the control block diagram of the formation of the operation signal generation unit of the control device of steam power plant among the expression embodiment shown in Figure 14.

Figure 18 is the key diagram that the function of unit is inferred in the arrival zone of the control device of steam power plant among the embodiment shown in Figure 14.

Figure 19 is the key diagram of the generation method of the operation signal of the operation signal generation unit of the control device of steam power plant among the embodiment shown in Figure 14.

Figure 20 is the flow chart of the order of operation of the learning algorithm of the unit of the control device of steam power plant among the expression embodiment shown in Figure 14.

Figure 21 is arranged on an example of the display frame of the image display device of the control device of steam power plant among the embodiment shown in Figure 14.

Figure 22 is arranged on another example of the display frame of the image display device of the control device of steam power plant among the embodiment shown in Figure 14.

Figure 23 is the control block diagram of all formations of control device of the steam power plant of expression other embodiment of the present invention.

Figure 24 is the key diagram that the function of unit is inferred in the distribution of control device of the steam power plant of other embodiment shown in Figure 23.

Figure 25 is the flow chart of control sequence of control device of the steam power plant of expression other embodiment shown in Figure 23.

Symbol description is following:

The 100-steam power plant; The 200-control device; 201,920-outer input interface; 202, the outside output interface of 940-; 210-measuring-signal database; The 220-graphic data base; 230-numeric value analysis result database; 240-operation signal database; 250-control logic database; 260-learning information database; The 270-equipment information database; The graphical unit of 300-; 400-numeric value analysis performance element; 500-operation signal generation unit; 600-unit; The 700-model; 800-evaluation of estimate computing unit; The 900-external input device; The 901-keyboard; The 902-mouse; 910 one dimension nurse tools; 930-data transmit-receive handling part; The 950-image display device; The 2100-steam power plant; The 2200-control device; The 2201-outer input interface; The outside output interface of 2202-; 2210-measuring-signal database; 2220-arrives regional database; 2230-numeric value analysis result database; 2240-operation signal database; 2250-control logic database; 2260-learning information database; 2300-arrives the zone and infers the unit; 2400-numeric value analysis performance element; 2500-operation signal generation unit; 2600-unit; The 2700-model; 2800-evaluation of estimate computing unit; The 2900-external input device; The 2901-keyboard; The 2902-mouse; The 2910-maintenance tool; The 2920-outer input interface; 2930-data transmit-receive handling part; The outside output interface of 2940-; The 2950-image display device.

The specific embodiment

Below, describe with reference to the control device of accompanying drawing the boiler of the embodiment of the invention.

(embodiment 1)

Fig. 1 is that the system of embodiment of the control device of expression boiler constitutes block diagram.In Fig. 1, steam power plant 100 is by control device 200 controls.

In the control device 200 controlled of steam power plant 100, be provided with graphical unit 300, numeric value analysis performance element 400, operation signal generation unit 500, unit 600, model 700, evaluation of estimate computing unit 800 as arithmetic unit to the control object.

In addition; In control device 200, be provided with measuring-signal database 210, graphic data base 220, numeric value analysis result database 230, operation signal database 240, control logic database 250, learning information database 260, equipment information database 270 as database.

In addition, in control device 200, as with the interface of outside, be provided with outer input interface 201 and outside output interface 202.

In control device 200, through outer input interface 201, measuring-signal 1 is taken into control device 200 from steam power plant 100, the measuring-signal 2 that is taken into is stored in the measuring-signal database 210.And the operation signal 17 in that operation signal generation unit 500 generates is kept in the operation signal database 240, simultaneously to outside output interface 202 transmission.Pass through the operation signal 18 of outside output interface 202, be transferred to steam power plant 100.

In numeric value analysis performance element 400; Facility information 19 and the physical model of simulation steam power plant 100 in the equipment information database of design information of steam power plant 100 preserved in utilization, and carrying out with steam power plant 100 is the numeric value analysis of object.Through carrying out the structure in the boiler will constitute steam power plant 100, burner, air port, and, predict the operation characteristic of steam power plant 100 to burner, the fuel flow rate of air port supply, the calculating that air mass flow is set at boundary condition.For example; Numeric value analysis performance element 400; Utilization is preserved the facility information 19 in the equipment information database of boiler design information and is simulated the physical model of above-mentioned boiler; The fuel flow rate that calculating is supplied with to burner or the service condition of the boilers such as air mass flow given of confession aweather, and the relation between at least one in carbonomonoxide concentration, the nitrous oxides concentration in the discharge gas of boiler.Through carrying out the numeric value analysis information 6 that numerical value resolution executing 400 obtains, be stored in the numeric value analysis result database 230.

In graphical unit 300; Utilization is kept at the measuring-signal data 4 in the measuring-signal database 210 and is kept at the numeric value analysis data 7 in the numeric value analysis result database 230, will be graphical to the fuel flow rate that the boiler mouth that constitutes steam power plant 100 is supplied with.In addition, for the formation of steam power plant 100, use Fig. 2 and Fig. 3 back to narrate.Graphical information 5 in that graphical unit 300 generates is stored in the graphic data base 220.

In unit 600, be object with model 700, the method for operating of steam power plant 100 is learnt.The control characteristic of model 700 simulation steam power plants 100.Just; Send to steam power plant 100 with the operation signal 18 that will be generated by control device 200; And it is identical as the situation of this control result's measuring-signal 1 by control device 200 receptions; Also will in model 700, send, and receive model output 10 as its control result by unit 600 by the model input 9 that unit 600 is generated.Model 700 utilizes and is kept at the numeric value analysis data 8 in the numeric value analysis result database 230 and is kept at the measuring-signal data 4 in the measuring-signal database 210, calculates with model and imports 9 corresponding model outputs 10.Model 700 for example utilizes the statistical model of neutral net etc. to set up.In model 700, before the steam power plant operation, only utilize the numeric value analysis data 8 that are kept in the numeric value analysis result database 230 to come model output 10 is calculated.Afterwards, 4 pairs of model outputs 10 of dual-purpose measuring-signal are calculated.Thus; The physical model of when making 400 actions of numeric value analysis performance element, using and the characteristic of steam power plant 100 are not simultaneously; Come model output 10 is calculated through paying attention to measuring-signal data 4, can make the characteristic of the characteristic of model 700 near steam power plant 100.

In unit 600, the generation method that can reach the model input 9 of desired value to the model output 10 of being calculated by model 700 is learnt.As the index of using in the generation method of unit 600 learning models input 9, can utilize the evaluation of estimate of calculating by evaluation of estimate computing unit 800 11.In evaluation of estimate computing unit 800, when model output 10 was the state of expectation, evaluation of estimate 11 was bigger, and departs from the state of expectation more, and the value of evaluation of estimate 11 reduces more.

Unit 600 is set up through the various optimization gimmicks such as calculating gimmick of using intensified learning, evolution.In unit 600, can reach maximum method of operating to the evaluation of estimate of calculating by evaluation of estimate computing unit 800 11 and learn.The learning information data 13 of the constraints of using in the study, model export target value etc. are stored in the learning information database 260.In addition, be learning information 12 as the result who learns by unit 600, be stored in the learning information database 260.

In operation signal generation unit 500, obtain following data as required, and utilize these information to generate the operation signal 17 that steam power plant 100 is controlled.These data are: be kept at measuring-signal data 3 in the measuring-signal database 210, be kept at graph data 16 in the graphic data base 220, be kept at the learning information data 14 in the learning information database 260 and be kept at the control logic data 15 in the control logic database 250.

The operator of steam power plant 100; Through utilizing the external input device 900 that constitutes by keyboard 901 and mouse 902 to generate maintenance tool input signal 51; And this signal is input in the maintenance tool 910, can the information that be configured in the database in the control device 200 be presented on the image display device 950 thus.

Maintenance tool 910 is made up of outer input interface 920, data transmit-receive portion 930 and outside output interface 940.

Externally the maintenance tool input signal 51 that generated of input unit 900 is taken in the maintenance tool 910 through outer input interface 920.In the data transmit-receive portion 930 of maintenance tool 910,, from the database that is configured in control device 200, obtain database information 50 according to the information of maintenance tool input signal 52.

In data transmit-receive portion 930, deal with data library information 50 resulting maintenance tool output signals 53 are sent to outside output interface 940.Maintenance tool output signal 54 is presented on the image display device 950.

In addition; In the control device 200 of the foregoing description; Though measuring-signal database 210, graphic data base 220, numeric value analysis result database 230, operation signal database 240, control logic database 250, learning information database 260, equipment information database 270, graphical unit 300, numeric value analysis performance element 400, unit 600, model 700 and evaluation of estimate computing unit 800 are configured in the inside of control device 200, also can with its all, or wherein a part of outside that is configured in control device 200.

Fig. 2 is the figure of schematic representation steam power plant 100.In the boiler 101 that constitutes steam power plant; Be provided with supply with coal dust that is through grinding mill (mill) 110 with coal grinding the burner 102 of auxiliary air of fuel, primary air that the coal dust conveyance is used and burning adjustment usefulness, make coal dust through 102 supplies of this burner in the internal-combustion of boiler 101.In addition, coal dust and primary air are from pipe arrangement 134, and auxiliary air imports to burner 102 from pipe arrangement 141.

In addition, the make-up gas (after air) that in boiler 101, is provided with two grate firings burning usefulness drops into the additional air port 103 in the boiler 101, and make-up gas is imported to by pipe arrangement 142 and replenishes 103 places, air port.

Produced the burning gases of high temperature by the burning of coal dust, flowed to the downstream, carried out heat exchange through the heat exchanger 106 that is configured in boiler 101 again, passed through air heater 104 afterwards along the inner track of boiler 101.Passed through the gas of air heater 104, after applying pump-down process, by smoke stack emission in atmosphere.

The feedwater of circulation in the heat exchanger 106 of boiler 101 is fed water through supplying with in feed pump 105 heat exchanger 106, and the burning gases heating that in heat exchanger 106, is flowed down along boiler 101, becomes the steam of HTHP.In addition, in the present embodiment,, also can dispose a plurality of heat exchangers though the number of heat exchanger is made as one.

Passed through the steam of the HTHP of heat exchanger 106, import in the steam turbine 108 through turbine type burner (タ one PVC Application ガ バ one Na one) 107, and the energy drives steam turbine 108 that utilizes steam to have has made generator 109 generate electricity.

In steam power plant, dispose the measuring appliance of the running status of multiple detection steam power plant, the measuring-signal of the equipment that from these measuring appliances, obtains sends in control device 200 as measuring-signal 1.For example, in Fig. 2, illustrate flow measuring probe 150, temperature meter 151, pressometer 152, generating output checker 153 and measurement of concetration device 154.

In flow measuring probe 150, the flow of the feedwater to boiler 101, supplied with from feed pump 105 is measured.In addition, temperature meter 151 and pressometer 152 are to measuring to the temperature and pressure of the steam of steam turbine 108 supplies from heat exchanger 106.

Through the electric weight of generator 109 generatings, measure by generating output checker 153.Relate to the composition (CO, the NO that are contained in the burning gases through boiler 101 _XDeng) information of concentration, the measurement of concetration device 154 in downstream that can be through being arranged on boiler 101 is measured.

In addition, in general, except that shown in Figure 2, also in steam power plant, dispose a plurality of measuring appliances, but omitted diagram here.

Below, the path of 103 input make-up gas describes to the path of putting into inner primary air of boiler 101 and auxiliary air from burner 102 and from additional air port.

Primary air imports pipe arrangement 130 by fan 120; And be diverted to halfway in pipe arrangement 132 and the pipe arrangement 131; And collaborate once more at pipe arrangement 133 places; Thereby import in the grinding mill 110 of the upstream side that is arranged on burner 102, wherein pipe arrangement 132 is through the air heater 104 in the downstream that is arranged on boiler 101, pipe arrangement 131 without and bypass (bypass) is crossed air heater 104.

Through the air of gas heater 104, receive the burning gases heating that flows down along boiler 101.Utilize this primary air, in grinding mill 110 by ground coal dust with air conveyance together in burner 102.

Auxiliary air and make-up gas import pipe arrangement 140 from fan 121; And receive equally after gas heater 104 heating; Be diverted in the pipe arrangement 142 that pipe arrangement 141 that auxiliary air uses and make-up gas use, thereby import to respectively in burner 102 and the additional air port 103.

Fig. 3 is the figure that the path of the coal dust supplied with of opposed firing mouth and air describes.

Shown in Fig. 3 (a), in boiler 101, dispose a plurality of burners along boiler 101 widths.In Fig. 3 (a), though represent that to disposing 5 burner 102A, 102B, 102C, 102D, 102E its quantity is arbitrarily.And in Fig. 3 (a),, also can be configured to multilayer though dispose burner with one deck along the short transverse of boiler.

Shown in Fig. 3 (b), each burner is connected with grinding mill through pipe arrangement 134A, 134B, 134C, 134D, 134E.In each pipe arrangement, dispose pulverized coal flow measuring appliance 155A, 155B, 155C, 155D, 155E respectively.Thus, the flow of input coal dust in burner 102A, 102B, 102C, 102D, 102E is all measured.The signal that measures sends to control device 200 as measuring-signal 1.In addition, in the present embodiment,, may not whole burners be configured though burner is all disposed the pulverized coal flow measuring appliance.In addition, also can come measuring mode configuration pulverized coal flow measuring appliance to compile the pulverized coal flow of supplying with by a plurality of burners.

In addition, auxiliary air is supplied with to each burner through pipe arrangement 141.Shown in Fig. 3 (b),, be provided with air damper (air damper) 160A, 160B, 160C, 160D, 160E in the inside of pipe arrangement 141.The flow of the secondary air flow of supplying with to each burner can be controlled through the aperture of adjustment air damper.The command signal of the aperture of air damper is also included within the middle of the operation signal 18 that is sent by control device 200.

Control device 200 utilizes the information to the pulverized coal flow of each burner supply, and decision supplies to the air mass flow in each burner.

Fig. 4 is the block diagram that the system of expression operation signal generation unit 500 constitutes, the block diagram of the command signal (command signal of air damper aperture) of the air mass flow be in the middle of the decision operation signal 17, supplying with to each burner.

In operation signal generation unit 500, be provided with reference signal generation unit 510, relative value calculating part 520, gain setting unit 530, upper lower limit value setup unit 540, multiplicative operator 550, switch 560,561, fixed value maker 570,571 and adder calculator 580 as arithmetic unit.

Reference signal generation unit 510 calculates the total air mass flow that is dropped into by the burning mouth, and it is assigned to each burner according to programme-control.When calculating reference signals 501 by reference signal generation unit 510, the mode that can reach homogeneous with the air capacity by whole burner inputs is calculated, and also can calculate the air mass flow that is dropped into by each burner according to the algorithm of being scheduled to (program).

In relative value calculating part 520,, calculate in each burner fuel flow rate with respect to the relative value of the fuel flow rate of the mean value of fuel flow rate based on the fuel flow rate that measures by measuring appliance.For example, utilize measuring-signal data 3, calculate the relative value 502 of the pulverized coal flow of each burner according to (1), (2) formula.Wherein, 1≤i≤i _Max, i _MaxBe the number of burner, r _iBe the relative value of the pulverized coal flow of burner i, CF _iBe measured value by the pulverized coal flow of burner i input, and CF _AverageBe mean value by the pulverized coal flow of burner input.

Several 1

$r_i=\frac{{\mathrm{Cf}}_i}{{\mathrm{CF}}_{\mathrm{average}}}-1...\left(1\right)$

Several 2

${\mathrm{CF}}_{\mathrm{average}}=\frac{1}{i_{\max }}\underset{i=1}{\overset{i_{\max }}{\mathrm{\Σ}}}{\mathrm{CF}}_i...\left(2\right)$

In addition, in the present embodiment, though with CF _iAs the measured value of the pulverized coal flow that drops into by burner i, but can calculate through the rolling average that utilizes measured value, perhaps low pass filter etc., with removed be included in noise in the measuring-signal signal as CF _i, and will utilize this CF _iThe r that calculates _iAs relative value 502.

In multiplicative operator 550, utilize relative value 502 and gain 509 basis (3) formula signal calculated 503.Wherein in (3) formula, s _iBe signal 503, G is gain 509.

Several 3

S _i＝G×r _i …(3)

Gain 509 utilizes gain setting unit 530, fixed value maker 570, switch 560 to calculate.

In switch 560; Two inputs of the fixed value (α) 508 that generates to the gain that calculates by gain setting unit 530 alternative 507 with at fixed value maker 570, will gain alternative 507 or fixed value (α) 508 in any signal export as gain 509.In addition, the value of the fixed value (α) 508 that in fixed value maker 570, generates can at random be set through control logic database 250 by the operator of steam power plant 100.

In gain setting unit 530, utilize measuring-signal data 3, generate gain alternative 507.In gain setting unit 530; Obtain the figure that is kept in the graphic data base 220 and the similarity of measuring-signal data 3; And from learning information database 260, extract air mass flow adjustment yield value with respect to the higher figure of similarity, as gain alternative 507.In the present embodiment, though will be fixed value G to the gain value settings that the relative value of each burner is taken advantage of the formula, also can change yield value by corresponding each burner as (3).In addition, the figure that is kept in the graphic data base 220 is narrated with the details back that is kept at the information in the learning information database 260.

In upper lower limit value setup unit 540, utilize signal 503, according to (4) formula correction signal alternative 504.Wherein, t _iBe that correction signal is alternative 504, t _MaxBe the higher limit of correction signal alternative 504, t _MinIt is the lower limit of correction signal alternative 504.And, t _MaxAnd t _MinCan at random set by the operator of steam power plant 100.

Several 4

$t_i=\left\{\begin{array}{cc}{S}_i& \mathrm{if}{t}_{m\mathrm{in}}\&le;\mathrm{Si}\&le;t_{\max }\\ t_{\max }& \mathrm{if}{t}_{\max }<S_i\\ t_{\min }& \mathrm{if}{S}_i<t_{\min }\end{array}\right....\left(4\right)$

In switch 561; Two inputs of the fixed value (0) 505 that generates to the correction signal of calculating by upper lower limit value setup unit 540 alternative 504 with by fixed value maker 571, with any signal of correction signal alternative 504 or fixed value (0) 508 as correction signal 506 outputs.In fixed value maker 571, generate null value.

In adder calculator 580, correction signal 506 and reference signal 501 are carried out add operation, calculating operation signal 17.The effect that utilization possesses switch 561 can make reference signal 501 consistent with operation signal 17.

As stated; Operation signal generation unit 500; Based on the graph data of fuel flow rate 16 decision air mass flows adjustment gain G, the fuel flow rate of the fuel flow rate of supplying with to burner that is based on that each burner measures calculates in each burner fuel flow rate with respect to the relative value of the fuel flow rate of the mean value of fuel flow rate; Based on air mass flow adjustment gain G and relative value, calculate the air mass flow Si that supplies with to burner.In addition, relative value is relative quantity both, also relative ratios.

Fig. 5 is an embodiment of reference signal 501, relative value 502, signal 503, operation signal 17, is the figure that the generation method to operation signal 17 describes.In addition, A～E is the symbol of establishing in order to discern burner among the figure.

Shown in Fig. 5 (a), the reference signal 501 by reference signal generation unit 510 generates is equal values at whole burners, if make it consistent with operation signal 17, the air mass flow of then being supplied with by whole burners is identical flow.Shown in Fig. 5 (b), the relative value 502 of utilizing (1) (2) formula to calculate is different each other at each burner.Fig. 5 (b) refer to is lacked than mean value by the pulverized coal flow that burner A, E supply with, the situation that the pulverized coal flow of being supplied with by burner B, C, D is Duoed than mean value.

Fig. 5 (c) multiply by the signal 503 that gain 509 calculates to relative value 502.Whole value of signal 503 is all at t _MinAbove t _MaxBelow, and in switch 561, when with correction signal alternative 504 during as correction signal 506, operation signal 17 is that the value of Fig. 5 (a) and Fig. 5 (c) is carried out the value after the add operation shown in Fig. 5 (d).

Through generating run signal 17 so, to the more burner of pulverized coal flow air supply more, for the less burner of pulverized coal flow air supply less.In addition, in the present embodiment,, also can adjust by replenishing the air mass flow that air port portion drops into according to figure though put down in writing to adjusting according to figure by the method for the air mass flow of burning mouth input.In addition, also can adjust the air mass flow of burning mouth and additional air port portion.

Below, will describe the generation method that is kept at the data in graphic data base 220, the learning information database 260 of reference in the gain setting unit 530 of Fig. 4 and alternative 507 the generation method of gaining.

Fig. 6 is the action flow chart of control device 200.As shown in Figure 6, control device 200 combination execution in step 1000,1010,1020,1030,1040,1050,1060,1070.Below will step separately be described.

At first in step 1000, make 400 actions of numeric value analysis unit, carry out the numeric value analysis of steam power plant 100.This as a result the numeric value analysis information 6 of gained be kept in the numeric value analysis result database 230.And the numeric value analysis data 7 that are kept in the numeric value analysis result database 230 are sent to graphical unit 300.The details back is narrated in Fig. 7 etc., in numeric value analysis data 7, contains the information relevant with the flow figure of coal dust.

In step 1010, judge whether implementing to make up the study of carrying out unit 600, model 700, evaluation of estimate computing unit 800.When implementing study, get into the path of " being (YES) ", when not implementing, get into the path of " not (NO) ".

In step 1020,, set up model 700 based on the data that are kept in numeric value analysis result database 230 and the measuring-signal database 210.After steam power plant 100 has just brought into operation, in measuring-signal database 210, do not store data.Under this situation, utilize and be kept at the numeric value analysis data 8 in the numeric value analysis result database 230, set up model 700.Afterwards, when obtaining measurement data 1, and when in measuring-signal database 210, storing data, so that the consistent mode correction model 700 of characteristic of model 700 and steam power plant 100.Utilize model 700 predictions by the CO, the NO that discharge in the steam power plant 100 _XConcentration etc.

In step 1030, ensemble learning unit 600, model 700, evaluation of estimate computing unit 800 are learnt the establishing method of the method for operating of steam power plant 100, the gain of air mass flow adjustment just.The learning information 12 that is obtained by step 1030 is kept in the learning information database 260.

In step 1040, make operation signal generation unit 500 action, generating run signal 17, and operation signal sent to steam power plant 100.

In step 1050, obtain with the operation signal 17 that generates in the step 1040 send to steam power plant 100 the result, be measuring-signal 1 and measuring-signal 2, and measuring-signal is kept in the measuring-signal database 210.

In step 1060, judge whether the flow figure to the pulverized coal flow of generation in the step 1010 has appended new figure.When having implemented figure and appended, enter into the path of " being ", when not implementing, enter into the path of " denying ".

In step 1070, carry out graphical unit 300, append figure.

In addition, in step 1010 and step 1060, enter into " being "/which of " denying ", can be in advance set by the operator of steam power plant 100.Also can estimate in addition, decide based on this evaluation result to enter into which of " being "/" denying " the performance, the performance of figure of study.

Below, utilize Fig. 7～Figure 10, will each movement content of key element among Fig. 6 be described.

Fig. 7 is the figure of the details of description of step 1000.Shown in Fig. 7 (a), can step 1000 segmentation be changed into step 1001,1002,1003.

In step 1001, the analysis condition that is used for carrying out numerical value resolution executing 400 is set.Fig. 7 (b) is an embodiment of analysis condition form.Shown in Fig. 7 (b), the pulverized coal flow and the air mass flow adjustment gain that are dropped into by burner A～E are set.The flow figure of the pulverized coal flow that is dropped into by burner A～E utilizes the facility information 19 that is kept in the equipment information database 270 to set.In Fig. 7 (b), enter into the situation (A～E all is 16kg/s) of burner A～E equably and the pulverized coal flow that dropped into by burner A is more and the less situation of pulverized coal flow that dropped into by burner B～E is an example with pulverized coal flow, put down in writing.

In facility information 19, contain: the characteristic information of the characteristic of expression grinding mill 110; With the length of the pipe arrangement that connects grinding mill and burner, be connected the relevant information of number of bends of the pipe arrangement of grinding mill and burner.Can calculate the pressure drop of the pipe arrangement that connects burner A～E and grinding mill through using in these information at least one.When the length of pipe arrangement is long, number of bends more for a long time, it is big that pressure drop becomes, the pulverized coal flow that supplies to burner might reduce.For example, suppose as Fig. 3 (b), grinding mill to be connected with burner.Shown in Fig. 3 (b), be zero comparing with the number of bends of the pipe arrangement that is connected burner A and grinding mill, the number of bends that connects the pipe arrangement of burner B～E and grinding mill is two.At this moment, the pressure drop of pipe arrangement that the pressure drop ratio that connects the pipe arrangement of burner B～E and grinding mill is connected burner A and grinding mill is bigger, and the pulverized coal flow of process still less.

Like this, utilize Equipment Design information, suppose the flow figure of multiple pulverized coal flow, and this flow figure is set at the boundary condition of parsing.

In addition, air mass flow adjustment gain is the value with gain 509 identical meanings of the multiplicative operator 50 of Fig. 4.Among Fig. 7 (b), make air mass flow adjustment gain each 0.1 size that changes in 0.1～2.0 scope, set analysis condition.The boundary condition of air mass flow multiply by air mass flow adjustment gain through the relative value to pulverized coal flow and decides.

In step 1002,, make 400 actions of numeric value analysis performance element based on the analysis condition that step 1001 is set.In step 1003, the numeric value analysis information 6 that implementation step 1002 is obtained is kept in the numeric value analysis result database 230.In addition, the flow figure of pulverized coal flow is kept in the graphic data base 220 through figure generation unit 300.

Fig. 8 is the figure that the movement content to the method for building up of model 700 and unit 600 describes.Utilize Fig. 8 that the step 1020 among Fig. 6,1030 movement content are described.

Fig. 8 (a) is the figure that the method for building up to model 700 describes.In numeric value analysis result database 230, each figure is all preserved when changing air mass flow adjustment gain the numeric value analysis result about CO concentration.In model 700, through utilizing neutral net, batten methods such as (spline) this numeric value analysis result is carried out interpolation, set up the model shown in Fig. 8 (a).Through utilizing this model 700, can infer and the relative CO concentration of air mass flow adjustment yield value.

In addition, in Fig. 8 (a),, also can utilize model 700 to infer air mass flow adjustment gain and NO though the relation of air mass flow adjustment gain and CO concentration is represented _XConcentration, O ₂The project relevant of concentration, unburned part, mercury etc. with boiler characteristics.

Shown in Fig. 8 (a), according to the flow figure of pulverized coal flow, the variation tendency of the relative air mass flow adjustment of CO concentration change in gain is different.In the example of Fig. 8 (a), in 0.0～2.0 scope, air mass flow adjustment gain is big more for figure 1 air mass flow adjustment gain, and CO concentration is low more, and with it relatively, figure 2 air flows adjustment gain is in minimum 0.8 o'clock CO concentration.Like this, because the variation tendency of the relative air mass flow of CO concentration adjustment change in gain is according to the flow figure of pulverized coal flow and difference, so be necessary that the flow figure of each pulverized coal flow is set the air mass flow adjustment that CO concentration is in hour to gain.

In unit 600, for the flow figure study optimal air flow adjustment yield value of each pulverized coal flow.Study is that index is implemented with the mean value 11 that is calculated by evaluation of estimate computing unit 800.In unit 600, the establishing method that evaluation of estimate 11 is in maximum air mass flow adjustment gain is learnt.

Fig. 8 (b) is the figure that the relation of air mass flow adjustment gain and evaluation of estimate when utilizing evaluation of estimate computing unit 800 very low the and higher mode of evaluation of estimate is calculated with CO concentration describes.In addition, in evaluation of estimate computing unit 800, be not to have only CO concentration, also can be based on NO _XConcentration, O ₂The value of concentration, unburned part, mercury etc., the value relevant with boiler characteristics calculate evaluation of estimate.

Shown in Fig. 8 (b), evaluation of estimate is in maximum, is that air mass flow adjustment gain is in 2.0 in figure 1, is in 0.8 in figure 2.In unit 600 owing to learn for maximum mode according to evaluation of estimate 11, so when figure 1 as long as it is 2.0 that air mass flow is adjusted gain setting, as long as and when figure 2, be set at 0.8 and can learn.

Fig. 9 is the form that is kept at the data in learning information database 260 and the graphic data base 220.As shown in Figure 9, the corresponding data of figure and air mass flow adjustment gain are preserved.This is by the method with explanation among Fig. 8 (b), decides the result of air mass flow adjustment gain to each figure.According to this value, in gain setting unit 530, calculated gains alternative 507.In step 1040,, and this signal sent to steam power plant 100 through operation signal generation unit 500 generating run signals 17.Thus, the operation signal in the time of can the evaluation of estimate of being calculated by evaluation of estimate computing unit 800 11 being in maximum sends to steam power plant 100.In step 1050, obtain measuring-signal 1 from steam power plant 100.

Figure 10 is the figure that the details to step 1060 describes.Shown in figure 10, step 1060 is subdivided into step 1061,1062,1063,1064.

In step 1061, from measuring-signal 1, extract the pulverized coal flow measured value information of supplying with to each burner, generate new flow figure.In step 1062, the similarity of the new flow figure that is kept at the existing graphics in the graphic data base 220 and generates in step 1061 is estimated.Similarity is obtained through for example formula (5).Wherein, I _nBe the similarity of n existing graphics and new figure, n (1≤n≤n _Max) be the figure numbering that is kept in the graphic data base 220, n _MaxBe the sum that is kept at the figure in the graphic data base 220, CF _NiBe to being kept at figure in the graphic data base 220 by the value of the input pulverized coal flow of burner i, CF _iIt is measured value by the pulverized coal flow of burner input.

Several 5

$I_n=\underset{i=1}{\overset{i_{\max }}{\mathrm{\&Sigma;}}}{({\mathrm{CF}}_{\mathrm{ni}}-{\mathrm{CF}}_i)}^2...\left(5\right)$

In step 1063,, calculate the error of the CO concentration value of the CO concentration value of inferring and actual measurement through the model of Fig. 8 (a) at first for the maximum figure of the similarity of calculating in the step 1062.

In step 1064, when the error of calculating in the step 1063 when threshold value is following, will not append figure, get into step 1010.At this moment, when entering into step 1020, will revise, so that the CO concentration value of actual measurement and the characteristic of model 700 reach consistent the CO characteristic curve of Fig. 8 (a).

When error when threshold value is above, enter into step 1070, the new figure that generates in the step 1061 is kept in the graphic data base 220.Afterwards, make the condition of air mass flow adjustment change in gain carry out numeric value analysis by step 1000 pair new figure utilization.

Below, utilize the flow process of Figure 10 that the effect of appending figure is described.

In control device 200, suppose the flow figure of pulverized coal flow in advance, and this flow figure actual figure value is resolved.But because the number of burner is more, and pulverized coal flow is successive value, so the flow figure exists with a plurality of forms.Therefore, be difficult to whole flow figure actual figure values is resolved.Narrating when explaining, set the flow figure based on the information that is kept in the equipment information database 270 like Fig. 7.Steam power plant 100 might generate the different new flow figure of flow figure that sets with this moment when operation.New flow figure is compared with existing flow figure, when boiler characteristics is identical with respect to the variation of air mass flow adjustment change in gain, with the corresponding new flow figure of the learning outcome of existing flow figure.In addition, at new flow figure and existing flow graphics feature not simultaneously each other, resolve, carry out correspondence thus through carrying out with new flow figure value corresponding.The result is if the experience of the new flow figure of accumulation can be set the optimum air flow adjustment gain relative with this flow figure.Its result can reduce the CO that is emitted by steam power plant 100.

In addition, through utilizing control device 200, the air mass flow of putting in the boiler is minimized.Its result can make fan power minimize, and can cut down the spent electric power of boiler.But also can reduce the size of boiler.

Figure 11 is the time dependent figure of explanation measured value.

Exist when measuring pulverized coal flow to measure and postpone with the pulverized coal flow measuring appliance.And as shown in Figure 3, coal dust is through after disposing the position of pulverized coal flow measuring appliance, supplies to burner, and imports in the stove.

Time till measuring time delay and beginning to the input stove through the pulverized coal flow measuring appliance from coal dust, measured value was consistent with the pulverized coal flow of putting in the stove when consistent.But, measure time delay in short-term at the time ratio that begins through the pulverized coal flow measuring appliance from coal dust till drop into the stove, shown in Figure 11 (a), the pulverized coal flow from put into stove, measured value will be later than the pulverized coal flow value.On the contrary, when time ratio measurement length time delay till coal dust begins to dropping into stove through the pulverized coal flow measuring appliance, shown in Figure 11 (b), the pulverized coal flow from put into stove, the pulverized coal flow value will be later than measured value.

With the length of the pipe arrangement that connects pulverized coal flow measuring appliance 155 and burner 102 flow velocity, can calculate coal dust time till the stove from beginning to putting into thus through pulverized coal flow measuring appliance 155 divided by coal dust.In addition, the measurement of pulverized coal flow measuring appliance 155 can be grasped time delay in advance.The length of the pipe arrangement through utilize connecting pulverized coal flow measuring appliance 155 and burner 102, the flow velocity of coal dust, measure the information of time delay, can proofread and correct measured value, so that measured value is consistent with the pulverized coal flow value in putting into stove.Though in the operation signal generation unit 500 of Fig. 4, illustrate; Measuring-signal data 3 are handled the measured value correcting unit of the function that the pulverized coal flow value in making measured value and putting into stove is consistent but operation signal generation unit 500 is had as above-mentioned.

In addition, also there is the operation air damper to make time delay up to the air mass flow method consistent with the value of hope.Can connect the pipe arrangement of pulverized coal flow measuring appliance and burner through lengthening, restrain the decline of the control performance of following this time delay.Also can consider above-mentioned pulverized coal flow measuring appliance 155 is configured in the steam power plant 100.

Figure 11 (c) be generating output when changing output and pulverized coal flow through the time variation.Shown in Figure 11 (c), the total amount of pulverized coal flow alters a great deal when output changes.Under these circumstances, as Figure 11 (a) is (b) said, might produce deviation at measured value and the pulverized coal flow value of putting into stove.In order to proofread and correct this influence, also can make operation signal generation unit 500 have the function of proofreading and correct the pulverized coal flow measured value based on the pulverized coal flow that supplies to grinding mill, power.And, for measured fuel flow rate value is proofreaied and correct, can utilize at least one information in measurement time delay of flow velocity, measuring appliance of length, the coal dust of the pipe arrangement that connects measuring appliance and burner.And can make graphical unit 300 have following function; Promptly estimate the pulverized coal flow of supplying with to grinding mill, the information of power; And the relation between the pulverized coal flow measured value generates information, and the relation of the flow figure of pulverized coal flow of the pulverized coal flow supplied with to grinding mill and power.

Figure 12, the 13rd is presented at the embodiment of the picture on the image display device 950.

Shown in figure 12; Through utilizing control device 200; The command value of the air mass flow that can calculate with the pulverized coal flow that is measured by pulverized coal flow measuring appliance 155A, 155B, 155C, 155D, 155E with at operation signal generation unit 500 is presented on the same picture.Thus, running status that can easy master steam power plant 100.

In addition, through each burner demonstration pulverized coal flow and air mass flow are grasped relative value.

Figure 13 carries out embodiment illustrated to the figure that utilizes graphical unit 300 to generate.

Shown in figure 13, can be on picture display graphics shape and characteristics/properties.Thus, can confirm the figure that generated by graphical unit 300 easily.

Below, with reference to accompanying drawing the control device of the steam power plant of another embodiment of the present invention and the control method of steam power plant are described.

(embodiment 2)

Figure 14 is the block diagram that the integral body of control device of the steam power plant of expression one embodiment of the invention constitutes.

In Figure 14, the control device that has possessed the steam power plant 2100 of the boiler that acts as a fuel with coal is controlled by control device 2200.Control device 2200 is respectively arranged with as arithmetic unit and arrives the zone and infer unit 2300, numeric value analysis performance element 2400, operation signal generation unit 2500, unit 2600, model 2700, evaluation of estimate computing unit 2800 and constitute.

In this control device 2200, be respectively arranged with measuring-signal database 2210, arrive regional database 2220, numeric value analysis result database 2230, operation signal database 2240, control logic database 2250 and learning information database 2260 as database.

In addition, in control device 2200, as with the interface of outside, dispose outer input interface 2201 and outside output interface 2202.

In this external this control device 2200; Through outer input interface 2201; From the various quantity of states of steam power plant 2100 with steam power plant, the measuring-signal 2001 of for example measuring oxygen concentration in the burning gases of boiler export or carbonomonoxide concentration is taken into this control device 2200, and; Through outside output interface 2202, send the operation signal 2018 of the air mass flow in the burner of controlling boiler for example and air port from the above-mentioned steam power plants 2100 of 2200 pairs of control device.

Be taken into the various quantity of states of the steam power plant 2100 of control device 2200 from steam power plant 2100; For example measure the measuring-signal 2001 of oxygen concentration in the burning gases of boiler export or carbonomonoxide concentration; After process outer input interface 2201, being kept at the database that is arranged at control device 2200 as measuring-signal 2002 is in the measuring-signal database 2210.

The operation signal 2017 that in the operation signal generation unit 2500 of the arithmetic unit that control device 2200 is provided with, generates; Send to outside output interface 2202 by operation signal generation unit 2500, and to be kept at the database that is arranged at control device 2200 be in the operation signal database 2240.

In addition in the numeric value analysis performance element 2400 of the arithmetic unit that control device 2200 is provided with; Set up the detailed physical model that steam power plant 2100 is carried out high-precision analog in inside; And constitute and utilize this detailed model, carrying out with steam power plant 2100 is the detailed of object and more accurate numerical parsing.

That is,, utilize detailed physical model to calculate, simulate and predict the operation characteristic of steam power plant 2100 accurately through carrying out following high-precision analog according to the operation of this numeric value analysis performance element 2400.It is that the structure in structure, the burner 2102 that is arranged at boiler 2101 and the air port 2103 of the boiler 2101 that constitutes steam power plant 2100 and fuel flow rate, the air mass flow supplied with to burner 2102 and air port 2103 respectively are set at boundary condition that this high-precision analog calculates.

Detailed and the more accurate numerical resolving information 2006 that then utilizes the numeric value analysis performance element 2400 of detailed physical model to be obtained through execution sends and is saved in the numeric value analysis result database 2230 from numeric value analysis performance element 2400.

Infer in the unit 2300 in the arrival zone of the arithmetic unit that is arranged at control device 2200; Utilization is kept at the measuring-signal data 2004 in the measuring-signal database 2210 and is kept at the numeric value analysis data 2007 in the numeric value analysis result database 2230, infers the air that supplies to boiler 2101 from burner 2102 and air port 2103 through computing and arrives the zone the boiler export stream.

The air that supplies to boiler 2101 from burner 2102 and air port 2103 of inferring through the computing that utilizes this arrival zone to infer unit 2300 arrives the arrival area information 2005 the boiler export stream, is kept at the database that is arranged at control device 2220, promptly arrives in the regional database 2220.

Be arranged at the unit 2600 of the arithmetic unit of control device 2200; The statistical model 2700 that constitutes with arithmetic unit is the method for operating of uncertain plant learning steam power plant 2100; Model 2700 constitutes and is utilized in the inner simple statistical model of setting up, and simulates the control characteristic of steam power plant 2100 at short notice.

Just; With will send to steam power plant 2100 from control device 2200 at the operation signal 2018 that being used for of generating of control device 2200 controlled steam power plant 2100; And it is identical from steam power plant 2100 receptions as the situation of the control result's of this steam power plant 2100 measuring-signal 2001 with control device 2200; Constitute as follows: will be in unit 2600 study of control device 2200 and the model input 2009 that generates; Send to model 2700 from unit 2600, it is model output 2010 in the simple analog result of the control characteristic of the steam power plant 2100 of these model 2700 simulations that unit 2600 receives conduct from model 2700.

In model 2700 as arithmetic unit; Utilize simple physical model; Simulate the control characteristic of steam power plant 2100 at short notice; And this analog result outputed in the unit 2600 as model output 2010; Wherein this simple physical model is to utilize to be kept at the numeric value analysis data 2008 in the numeric value analysis result database 2230 and to be kept at the measuring-signal data 2004 in the measuring-signal database 2210, is based upon the inside of model 2700 based on the model input 2009 of being sent by unit 2600.

This model 2700 utilizes the statistical model of neutral net for example etc. to set up; Through operational model 2700; Before steam power plant 2100 operations; Only utilize the numeric value analysis data 2008 be kept in the numeric value analysis result database 2230 to come the above-mentioned model output 2010 of analog computation, after steam power plant 2100 operations, the measuring-signal 2004 that dual-purpose is kept in the measuring-signal database 2210 of the quantity of state of measuring steam power plant 2100 comes analog computation model output 2008.

Like this in the analog result of the steam power plant 2100 that has used model 2700; The analog result of the steam power plant 2100 of the detailed physical model of using when having used computing numerical value resolution executing 2400 is different; Under the control characteristic condition of different of steam power plant 2100; Through the measuring-signal data 2004 in the measuring-signal database 2210 of paying attention to being kept at the quantity of state of measuring steam power plant 2100; By above-mentioned model 2700 analog computation models output 2010, thereby can make the control characteristic of the characteristic of model 2700 near steam power plant 2100.

In unit 2600, to learning to the generation method of the model input 2009 of above-mentioned model 2700 inputs, so that reach desired value by the model output 2010 of model 2700 analog computations from this unit 2600.

As the index of using through the generation method of this unit 2600 learning models input 2009; Can utilize evaluation of estimate 2011 to come to learn through above-mentioned unit 2600, this evaluation of estimate 2011 is to be calculated by the evaluation of estimate computing unit 2800 as arithmetic element that is arranged on control device 2200.

In this evaluation of estimate computing unit 2800; When the model output 2010 that is gone out by model 2700 analog computations is the state of expectation; The value that is input to the evaluation of estimate 2011 in the above-mentioned unit 2600 is set at bigger, and along with the state that departs from expectation, the value of evaluation of estimate 2011 is set at less.

In addition, in this evaluation of estimate computing unit 2800, also can utilize to be kept at the arrival area data 2016 that the database that is arranged at control device 2200 promptly arrives in the regional database 2220 and to calculate above-mentioned evaluation of estimate 2011.

Above-mentioned unit 2600 is set up through the various optimization gimmicks such as calculating gimmick of using intensified learning, evolution.

In addition, in this unit 2600, the evaluation of estimate 2011 of being calculated and be imported into this unit 2600 by evaluation of estimate computing unit 2800 is reached maximum method of operating learn.

The learning information data 2013 of the constraints of in the study of unit 2600, using, model export target value etc., being kept at the database that is arranged at control device 2200 is in the learning information database 2260.

In addition, as being learning information data 2012, exported to learning information database 2260 from unit 2600, and be kept in this learning information database 2260 by the result after unit 2600 study.

At the arithmetic unit that is arranged at control device 2200 is in the operation signal generation unit 2500; Obtain following data as required, and utilize these information to generate the operation signal 2017 carbonomonoxide concentration, that for example control to the air mass flow of boiler supply of discharging by steam power plant 2100 being used for reducing from the burner 2102 and the air port 2103 of boiler 2101.These data are: be kept at measuring-signal data 2003 in the measuring-signal database 2210, be kept at the arrival area data 2016 that arrives in the regional database 2220, be kept at the control logic data 2015 in the control logic database 2250 and be kept at the learning information data 2014 in the learning information database 2260.

In addition, the operation signal 2017 that has generated at aforesaid operations signal generation unit 2500 is constituted as, and from control device 2200, sees off through outside output interface 2202 as the operation signal 2018 to steam power plant 2100.

In addition, shown in figure 14, near control device 2200, be provided with the external input device 2900, maintenance tool 2910 and the image display apparatus 2950 that constitute by keyboard 2901 and mouse 2902.

In addition; The operator of steam power plant 2100; Through utilizing the external input device 2900 that constitutes by keyboard 2901 and mouse 2902 to generate maintenance tool input signal 2051; And this maintenance tool input signal 2051 is input in the maintenance tool 2910, can the information that be configured in the various databases in the control device 2200 be presented on the image display device 2950 thus.

Above-mentioned maintenance tool 2910 is made up of outer input interface 2920, data transmit-receive portion 2930 and outside output interface 2940.

Externally the maintenance tool input signal 2051 that generated of input unit 2900 imports in the maintenance tool 2910 through outer input interface 2920.

In the data transmit-receive portion 2930 of maintenance tool 2910, constitute information according to maintenance tool input signal 2052, from the various databases that are configured in control device 2200, obtain database information 2050.

In the data transmit-receive handling part 2930 of maintenance tool 2910, deal with data library information 2050 resulting maintenance tool output signals 2053 are sent to outside output interface 2940.

Outside output interface 2940 will send to image display apparatus 2950 based on the output signal 2054 of this maintenance tool output signal 2053, and shows through this image display device 2950.

In addition; In the control device 2200 of the steam power plant of the foregoing description; Though be arranged on measuring-signal database 2210, arrival regional database 2220, numeric value analysis result database 2230, operation signal database 2240, control logic database 2250 and the learning information database 2260 in the composition data storehouse of above-mentioned control device 2200; Infer the inside that unit 2300, numeric value analysis performance element 2400, unit 2600, model 2700 and evaluation of estimate computing unit 2800 all are configured in above-mentioned control device 2200 with the arrival zone that constitutes arithmetic unit, but also can with its all, or wherein a part be configured in the outside of control device 2200.

Figure 15 represent to use coal as the control object of the control device of steam power plant among the embodiment shown in Figure 14, be that the summary that the boiler of fuel is in the steam power plant constitutes.

At first utilize Figure 15 (a) that the structure of the generating of the steam power plant 2100 that possessed boiler is described.

In Figure 15 (a); The coal that acts as a fuel is ground by grinding mill 2110 and the primary air used as coal dust and coal conveyance; And the auxiliary air of burning adjustment usefulness is together through being arranged on the burner 2102 of boiler 2101 and putting in the boiler 2101, and the coal that acts as a fuel burns in the stove inside of boiler 2101.

Coal that acts as a fuel and primary air are by pipe arrangement 2134, and auxiliary air is directed in the burner 2102 by pipe arrangement 2141.

In addition, the make-up gas of two grate firings burning usefulness is put in the boiler 2101 through the additional air port 2103 that is arranged on boiler 2101.This make-up gas is to import to via pipe arrangement 2142 to replenish air port 2103.

Make the stove internal-combustion of the coal that acts as a fuel at boiler 2101; The burning gases of the high temperature that is produced; In the stove of boiler 2101, flow to the downstream along the path shown in the arrow; Carry out heat exchange through the heat exchanger 2106 that is configured in boiler 2101 again, become burning and gas-exhausting afterwards and discharge from boiler 2101, and run down into the air outside heater 2104 that is arranged on boiler 2101.

Passed through the burning and gas-exhausting of air heater 2104, removed the harmful substance that is included in the burning and gas-exhausting by not shown exhaust gas treatment device afterwards, then by smoke stack emission in atmosphere.

The feedwater of circulation in boiler 2101; From the not shown condenser that is arranged at steam turbine 2108, import in the boiler 2101 by feed pump 2105; And in the heat exchanger in the stove that is arranged at boiler 2,101 2106; By the burning gases heating that the stove inside along boiler 2101 flows down, become the steam of HTHP.

In addition, in the present embodiment, illustrate, also can dispose a plurality of heat exchangers though the number of heat exchanger 2106 is made as one.

The steam of the HTHP that in heat exchanger 2106, has generated; Import in the steam turbine 2108 by turbine type burner valve (タ one PVC Application ガ バ Na man's cap used in ancient times) 2107; And the energy drives steam turbine 2108 that utilizes steam to have, make generator 2109 rotations and the generating that have linked with this steam turbine 2108.

Below, the path of the burner from the stove that is arranged on boiler 2,101 2102 being put into the make-up gas in the path of primary air and auxiliary air in the stove of boiler 2101 and the stove that boiler 2101 is put in the additional air port 2103 from the stove that is arranged on boiler 2101 describes.

Primary air imports pipe arrangement 2130 by fan 2120; And the pipe arrangement 2132 that is diverted to the inside of process air heater 2104 is halfway crossed in the pipe arrangement 2131 of air heater 2104 with bypass; The primary air that in this pipe arrangement 2132 and pipe arrangement 2131, flows down collaborates at pipe arrangement 2133 places once more, thereby imports in the grinding mill 2110.

Through the air of air heater 2104, receive the burning gases of discharging from the stove of boiler 2101 and be heated.

Utilize this primary air, conveyance is in burner 2102 by coal (coal dust) the process pipe arrangement 2133 of grinding mill 2110 generations.

Auxiliary air and make-up gas import pipe arrangement 2140 from fan 2121; In pipe arrangement 2140, flow down and after being heated through the inside of air heater 2104; Be diverted in the pipe arrangement 2142 that pipe arrangement 2141 that auxiliary air uses and make-up gas use in the downstream of pipe arrangement 2140, constitute in the burner 2102 that imports to the stove that is arranged at boiler 2101 respectively and the additional air port 2103.

The control device 2200 of steam power plant 2100 that has possessed the boiler of present embodiment is in order to reduce the NO in the boiler exhaust _XWith CO concentration, have putting into the air capacity the boiler 2101 from burner 2102 and putting into the function that the air capacity the boiler 2101 is adjusted from replenishing air port 2103.

In steam power plant 2100, dispose the various measuring appliances that the running status to this steam power plant 2100 detects, by the measuring-signal of the equipment of obtaining in these measuring appliances, be sent in the control device 2200 as measuring-signal 1.

The various measuring appliances that detect as the running status to steam power plant 2100 for example, illustrate flow measuring probe 2150, temperature meter 2151, pressometer 2152, generating output checker 2153 respectively and to O in Figure 15 ₂The measurement of concetration device 2154 that concentration and/or CO concentration are measured.

Flow measuring probe 2150 is measured the flow that supplies to the feedwater the boiler 2101 from feed pump 2105.In addition; Temperature meter 2151 and pressometer 2152; In the heat exchanger that is equipped on boiler 2,101 2106, the steam to producing respectively by carrying out heat exchange with the burning gases that in this boiler 2101, flow down, the temperature and the pressure that are fed into the steam in the steam turbine 2108 are measured.

Steam driving steam turbine 2108 rotations that generated at above-mentioned heat exchanger 2106, the rotation through this steam turbine 2108 impels generator 2109 generatings, and its electric weight can be measured through generating output checker 2153.

In addition, relate to composition (CO, the NO that in the burning gases that boiler 2101 flows down, is contained _XDeng) information of concentration, can be through the O downstream, that is arranged on boiler 2101 to being provided with in the boiler export stream ₂The measurement of concetration device 2154 that concentration and/or CO concentration are measured is measured.

In addition, in general, except that shown in Figure 15, also dispose a plurality of measuring appliances in steam power plant 2100, but omitted diagram here.

Figure 15 (b) is the air heater 2104 in the expression downstream that is arranged on the boiler 2101 that constitutes steam power plant 2100 and the partial enlarged drawing that is provided in the pipe arrangement of this air heater 2104.

Shown in Figure 15 (b); The auxiliary air of having shunted in the downstream of the pipe arrangement 2140 of the inside that is equipped on air heater 2104 disposes air damper 2162,2163 with pipe arrangement 2141 and make-up gas respectively in pipe arrangement 2142, and at the pipe arrangement 2132 of the inside that is equipped on air heater 2104, reach bypass and cross in the pipe arrangement 2131 of air heater 2104 and to dispose air damper 2161,2160 respectively.

So through operate these air dampers 2160～2163 change in pipe arrangement 2131,2132,2141,2142 air the area of process, and the air mass flow through this pipe arrangement 2131,2132,2141,2142 individually adjusted.

In addition; Utilization is also outputed to the operation signal 2018 in this steam power plant 2100 by control device 2200 generations of control steam power plant 2100, comes the machine of feed pump 2105, grinding mill 2110, air damper 2160,2161,2162,2163 etc. is operated.

In addition in the control device of the steam power plant of present embodiment; Machine for the quantity of state of the steam power plant of regulating feed pump 2105, grinding mill 2110, air damper 2160,2161,2162,2163 etc.; They are called the operating side, and will make it to operate required command signal and be called operation signal.

In addition; When the fuel that also burner that is arranged on boiler 2,101 2102 and additional air port 2103 has been added the air that can use etc. in burning or coal dust etc. drops in the boiler 2101; Come to move up and down the function of its angle that spues, these command signals of regulating the angle that spues in burner 2102 and additional air port 2103 also can be included among the aforesaid operations signal 2108.

Figure 16 be in the control device of the steam power plant 2100 of the boiler that possesses embodiment shown in Figure 14, the flow chart of control sequence of expression control device 2200.

In Figure 16, control device 2200 combination step 3000,3010,3020,3030,3040,3050, carry out control to steam power plant 2100.

Step 3030a and step 3030b and step 3050a and step 3050b are respectively identical actions in addition.Below will step separately be described.

In the step 3000 that at first initial numeric value analysis is on stream carried out, make numeric value analysis unit 2400 actions of the detailed physical model of having set up the control device 2200 that is arranged on Figure 14, carry out the high accuracy numeric value analysis of steam power plant 2100.

The high accuracy numeric value analysis result 2006 of the steam power plant 2100 of the detailed physical model action gained of this numeric value analysis unit 2400 is kept in the numeric value analysis result database 2230.

Then, study on stream has or not in the determining step 3010, and combination is arranged on unit 2600, model 2700 and the evaluation of estimate computing unit 2800 in the control device 2200 of Figure 14, comes to judge whether implementing performed study.

And have or not in the step 3010 of judgement in this study; When being judged as execution study; Have or not the step 3010 of judgement to get into the path of " being " from study; And then get into model establishment step 3020 subsequently, and, have or not the path of determining step 3010 entering " denying " from study when being judged as when not implementing to learn.

Wherein, for being judged as when implementing study and being judged as contents processing when not implementing to learn (arrive step 3030 is inferred in the zone, operation signal generates step 3050), added the processing of model establishment step 3020 and method of operating learning procedure 3040.

Follow in the step 3020 of model foundation on stream; The various quantity of states to steam power plant 2100 based in numeric value analysis result database 2230 that is kept at the control device 2200 that is arranged at Figure 14 and the measuring-signal database 2210 carry out data measured; Foundation is arranged at the model 2700 of the control device 2200 of Figure 14, and utilizes the characteristic of these model 2700 simulation trial steam power plants of having set up 2100.

In addition, after steam power plant 2100 had just brought into operation, in measuring-signal database 2210, storage had not been carried out data measured to the various quantity of states of steam power plant 2100.

Therefore under this situation, utilize and be kept at the numeric value analysis data 2008 in the numeric value analysis result database 2230, set up the simple physical model of model 2700 and simulate the characteristic of steam power plant 2100.

After steam power plant 2100 operation beginnings; Through after blink; The various quantity of states that obtain steam power plant 2100 have carried out measurement data measured 2001; And when being stored in the data of the various quantity of states of above-mentioned steam power plant 2100 in the measuring-signal database 2210, through the above-mentioned model 2700 of calculation process correction, so that the characteristic of model 2700 and steam power plant 2100 is reached an agreement.

As stated, in the control device 2200, be the function of the numeric value analysis of object with steam power plant 2100 through making model 2700 possess execution, can predict CO concentration, NO in the discharge gas of from the boiler of steam power plant 2100, discharging easily in the short time _XConcentration is perhaps in gas concentration distribution of boiler export etc.

Follow in the step 3030 of inferring in arrival zone on stream; Make the arrival zone in the control device 2200 that is arranged on Figure 14 infer unit 2300 actions; Infer the air mass flow of from the burner 2102 of the boiler 2101 that is configured in steam power plant 2100 and air port 2103, supplying with, arrive which part of boiler export stream.

Follow in the step 3040 of method of operating study on stream, make up unit 2600, model 2700 and evaluation of estimate computing unit 2800 in the control device 200 that is separately positioned on Figure 14, the method for operating of steam power plant 2100 is learnt.

For the learning algorithm that is arranged on the unit 2600 in the control device 2200; Though using Figure 20 back narrates; But the learning outcome that the unit 2600 that obtains through the step 3040 that this method of operating is learnt is produced is kept in the learning information database 2260.

Then; In the step 3050 that operation signal on stream generates; Make operation signal generation unit 2500 action in the control device 2200 that is arranged on Figure 14, generate from control device 2200 to the command signal of the operating side of steam power plant 2100 output, be operation signal 2017.

Figure 17 is the control block diagram that expression is arranged on the formation of the operation signal generation unit 2500 in the control device 2200 of the Figure 14 that generates this operation signal 2017.

Shown in figure 17; Operation signal generation unit 2500 constitutes to be possessed: reference signal generation portion 2510; Its utilization is kept at the control logic data 2015 in the control logic database 2250 and is kept at the measuring-signal data 2003 in the measuring-signal database 2210, generates reference signal 2030; The first correction signal generation portion 2520, its utilization is kept at and arrives the arrival area data 2016 in the regional database 2220 and be kept at the measuring-signal data 2003 in the measuring-signal database 2210, generates first correction signal 2033; The second correction signal generation portion 2530, its utilization is kept at the learning information data 2014 in the learning information database 2260 and is kept at the measuring-signal data 2003 in the measuring-signal database 2210, generates second correction signal 2036; Null value maker 2540,2550, it generates the signal of null value; And switch 2560,2570, it selects a signal in two signals being imported, and the signal that will select is as the output signal.

In addition; Constitute the said reference signal generation portion 2510 of the operation signal generation unit 2500 of Figure 17; Utilization is kept at the control logic data 2015 in the control logic database 2250 and is kept at the measuring-signal data 2003 in the measuring-signal database 2210, generates reference signal 2030.

In this reference signal generation portion 2510, the air mass flow that decision is supplied with from the burner that is arranged on boiler 2,101 2102 and air port 2103 is so that the oxygen (O in the burning gases at boiler export stream place ₂) concentration reaches consistent with predetermined value.

Constitute the above-mentioned first correction signal generation portion 2520 of the operation signal generation unit 2500 of Figure 17; Utilization is kept at and arrives the arrival area data 2016 in the regional database 2220 and be kept at the measuring-signal data 2003 in the measuring-signal database 2210, generates first correction signal 2033.

In this first correction signal generation portion 2520; Constitute the control loop of air mass flow that decision is arranged on burner 2102 and the air port 2103 of boiler 2101; Air mass flow with the low zone of the oxygen concentration of increase arrival boiler export stream perhaps increases the air mass flow that arrives the high zone of carbonomonoxide concentration.

Being kept at the information that arrives in the regional database 2220 and the generation method of first correction signal 2033 utilizes Figure 18 and Figure 19 back to narrate.

Constitute the above-mentioned second correction signal generation portion 2530 of the operation signal generation unit 2500 of Figure 17; Utilization is kept at the learning information data 2014 in the learning information database 2260 and is kept at the measuring-signal data 2003 in the measuring-signal database 2210, generates second correction signal 2036.

The generation method that is kept at information and second correction signal 2036 in the learning information database is utilized the narration of Figure 20 back.

Constitute the above-mentioned switch 2560,2570 of the operation signal generation unit 2500 of Figure 17, in two signals being imported, select a signal, and with selected signal as the output signal.

Therefore, select and the signal 2034 of output by switch 2560, consistent with one of them of first correction signal 2033 that generates in the first correction signal generation portion 2520 and the signal value of zero 2032 that in null value maker 2540, generates.

In addition, select and the signal 2037 of output by switch 2570, consistent with one of them of second correction signal 2036 that generates in the second correction signal generation portion 2530 and the signal value of zero 2035 that in null value maker 2550, generates.

Shown in figure 17; The operation signal 2017 of output from operation signal generation unit 2500; Be reference signal 2030, in addition select and the signal 2034 exported and select and the signal 2037 exported and the signal that generated by switch 2570 by switch 2560 to having generated and exported in reference signal generation portion 2510.

Therefore, this operation signal 2017 possibly be following value: be equal values with the reference signal 2030 that has generated and exported in reference signal generation portion 2510; It is the aggregate value of the reference signal 2030 and first correction signal 2033 that generated in the first correction signal generation portion 2520; It is the aggregate value of the reference signal 2030 and second correction signal 2036 that generated in the second correction signal generation portion 2530; Or the aggregate value of reference signal 2030 and first correction signal 2033 and second correction signal 2036.

In addition, be imported in the signal of switch 2560,2570, determine the data of which signal, be kept in the control logic database 2250 of the control device 2200 that is arranged at Figure 14 as the output signal.

Figure 18 infers the function of unit 2300 and arrives the figure that the information in the regional database 2220 describes to being kept at the arrival zone in the control device 2200 that is arranged on present embodiment shown in Figure 14.And numeric value analysis performance element 2400 shown in Figure 14 constitutes and can resolve flowing of fluid in the boiler 2101 through three-D.

The result who infers the parsing that is produced with numeric value analysis performance element 2400 that unit 2300 is produced is inferred in the arrival zone that is arranged in the control device 2200 of Figure 14; Shown in Figure 18 (a), can calculate the air mass flow that from the air port that is arranged on boiler 2,101 2103 and burner 2102, drops into; Arrive 12 parts of which zones of cutting apart in the boiler export stream, then the numeric value analysis data 2006 of this result of calculation are stored in the numeric value analysis result database 2230.

Infer in the unit 2300 in the arrival zone; Can be kept at the numeric value analysis data in the numeric value analysis result database 2230 through reference; The air of boiler is put in grasp from the air port that is arranged on boiler 2,101 2103 and burner 2102, arrive 12 parts of which zones of cutting apart the boiler export stream.

In addition; In also can measured value information according to the gas concentration of the operation signal of air mass flow and boiler export; Obtain the air that from the air port that is arranged on boiler 2,101 2103 and burner 2102, drops into, arrive 12 parts of which zones of cutting apart the boiler export stream.

Utilize Figure 18 (b) to describe, Figure 18 (b) is the air mass flow from the A of boiler 2101 to each air port of E, with oxygen (O in 12 parts of each cut zone of cutting apart the boiler export stream ₂) relation of measured value of concentration.

In the present embodiment shown in Figure 18 (b), suppose the stream of boiler export is divided into 3 parts being divided on the width on 4 parts, depth direction, and in these 12 parts each zones of having cut apart, dispose the measured oxygen concentration device respectively as measurement of concetration device 2154.

Therefore; In the present embodiment; In the zone of 12 parts of streams of having cut apart boiler export, disposed 12 measured oxygen concentration devices, can certainly change boiler export stream cut apart number, perhaps be increased in the number of the measured oxygen concentration device that disposes in each zone of having cut apart.

In Figure 18 (b), the air mass flow that drops among the A～E by the air port 2103 of boiler 2101 at first oxygen (O do not occur for fixing in the stream of boiler export ₂) dense part.But, when increasing the air mass flow of from the 2103A of air port, supplying with, increase at the measured value of the right regions oxygen concentration of 12 parts of streams of cutting apart boiler export when the balance that changes air mass flow.

Can judge from the air port air that 2103 A has supplied with from this result, arrive 12 parts of right regions of cutting apart the boiler export stream.

Like this; The numeric value analysis performance element 2400 of control device 2200; Having can be from the operation signal of air mass flow and the gas concentration measurement information the boiler export stream, obtains the air of from the air port that is arranged on boiler 2101 and burner 2102, putting into boiler and arrives 12 parts of functions of cutting apart which zone behind the boiler export stream.

In addition, in Figure 18 (b), to dispose oxygen (O at boiler export ₂) the measurement of concetration device for example is illustrated, replaces measured oxygen concentration device but also can dispose carbon monoxide (CO) measurement of concetration device as measurement of concetration device 2154 as the measurement of concetration device 2154 that the gas concentration in the exhaust is measured.At this moment, judging the zone of the stream of the boiler export of from the A in air port 2103, supplying with that air mass flow arrived, is the part that CO concentration reduces when increasing the air mass flow of air port 2103A.

Figure 19 is the figure that the generation method to the operation signal of the operation signal generation unit 2500 in the control device 2200 of the present embodiment that is arranged on Figure 14 describes.In operation signal generation unit 2500; Shown in the control device 2200 of Figure 14; Utilization is kept at the measuring-signal data 2003 that arrive the arrival area data 2016 in the regional database 2220 and be kept at the quantity of state of the steam power plant 2100 in the measuring-signal database 2210 and carries out computing, generating run signal 2017.

Shown in Figure 19 (a); Expression with Figure 18 explanation, the measurement of concetration device 2154 that the CO concentration in the exhaust is measured is set in the stream of boiler export; Measure the situation of the concentration measurement of carbon monoxide (CO); In aforesaid operations signal generation unit 2500, in 12 parts of each cut zone of cutting apart the boiler export stream, the arrival amount of the air mass flow that will from the burner 2102 that is arranged at boiler 2101 or air port 2103, drop into; The CO concentration measurement of measuring with the carbonomonoxide concentration measuring appliance that is arranged on each cut zone as measurement of concetration device 2154 usefulness multiplies each other, and calculates their summation.This summation is defined as first influence value.

In aforesaid operations signal generation unit 2500; With the mode of the air mass flow that increases proportion is bigger in first influence value burner 2102 or air port 2103, utilize first influence value and through 2500 computings of operation signal generation unit and generate command signal to steam power plant 2100, be operation signal 2017.

At Figure 19 (a), in the air port 2103 of expression from A to E, the situation that first influence value of air port B is bigger than first influence value of air port A.At this moment, generate the operation signal 2017 that makes the air mass flow increase that B drops into from the air port through operation signal generation unit 2500.

Utilize said method to calculate and generate command signal through operation signal generation unit 2500, be operation signal 2017 (air mass flow command value) to steam power plant 2100, can increase thus be used for that air supply arrives the high zone of carbonomonoxide concentration be arranged on the burner 2102 in the boiler 2101, the air mass flow in air port 2103.

Thus, in the boiler 2101 of the steam power plant 2100 of present embodiment, arrive the high zone of carbonomonoxide concentration through air supply, can make carbon monoxide (CO) and the air supplied with in oxygen (O ₂) react oxidizing carbon monoxide, therefore, can reduce the concentration of carbon monoxide in the exhaust of from boiler 2101 outlets, discharging.

In addition, shown in Figure 19 (b), expression with Figure 18 explanation, be provided with oxygen (O in the burning gases of the flow path of boiler export ₂) the measurement of concetration device 2154 measured of concentration measures the situation of the measured value of oxygen concentration; In aforesaid operations signal generation unit 2500; For 12 parts of each zones of cutting apart the boiler export stream; The arrival amount of the air mass flow that will from the burner 2102 that is arranged at boiler 2101 or air port 2103, drop into and multiply each other as the measured oxygen concentration value that measurement of concetration device 2154 is arranged on each cut zone is calculated their summation.This summation is defined as second influence value.

At this moment; In aforesaid operations signal generation unit 2500; With the mode of the air mass flow that reduces proportion is bigger in second influence value burner 2102 or air port 2103, through 2500 computings of operation signal generation unit and generate command signal to steam power plant 2100, be operation signal 2017.

For the burning gases that flow at boiler export, in 12 parts of each cut zone of cutting apart the boiler export stream, (O produces oxygen ₂) under the situation of concentration upper zone, might be not enough at other regional oxygen, this moment, the concentration at oxygen other regional carbon monoxide (CO) uprised.

In addition, under the situation of the concentration that produces oxygen upper zone, owing to might drop into excessive air, so the excessive air that has dropped into through reducing has the power consumption advantages that can reduce air fed fan power etc.Therefore, high regional the least possible of oxygen concentration is good.

Therefore in aforesaid operations signal generation unit 2500, utilize second influence value through 2500 computings of operation signal generation unit and derive command signal to steam power plant 2100, be operation signal 2017, thus can the control air flow with at O ₂Do not produce skew on the CONCENTRATION DISTRIBUTION, in the boiler 2101 of the steam power plant 2100 of present embodiment, not only can restrain the generation of the carbon monoxide in the burning gases of from boiler 2101, discharging like this, also can obtain to reduce the energy-saving effect of fan power.

Figure 20 is to the arithmetic unit of the control device that is arranged at Figure 14 2200 of present embodiment, is the flow chart that the order of operation of learning algorithm in the unit 2600 describes, is the figure that the action of method of operating learning procedure 3040 in the flow chart shown in Figure 16 is represented in detail.

In addition, Figure 20 puts down in writing in use intensified learning in above-mentioned learning algorithm as an example, but also can in the study computing of present embodiment, use various Techniques of Optimum such as genetic algorithm, annealing method.

Shown in the flow process of Figure 20, step 3100,3110,3120,3130,3140,3150,3160,3170,3180,3190 each step groups are moved altogether by the learning algorithm that is arranged on unit 2600 computings in the control device 2200.

At first in the step 3100 of the initial value that any setting model is imported, the initial value of model input 2009 is at random set.

Then in the step 3110 of computation model output; The model input 2009 that will in step 3110, set is input in the model 2700 of the control device 2200 that is arranged at Figure 14, and the simulation trial that utilizes this model 2700 to carry out steam power plant 2100 obtains model output 2010.

Then; In initial value determination step 3120; Model output 2010 and model export target value that the step of relatively exporting through computation model 3110 obtains; If model output 2010 reaches model export target value, promptly turn back to the step 3100 of computation model output, do not reach the step 3130 that gets into decision model input amplitude of variation under the situation of model export target value in model output 2010.

Next; In the step 3130 of decision model input amplitude of variation; In the unit 2600 of the control device that is arranged at Figure 14 2200, utilize the information of the learning information data 2013 of being preserved in the learning information database 2260 of the control device 2200 that is arranged at Figure 14 to decide the model input to change amplitude, ao a.

Next, in the step 3140 of decision model input, in above-mentioned unit 2600, utilize formula (1) to obtain to carry out the learning information data 2012 of study.

a(t+1)＝a(t)+Δa…(1)

Wherein a refers to operation.

Next; In the step 3150 of computation model output; The model input 2009 that has generated by unit 2600 that will in the step 3140 of decision model input, obtain; Be input in the model 2700 of the control device 2200 that is arranged at Figure 14, and this model 2700 is carried out the analog computation of steam power plant 2100, obtain model output 2010.

Next; In the step 3160 of calculating evaluation of estimate; The model output 2010 by model 2700 analog computations that obtains with the step 3150 in computation model output be the basis, and the evaluation of estimate computing unit 2800 of the control device 2200 through being arranged at Figure 14 utilizes formula (2) calculating evaluation of estimate 2011.

Q(s，a)＝E(∑γ ^k?r _t+k+1)…(2)

(in addition, calculate with ∑ and scope be to begin ∞ to k=from k=0)

Wherein, (s a) is the value of selection operation a under state s to Q, and γ (0≤γ＜1) is a slip, and γ t is the remuneration of t constantly.Though by the summation through constantly decide be worth Q (s, a), meaningful therein.

State s is meant the condition of air mass flow, and operation a is meant the for example increase of air mass flow or the minimizing of air mass flow.In addition, (s is meant is worth that the concentration of carbon monoxide reduces when big a) that the concentration that is worth hour carbon monoxide increases to be worth Q.

In fact operate a, the result who utilizes the unit 2600 of control device 2200 to learn here operates the response under the situation of steam power plant 2100, how with time delay.

Therefore, be not to utilize with the operation of operating a to decide value, but decide value more actual by the summation of resulting remuneration in the future shortly past the relative remuneration in back.In addition; Utilize the importing of slip γ; The operation of setting operation a for shortly past after the remuneration that obtains increase, can calculate the evaluation of estimate 2011 of also having considered response through the evaluation of estimate computing unit 2800 that calculates the control device 2200 of evaluations of estimate based on the model output 2010 of model 2700 thus.

In addition, as the remuneration γ that sets at evaluation of estimate computing unit 2800 _tEstablishing method only has the method that when model output 2010 reaches its desired value, gives remuneration and along with model output 2010 is approaching more with desired value, the method that consideration value increases.

And, also can use gimmick shown in Figure 180 and calculate evaluation of estimate 2011.

The model output 2010 that utilizes model 2700 to come analog computation steam power plant 2100 comprises CO CONCENTRATION DISTRIBUTION, O in the exhaust of boiler export stream ₂CONCENTRATION DISTRIBUTION.Through with this CO CONCENTRATION DISTRIBUTION, O ₂Multiplying each other in zone in the boiler export stream that CONCENTRATION DISTRIBUTION and the air of from each air port that is arranged on boiler, supplying with arrive, can calculate influence value shown in Figure 180.

For evaluation of estimate computing unit 2800, identical in calculating with operation signal generation unit 2500, set with the big more mode of the big more evaluation of estimate of first influence value.In addition, also can set with the more little mode of the big more evaluation of estimate of second influence value.

Next; In the step 3170 of upgrading learning parameter; The evaluation of estimate that calculates with incident (episode) end determination step 3180 is the basis; The above-mentioned unit 2600 of the control device 2200 through being arranged on Figure 14 also utilizes formula (3) to carry out computing, the update event parameter, and will upgrade in the learning information database 2260 that the result is kept at control device 2200.

Q(s _t，a _t)←Q(s _t，a _t)+α(r _t+γmax?Q(s _t+1，a _t+1)-Q(s _t，a _t))…(3)

Wherein α (0≤α＜1) is a learning rate.

Finish in the determination step 3180 in incident;, model output 2010 gets into the step 3190 that study finishes judgement when reaching model export target value; When model output 2010 does not reach model export target value, return the step 3130 of decision model input amplitude of variation, wherein model output 2010 is to carry out simulation trial in the step 3150 of computation model output through the model 2700 that calculates to draw.In addition, finish to judge through unit 2600 decision event.

More than; At the process description that utilizes Figure 20 the learning algorithm of unit 2600 of control device 2200 in; Can learn following method of operating; That is, make, contain CO concentration, O in the burning gases of stream of boiler export by 2600 study of above-mentioned unit and in the model input 2009 of model 2700 inputs ₂Concentration is learnt through this unit 2600 again, comes thus the method for operating that reduces the CO concentration in the burning gases of from boiler, discharging is learnt.

Here; In unit 2600, big more when being set at first influence value, when evaluation of estimate is big more; In unit 2600, can learn to increase the method for operating of air mass flow in burner 2102, the air port 2103 of the boiler 2101 that is used for supplying with the air that arrives the low zone of CO concentration.

In addition, likewise in unit 2600, big more when being set at second influence value, evaluation of estimate in unit 2600, can be learnt at O more hour ₂Do not produce the method for operating of air mass flow in burner 2102, the air port 2103 of the boiler 2101 of skew on the CONCENTRATION DISTRIBUTION.

Then; Utilize Figure 21 and Figure 22, to one embodiment of the invention of Figure 14, promptly be used for showing that the instance of attending institute's picture displayed from the image display device 2950 of the output signal of the data transmit-receive handling part 2930 of the maintenance tool 2910 of the control device that constitutes steam power plant 2100 describes.

Figure 21 is with shown in Figure 180 identical, on image display device 2950, shows in the burning gases of each cut zone of stream of 12 parts of boiler exports of cutting apart boiler 2101 CO concentration and O as gas concentration ₂Concentration is with the arrival that air the arrived zone of dropping into from the air port 2103 of boiler 2101 on each cut zone of the stream of having cut apart this boiler export at 12 parts.

In addition; Though the air that Figure 21 expression has dropped into from air port 2103 is in the arrival zone of boiler export; But the air that also can represent from the burner 2102 of boiler 2101, to drop into is in the arrival zone of boiler export, and perhaps the air that from burner 2102 and air port 2103, dropped into of combination expression at random is in the arrival zone of boiler export.

Figure 22 is with shown in Figure 21 identical in addition, expression respectively on image display device 2950: in the burning gases of each cut zone behind the stream of 12 parts of boiler exports of cutting apart boiler 2101 as the CO concentration and the O of gas concentration ₂Concentration, the arrival that air the arrived zone of on each cut zone of 12 parts of streams of having cut apart this boiler export, dropping into from the air port 2103 of boiler 2101, with the CO CONCENTRATION DISTRIBUTION of above-mentioned gas concentration with arrive figure and Figure 19 of multiplying each other in the zone represented influence value is described.

Yet, also have through manually implementing adjustment to the air mass flow that from the burner 2102 of boiler 2101, air port 2103, drops into.At this moment; Can be through on image display device 2950, showing the picture of Figure 21 and Figure 22; The adjustment operation of the air mass flow of supporting to drop into the burner 2102, air port 2103 from boiler 2101 separately obtains shortening required time of adjustment operation of the air mass flow that drops into burner 2102 from boiler 2101, the air port 2103 thus.

According to above-mentioned embodiments of the invention; When on oxygen concentration distributes in the exhaust of the boiler export that acts as a fuel with coal skew being arranged, can realize supplying with necessary air mass flow and come to reduce effectively the control device of the steam power plant of carbonomonoxide concentration in the exhaust of boiler and the control method of steam power plant to the less zone of oxygen concentration.

In addition, owing to can be reduced to the necessary flow of fuel combustion the boiler, come the electric power energy-saving that consumed in steam power plant so can reduce air fed fan power from the air mass flow of boiler input.

(embodiment 3)

Figure 23 is the control block diagram that the integral body of control device of the steam power plant of expression other embodiment of the present invention constitutes.

The control device of the steam power plant of other embodiment of the present invention shown in Figure 23; With before at the control device of the steam power plant of embodiments of the invention shown in Figure 14; Basic device constitute with control method on identical; Therefore omit for the formation of both common devices and the explanation of control method, below different piece is described.

In Figure 23; In the control device 2200 of the control device that is arranged at the steam power plant 2100 that has possessed boiler; As the arithmetic unit in the control device 2200 of the control steam power plant 2100 of the embodiment that is arranged on Figure 14, having appended distributes infers unit 2350 and distributed intelligence database 2270.

Infer in the unit 2350 in distribution; Based on the measuring-signal data 2004 that are kept in the measured value signal database 2210; And the numeric value analysis data 2007 that are kept in the numeric value analysis result database 2230 infer computing, generate the oxygen (O in the exhaust that contains the boiler export stream ₂) distributed intelligence 2019 of CONCENTRATION DISTRIBUTION, carbon monoxide (CO) CONCENTRATION DISTRIBUTION.

Then, this distribution infer unit 2350 infer computing the above-mentioned O of containing ₂The distributed intelligence 2019 of CONCENTRATION DISTRIBUTION, CO CONCENTRATION DISTRIBUTION is kept in the distributed intelligence database 2270.

In the control device 2200 of previous embodiment shown in Figure 14, the O that will measure at measurement of concetration device shown in Figure 18 (b) ₂The measured value of concentration, CO concentration is as the typical value of the cut zone of the boiler export stream that disposes this measurement of concetration device.

In addition, shown in figure 19, in being arranged at the operation signal generation unit 2500 of control device 2200, adopt typical value that concentration in the exhaust in the cut zone of boiler export stream is made as necessarily, calculate influence value.

Yet, in fact because the concentration in the exhaust in the cut zone of boiler export stream is not certain, and be formed with CONCENTRATION DISTRIBUTION, are the O that have in the exhaust of inferring the boiler export stream so unit 2350 is inferred in the distribution of control device 2200 ₂The function of the distribution situation of the distribution situation of concentration, CO concentration, so generate contain infer computing above-mentioned O ₂The device of the distributed intelligence 2019 of CONCENTRATION DISTRIBUTION, CO CONCENTRATION DISTRIBUTION.

And the operation signal generation unit 2500 in the control device 2200 of the steam power plant that is arranged at present embodiment 2100 shown in Figure 23; Not only through generating operation signal 2017, also utilize the distributed intelligence data 2020 that are kept in the distributed intelligence database 2270 to generate the operation signal 2017 as the steam power plant 2100 of control object is formed steam power plant 2100 with the operation signal generation unit 2500 identical formation and the methods of the previous control device that embodiment put down in writing 2200 of Figure 14.

Just, to the first correction signal generation portion 2520 of formation operation signal generation unit 2500 shown in Figure 17 and the input signal of the second correction signal generation portion 2530, append above-mentioned distributed intelligence data 2020 and constitute.

Figure 24 is that the figure that the function of unit 2350 describes is inferred in the distribution of the arithmetic unit of the formation control device for thermal power plant 2200 of other embodiment shown in Figure 23.Wherein, be that example describes with following situation, this situation is to utilize to be kept at the measuring-signal data 2004 in the measuring-signal database 2210, infers the O of the boiler export stream of boiler 2101 ₂CONCENTRATION DISTRIBUTION.

Shown in Figure 24 (a), the cut zone through being arranged on 12 parts of boiler export streams of having cut apart boiler 2101, to the O in the exhaust ₂The O that concentration is measured ₂Measurement of concetration device 2154 is measured the O in the exhaust that each cut zone at boiler export flows down ₂Concentration.

In Figure 24 (a), for through be configured in 12 parts of cut zone of cutting apart the boiler export stream, the O in upper left four zones ₂O in the exhaust that measurement of concetration device 2154a, 2154b, 2154c, 2154d have measured respectively ₂Concentration measurement is expressed as O in the exhaust in Figure 24 (b) ₂The mean value of concentration measurement, dispersion value, maximum and minimum of a value.

Adopt shown in the Trends Sheet, like Figure 24 (b) through O ₂O in the exhaust that measurement of concetration device 2154 measures ₂Concentration value in time process and change.

In Figure 18 of the function of the expression machine of the previous embodiment of Figure 14, Figure 19; For example with the mean value of the concentration measurement of the exhaust of boiler export; As typical value of the boiler export stream being carried out a plurality of zones of having cut apart etc., with smoothing processing the result of measured value as the typical value in the zone of having cut apart.

Relative therewith, unit 2350 is inferred in the distribution in the control device 2200 of present embodiment, shown in Figure 24 (b), utilizes mean value, dispersion value, maximum and the minimum of a value of concentration measurement of the exhaust of boiler export stream, infers concentration, the for example O of exhaust ₂The CONCENTRATION DISTRIBUTION of concentration constitutes.

At first, in 12 parts of cut zone of cutting apart the boiler export stream, through being configured in the O in four upper left zones ₂O in the exhaust that measurement of concetration device 2154a, 2154b, 2154c, 2154d have measured respectively ₂Concentration measurement a, b, c, d, wherein measured value a, b compare with measured value c, d, and the mean value that draws measured value a, b value is lower than the mean value of measured value c, d value.

Based on these measuring-signal data 2004, infer unit 2350 in distribution and judge in 12 parts of cut zone of cutting apart the boiler export stream O ₂The zone that concentration is low is positioned at and is provided with the O that detects measured value a, b ₂In the zone of measurement of concetration device 2154a and 2154b.

Then, infer the dispersion value that unit 2350 utilizes measured value a and measured value b, judge to be provided with O in distribution ₂Which part of O in the zone of measurement of concetration device 2154a and 2154b ₂Concentration is lower.

The dispersion value of measured value a and measured value b than the threshold value of being scheduled to hour, is judged and is for example disposed O ₂The O in the zone of measurement of concetration device 2154a and 2154b ₂Concentration is low, when the dispersion value of measured value a and measured value b is bigger than threshold value, judges and for example disposes O ₂The O on the border in the zone of measurement of concetration device 2154a and 2154b ₂Concentration is low.

This is a determination methods of having utilized following foundation; Promptly; Even if want to control regularly from the burner 2102 of boiler 2101 and air port 2103 and supply with the air mass flow of coming out; But because in fact air mass flow changes in certain scope, so the O for example of the judgement in 12 parts of exhausts of cutting apart the cut zone behind the boiler export stream ₂The CONCENTRATION DISTRIBUTION of concentration also can change in certain scope.

In addition, infer in the unit 2350, when being arranged at 12 parts of specific O of cutting apart the cut zone of boiler export stream in distribution ₂The O that the measurement of concetration device is measured ₂The dispersion value of concentration is judged and is disposed this O than threshold value hour ₂The position of the cut zone of measurement of concetration device, common and O ₂The area coincidence that concentration is low is so this O is disposed in judgement ₂The O in the zone of the part of measurement of concetration device ₂Concentration is low.

On the other hand, as the O that confirms ₂The O that the measurement of concetration device is measured ₂When the dispersion value of concentration is bigger than threshold value, is meant and disposes this O ₂The O of the position of the cut zone of measurement of concetration device ₂Concentration is high or low.

Therefore, judgement is not to dispose this O ₂The position of the cut zone of measurement of concetration device, but on the zone of the position of departing from a little, have O apart from it ₂The position that concentration is low.

In addition, infer in the unit 2350, utilize to be kept at the numeric value analysis data 2007 in the numeric value analysis result database 2230, can also infer CO concentration and O in the exhaust of boiler export stream in the distribution of present embodiment ₂CONCENTRATION DISTRIBUTION.

In the operation signal generation unit 2500 in the control device that is arranged at present embodiment 2200, be utilized in above-mentioned distribution and infer CO concentration and O in the exhaust of the boiler export stream of having inferred unit 2350 ₂The CONCENTRATION DISTRIBUTION information of concentration, calculating operation signal 2017.

Just, with CO concentration measurement and O in the control device 2200 of previous embodiment shown in Figure 19 ₂Concentration measurement is replaced as CO CONCENTRATION DISTRIBUTION presumed value and O respectively ₂The CONCENTRATION DISTRIBUTION presumed value is calculated influence value.

In addition; Even if in the order of operation of the learning algorithm of the unit that is arranged at control device 2,200 2600 of present embodiment; In the step 3160 of calculating evaluation of estimate of flow process Figure 20 of the order of operation of expression learning algorithm; During evaluation of estimate computing unit 2800 calculating mean values through control device 2200, also can utilize above-mentioned influence value.

At this moment, with only utilize CO concentration measurement and O ₂The contrast of the previous embodiment of the measured value information of concentration measurement is utilized CO CONCENTRATION DISTRIBUTION presumed value and O ₂CO CONCENTRATION DISTRIBUTION presumed value and O in the exhaust that the situation of the present embodiment of CONCENTRATION DISTRIBUTION presumed value can obtain and reality is approaching ₂Therefore the CONCENTRATION DISTRIBUTION presumed value can realize the reduction effect of more CO concentration in the exhaust.

This be because; CONCENTRATION DISTRIBUTION in the exhaust of the boiler export stream of in control device 2200, being grasped; Contrast with the typical value of utilizing 12 parts of cut zone of cutting apart the boiler export stream; Utilization is inferred the situation of the CONCENTRATION DISTRIBUTION in the exhaust of having inferred unit 2350 more near actual through distribution, therefore in order to reduce CO concentration, can generate and more can suit to regulate the command signal of from the burner 2102 of boiler 2101 and air port 2103, supplying with the air mass flow of the air mass flow of coming.

Figure 25 is the action flow chart of control device of the steam power plant of other embodiment of the present invention shown in Figure 23.

The flow process of Figure 25 is to flow process shown in Figure 16, and in step 3020 that model is set up with supply with between the step 3030b of position deduction and appended the step 3060 that distributes and infer, other step is identical with the flow process of Figure 16.

In the step 3060 that this distribution is inferred, carry out computing through inferring unit 2350 in the distribution of the control device that is arranged at Figure 23 as stated, infer gas concentration in the exhaust of the boiler export stream (O that distributes ₂CONCENTRATION DISTRIBUTION and CO CONCENTRATION DISTRIBUTION).

If utilize the control device of the steam power plant of above-mentioned other embodiment of the present invention shown in Figure 23, can on image display device 2950, show the information in the distributed intelligence database 2270 that is kept at.Like this; When the control device of the steam power plant that has utilized previous embodiment shown in Figure 14; Like Figure 21, shown in 22; Each regional zoning shows the gas concentration in the exhaust of boiler export on image display device 2950, but if utilize the control device of the steam power plant of present embodiment, then can be more near actual CONCENTRATION DISTRIBUTION state demonstration boiler export gas concentration down.

According to above-mentioned embodiments of the invention, the O in the exhaust of the boiler export that uses coal to act as a fuel ₂CONCENTRATION DISTRIBUTION has under the situation of skew, can realize supplying with necessary air mass flow to O ₂The zone that concentration is less, thus control device and the control method of steam power plant of the steam power plant of the CO in the exhaust of boiler effectively reduced.

In addition, because the air mass flow of putting in the boiler can be reduced to the necessary flow of fuel combustion in the boiler, so the fan power of the air that can abridge supply is reduced in the electric power that is consumed in the steam power plant.

The present invention can be useful in the control device of the steam power plant that is used for reducing the carbonomonoxide concentration of from possessed the steam power plant that uses the boiler that coal acts as a fuel, having discharged and the control method of steam power plant.

Claims (15)

1. the control device of a steam power plant possesses boiler, and this boiler has: with air and the burner of the coal that acts as a fuel to the boiler supply; And the air port, it is arranged in the downstream of the flow direction that makes the burning gases that the air of supplying with from this burner generates with the coal combustion that acts as a fuel and to this burning gases air supply, wherein,

In the boiler of steam power plant, possess measuring appliance, it is measured oxygen concentration or carbonomonoxide concentration in the burning gases of this boiler export,

In the controller of the control device that constitutes steam power plant, possess respectively:

Arrive the zone and infer the unit, it infers the zone that the air of from the burner of above-mentioned boiler or air port, being supplied with arrives boiler export; With

The operation signal generation unit; Its based on the measured value of the measured value of carbonomonoxide concentration in the burning gases of the boiler export of measuring by above-mentioned measuring appliance or oxygen concentration and by above-mentioned arrival zone infer the air that is supplied to of inferring out the unit arrive boiler export the zone infer the result; Setting supplies to the air mass flow the boiler from burner or air port, increases so that arrive the air mass flow in the low zone of carbonomonoxide concentration is high in the burning gases of the boiler export of being measured by above-mentioned measuring appliance zone or oxygen concentration.

2. the control device of a steam power plant has: the burner that air and the coal that acts as a fuel are supplied with in boiler; And the air port, it is positioned at the downstream of the flow direction that makes the burning gases that generate from this burner air of supplying with and the coal combustion that acts as a fuel and to this burning gases air supply, wherein,

In the boiler of steam power plant, possess measuring appliance, it is measured oxygen concentration or carbonomonoxide concentration in the burning gases of this boiler export,

In the controller of the control device that constitutes steam power plant, possess respectively:

Arrive the zone and infer the unit, it infers the zone that the air of from above-mentioned burner or air port, supplying with arrives boiler export;

The unit is inferred in distribution, and it infers the distribution of oxygen concentration in the burning gases of boiler export or the distribution of carbonomonoxide concentration; With

The operation signal generation unit; Its distribution based on carbonomonoxide concentration or oxygen concentration in the burning gases of inferring the boiler export of inferring the unit by above-mentioned distribution infer the result and by above-mentioned arrival zone infer the air that is supplied to of inferring the unit arrive boiler export the zone infer the result; Setting supplies to the air mass flow the boiler from burner or air port, so that the air mass flow that arrives by the high zone that perhaps oxygen concentration is low, zone of carbonomonoxide concentration in the burning gases of the measured boiler export of above-mentioned measuring appliance increases.

3. according to the control device of claim 1 or 2 described steam power plants, it is characterized in that,

In above-mentioned controller, possess:

The numeric value analysis performance element; It has the physical model of simulation steam power plant in inside; Based on the calculating in the path of the air mass flow of supplying with to boiler from the burner that utilizes this physical model or air port, calculate the zone that the air mass flow that supplies to from burner or air port the boiler arrives boiler export; With

Arrive the zone and infer the unit; It utilizes the measuring-signal data of oxygen concentration or carbonomonoxide concentration from the burning gases of the boiler export that steam power plant obtained is measured by above-mentioned measuring appliance or carries out at least one data in the numeric value analysis data that above-mentioned numeric value analysis performance element obtained, and infers the zone that the air mass flow of supplying with to boiler from burner or air port arrives boiler export.

4. the control device of steam power plant according to claim 2 is characterized in that,

In above-mentioned controller, possess:

The numeric value analysis performance element, it has the simulation steam power plant in inside physical model calculates in the burning gases of the boiler export utilize this physical model that oxygen concentration distributes or carbonomonoxide concentration distributes; With

The unit is inferred in distribution; It utilizes in the following data at least one data to infer in the boiler export that oxygen concentration distributes or carbonomonoxide concentration distributes; The oxygen concentration of promptly being measured by above-mentioned measuring appliance from the boiler export burning gases that steam power plant obtains or the measuring-signal data of carbonomonoxide concentration are perhaps carried out the numeric value analysis data that oxygen concentration distributes or carbonomonoxide concentration distributes in the burning gases of the boiler export that above-mentioned numeric value analysis performance element obtained.

5. according to the control device of claim 1 or 2 described steam power plants, it is characterized in that,

Above-mentioned measuring appliance is according to can measuring the measured value of carbonomonoxide concentration or the mode of oxygen concentration and set each cut zone of the cut zone that the boiler export flowing path section is divided into any number,

The operation signal generation unit that in above-mentioned controller, is provided with constitutes has following function; Can calculate respectively the measured carbonomonoxide concentration of the above-mentioned measuring appliance that will set by each cut zone at the boiler export flowing path section measured value or oxygen concentration measured value and multiply each other by the presumed value that the air of from the burner of boiler or air port, being supplied with that above-mentioned arrival zone infers that the unit obtains arrives the zone of boiler export after value summation, be first influence value of carbonomonoxide concentration or second influence value of oxygen concentration, and according to the air mass flow that increases bigger burner of this first influence value proportion or air port, or the mode that reduces the air mass flow in bigger burner of the second influence value proportion or air port control.

6. the control device of steam power plant according to claim 2 is characterized in that,

The operation signal generation unit that in above-mentioned controller, is provided with constitutes has following function; Can calculate first influence value of carbonomonoxide concentration or second influence value of oxygen concentration respectively; And control with the air mass flow that increases bigger burner of this first influence value proportion or air port or the mode that reduces the air mass flow in bigger burner of the second influence value proportion or air port; Wherein, First influence value of this carbonomonoxide concentration or second influence value of oxygen concentration; Be the presumed value of the presumed value that the boiler export flowing path section is divided into carbonomonoxide concentration in the cut zone after the cut zone arbitrarily that will infer that the unit infers by above-mentioned distribution or oxygen concentration and the summation of inferring the value after presumed value that the air of supplying with resulting burner or the air port from boiler, unit arrives the zone of boiler export multiplies each other by above-mentioned arrival zone.

7. the control device of steam power plant according to claim 5 is characterized in that,

In above-mentioned controller, possess: the evaluation of estimate computing unit; It has the simple physical model of the characteristic of simple simulation steam power plant in inside, and is that object calculates above-mentioned first influence value of carbonomonoxide concentration and above-mentioned second influence value of oxygen concentration respectively with this simple physical model; And unit, the method for operating that first influence value that its study is calculated by above-mentioned evaluation of estimate computing unit increases, or the method for operating that reduces of second influence value,

Aforesaid operations signal generation unit is constituted as based on the learning outcome of above-mentioned unit and sets the air mass flow of to boiler, supplying with from burner or air port.

8. according to the control device of claim 1 or 2 described steam power plants, it is characterized in that,

Possess display unit, its arrival zone that will be arrived boiler export by the air that supplies to boiler from burner or air port that the above-mentioned arrival zone that in controller, is provided with infers that the unit infers is presented on the picture.

9. the control device of steam power plant according to claim 5 is characterized in that,

Possess display unit, it will be presented on the picture by first influence value or second influence value that the aforesaid operations signal generation unit that in controller, is provided with calculates.

10. the control method of a steam power plant; Burner to boiler in the air supply and the coal that act as a fuel of above-mentioned steam power plant from being arranged on boiler; And make the air of from this burner, supplying with and the coal that acts as a fuel generate burning gases in the boiler internal combustion, again from the air port that is arranged at boiler to above-mentioned burning gases air supply, above-mentioned air port is in the downstream that flows to of the burning gases that generated; Wherein

By oxygen concentration or carbonomonoxide concentration in the burning gases of the measuring appliance measurement boiler export that in the boiler export of steam power plant, is provided with; And infer the unit by the arrival zone that in the controller of control steam power plant, is provided with and infer the zone that arrives boiler export from the air of above-mentioned burner or the supply of above-mentioned air port; And be utilized in the operation signal generation unit that is provided with in the above-mentioned controller; Arrive based on the measured value of the measured value of carbonomonoxide concentration in the burning gases of the boiler export of measuring by above-mentioned measuring appliance or oxygen concentration with by the air of supplying with from burner or air port that above-mentioned arrival zone infers that the unit infers out boiler export the zone infer the result; Set and control supplies to the air mass flow the boiler from burner or air port, increase so that arrive the air mass flow in the low zone of carbonomonoxide concentration is high in the burning gases of the boiler export of being measured by above-mentioned measuring appliance zone or oxygen concentration.

11. the control method of a steam power plant; The burner of above-mentioned steam power plant from be arranged on boiler be air supply and the coal that acts as a fuel in boiler; And make from this burner air of supplying with and the coal that acts as a fuel and generate burning gases in the boiler internal combustion, the air port from be arranged at boiler is to above-mentioned burning gases air supply again, and above-mentioned air port is in the downstream that flows to of the burning gases that generated; Wherein

By oxygen concentration or carbonomonoxide concentration in the burning gases of the measuring appliance measurement boiler export that in the boiler export of steam power plant, is provided with; And infer the unit and infer the zone that the air of supplying with from above-mentioned burner or above-mentioned air port arrives boiler export by being arranged on arrival zone in the controller that steam power plant is controlled; Infer the distribution that oxygen concentration in the burning gases of boiler export or carbonomonoxide concentration are inferred in the unit by being arranged on distribution in this controller; By the operation signal generation unit that is arranged on this controller; Infer the result and infer the arrival zone that the air that is supplied to of inferring out the unit arrives the zone of boiler export by above-mentioned arrival zone and infer the result based on the distribution of inferring the carbonomonoxide concentration of inferring the unit or oxygen concentration by above-mentioned distribution; Set and control supplies to the air mass flow the boiler from burner or air port, increase so that arrive the air mass flow in the low zone of carbonomonoxide concentration is high in the burning gases of the boiler export of being measured by above-mentioned measuring appliance zone or oxygen concentration.

12. the control method according to claim 10 or 11 described steam power plants is characterized in that,

Through being arranged on the numeric value analysis performance element in the controller that steam power plant is controlled; Come calculating based on the physical model that steam power plant is simulated that is utilized in that its inside has to the path of the air mass flow supplied with to boiler from burner or air port; Infer the unit and calculate the zone that the air mass flow of supplying with to boiler from burner or air port arrives boiler export through being arranged on arrival zone in this controller

The unit is inferred in arrival zone through in above-mentioned controller, being provided with; The oxygen concentration the burning gases of the boiler export of measuring by above-mentioned measuring appliance that utilization obtains from steam power plant or the measuring-signal data of carbonomonoxide concentration, or carry out above-mentioned numeric value analysis performance element and at least one data in the numeric value analysis data that obtain, infer the zone that the air mass flow of supplying with to boiler from burner or air port arrives boiler export.

13. the control method of steam power plant according to claim 11 is characterized in that,

Through the numeric value analysis performance element that in the controller of control steam power plant, is provided with; The oxygen concentration that the physical model that steam power plant is simulated that is utilized in that its inside has calculates boiler export distributes or carbonomonoxide concentration distributes; And infer the unit through the distribution that is arranged in the above-mentioned controller; The measuring-signal data of oxygen concentration or carbonomonoxide concentration or carry out above-mentioned numeric value analysis performance element and the oxygen concentration distributed data that obtains or at least one data in the carbonomonoxide concentration distributed data the burning gases of the boiler export that is measured by above-mentioned measuring appliance that utilization obtains from steam power plant infer that oxygen concentration in the boiler export distributes or carbonomonoxide concentration distributes.

14. the control method according to claim 10 or 11 described steam power plants is characterized in that,

The flowing path section of boiler export is divided into the cut zone of any number; Measure carbonomonoxide concentration or oxygen concentration in the burning gases in each cut zone through measuring appliance; And obtain summation after the measured value of the measured value of the carbonomonoxide concentration of each cut zone of these boiler exports or oxygen concentration and the presumed value of utilizing the air of from the burner of boiler or air port, being supplied with that arrives that the zone infers that the unit obtains to arrive the zone of boiler export multiplied each other, be first influence value of carbonomonoxide concentration or second influence value of oxygen concentration through being arranged on operation signal generation unit in the controller; According to increasing the bigger burner of this first influence value proportion or the air mass flow in air port, perhaps reduce the mode of the air mass flow in bigger burner of the second influence value proportion or air port and control.

15. the control method of steam power plant according to claim 14 is characterized in that,

The physical model that steam power plant is simulated with in the controller that is arranged on the control steam power plant is an object; Calculate above-mentioned second influence value of first influence value and oxygen concentration through the evaluation of estimate computing unit that is arranged on this controller; And the method for operating that increases through first influence value that the unit that is arranged on this controller is learnt to be calculated by this evaluation of estimate computing unit; Perhaps learn the method for operating that second influence value reduces; Again in being arranged at the aforesaid operations signal generation unit of controller, based on by the resulting above-mentioned learning outcome of above-mentioned unit, set the air mass flow of supplying with to boiler from burner or air port.

Images