JP2006132902A - Combustion condition estimating device and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To easily estimate optimum combustion conditions corresponding to the load of a combustion apparatus under operation. <P>SOLUTION: A storage part 14 of a combustion condition estimating device 1 stores an estimation model for estimating CO and O<SB>2</SB>exhausted from a boiler 2 based on the operating state of the boiler 2, and an estimation model for estimating the operation efficiency of the boiler 2 based on the operating state, wherein the estimation models are formed beforehand based on measured data acquired from the combustion apparatus. An optimum condition estimating means 12 estimates optimum combustion conditions 7 in the boiler 2 based on evaluation indexes formed from the estimation models. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a combustion control technique, and more particularly to a combustion condition estimation technique for estimating optimum combustion conditions in a combustion apparatus based on measurement data obtained from a combustion apparatus such as a boiler.

  In combustion control devices that control the combustion state of a combustion device such as a boiler, the reduction of harmful exhaust gases such as CO generated when fuel and air are mixed and burned is a response to recent environmental problems. Therefore, it has become an important issue. On the other hand, from the viewpoint of energy saving, it is necessary to efficiently operate with a small amount of fuel to obtain a desired amount of heat.

Normally, the combustion conditions of the boiler can be controlled by the mixing ratio of fuel and air, that is, the air-fuel ratio (mass ratio of air to fuel). Conventionally, with regard to this combustion condition, a combustion condition that can suppress the CO concentration can be obtained by utilizing the relationship between the concentration of O _{2 to be} supplied and the concentration of CO or CO ₂ contained in the exhaust gas. know.

FIG. 13 is a graph showing combustion characteristics in a general boiler, in which the horizontal axis indicates the air-fuel ratio, and the vertical axis indicates the concentrations of CO ₂ and O _{2 in} the exhaust gas (for example, non-patent document). 1 etc.). In this graph, the characteristic 91 is CO ₂ concentration in the exhaust gas, when raising the air fuel ratio will increase the amount of air to constant fuel (heavy oil), tend to decrease the CO ₂ concentration, This also applies to the CO concentration. Note that the O ₂ concentration indicated by the characteristic 92 increases as the air-fuel ratio increases. Further, the operating efficiency (boiler efficiency) is changed from rising to lowering with a certain air-fuel ratio at the top as the amount of air with respect to a certain fuel, that is, the supply amount of O ₂ is increased. Therefore, the optimal combustion condition is an air-fuel ratio in which the CO ₂ concentration and the CO concentration are not more than a reference value and the operation efficiency is good.

The applicant has not yet found prior art documents related to the present invention by the time of filing other than the prior art documents specified by the prior art document information described in this specification.
JP 2002-183111 A Japanese Patent No. 2632117 Kumao Ebihara, “Thermal Management Technology Lecture”, published on April 25, 1979, Maruzen Co., Ltd., pp.108-111 Nishikawa, Sannomiya, Ibaraki, "Iwanami Lecture Information Science-19 Optimization", published September 10, 1982, Iwanami Shoten, pp.162-171

Usually, the combustion characteristics of a combustion device vary according to its load. For example, in the case of a boiler, combustion characteristics fluctuate according to boiler loads such as the required steam flow rate and the temperature of supplied water. At this time, the concentrations of CO and CO ₂ in the exhaust gas increase as the boiler load increases even at the same air-fuel ratio.

Conventionally, such a combustion apparatus such as a boiler has a function of setting a desired air-fuel ratio by an operator operation. However, in consideration of the safety against the load fluctuation described above, the CO concentration can be reduced even when the load fluctuation occurs. The air-fuel ratio is set higher so that it can be suppressed below the reference value. For this reason, although the CO concentration can be suppressed, the operation efficiency is poor, and as a result, there is a problem that a large amount of fuel is wasted and the amount of CO ₂ emission increases.
An object of the present invention is to provide a combustion condition estimation apparatus and method that can easily estimate the optimal combustion condition according to the load of the combustion apparatus during operation.

In order to achieve such an object, a combustion condition estimation device according to the present invention acquires measurement data indicating an operating state from a combustion device that mixes fuel with air and burns, and supplies fuel and air to the combustion device. Is a combustion condition estimation device that estimates the optimal combustion condition indicating the supply amount of the fuel based on the measurement data, and is generated for any CO contained in the exhaust gas of the combustion device that is generated based on the measurement data acquired from the combustion device in advance. and CO concentration estimation model for estimating the CO concentration in accordance with the combustion conditions, and the O ₂ concentration estimation model for estimating the O ₂ concentration in accordance with any of the combustion conditions for the O ₂ contained in the exhaust gas of the combustion apparatus, the combustion apparatus storage means for storing the operating efficiency estimation model for estimating the operating efficiency in accordance with any combustion conditions for operating efficiency, these CO concentration estimation model, O ₂ concentration estimation model, and And a optimal condition estimating means for estimating the desired optimum combustion conditions in the combustion device on the basis of the evaluation index generated from a converter efficiency estimation model.

  At this time, for each estimation model, a plurality of sets of input parameters and output parameters of the estimation model, which are generated in advance based on measurement data from the combustion apparatus, are stored as an estimation model generation database of the estimation model. You may further provide the estimation model production | generation means which produces | generates the said estimation model from each group of a database part and the database for estimation models.

In addition, each estimation model includes, among the measurement data, a fuel state quantity indicating a state of fuel supplied to the combustion apparatus, an air state quantity indicating a state of air supplied to the fuel apparatus, and an operating state of the combustion apparatus. and operation state quantity indicating, and a load state quantity indicating a state of the load applied to the combustion apparatus, CO concentration, O ₂ concentration, and the operating efficiency may be constructed from the model that it it estimates.

Further, the combustion condition estimation method according to the present invention acquires measurement data indicating the operating state from a combustion apparatus that mixes fuel with air and burns it, and optimal combustion conditions that indicate the amount of fuel and air supplied to the combustion apparatus Is a combustion condition estimation method used in a processing apparatus that estimates the measurement data based on measurement data, and is generated based on measurement data acquired in advance from the combustion apparatus, and is set to an arbitrary combustion condition for CO contained in the exhaust gas of the combustion apparatus. and CO concentration estimation model for estimating the CO concentration corresponding, and O ₂ concentration estimation model for estimating the O ₂ concentration in accordance with any of the combustion conditions for the O ₂ contained in the exhaust gas of the combustion apparatus, the operating efficiency of the combustion device and storing the operating efficiency estimation model for estimating the operating efficiency in accordance with any of the combustion condition in the storage unit, these CO concentration estimation model, O ₂ concentration estimation model, Oyo And a optimal condition estimation step of estimating the desired optimum combustion conditions in the combustion device on the basis of the evaluation index generated from operating efficiency estimation model.

  At this time, for each estimation model, a plurality of sets of input parameters and output parameters of the estimation model, which are generated in advance based on measurement data from the combustion apparatus, are used as a database for estimation model generation of the estimation model. And a step of generating the estimation model from each set of the estimation model database.

In addition, each estimation model includes, among the measurement data, a fuel state quantity indicating a state of fuel supplied to the combustion apparatus, an air state quantity indicating a state of air supplied to the fuel apparatus, and an operating state of the combustion apparatus. and operation state quantity indicating, and a load state quantity indicating a state of the load applied to the combustion apparatus, CO concentration, O ₂ concentration, and the operating efficiency may be constructed from the model that it it estimates.

According to the present invention, an estimation model that is generated based on measurement data acquired in advance from a combustion device and estimates CO and O ₂ emitted from the combustion device based on the operation state of the combustion device; An estimation model for estimating the operation efficiency of the combustion apparatus is stored in the storage unit, and the optimum condition estimation means estimates the optimum combustion condition in the combustion apparatus based on the evaluation index generated from these estimation models. Therefore, it is possible to easily estimate the optimum combustion condition according to the load of the combustion apparatus during operation.

Thereby, in order to suppress generation | occurrence | production of the exhaust gas from a combustion apparatus, the air which was supplied excessively can be decreased and it becomes unnecessary to supply excess air to a combustion apparatus. As a result, the amount of generated steam per unit fuel, that is, the operation efficiency can be improved, and the combustion time can be shortened and the CO ₂ emission amount can be reduced.
In general, the CO concentration tends to increase due to the reduction in the amount of air supplied to the combustion device, but according to the present embodiment, the change in the CO concentration due to the change in the combustion conditions can be appropriately predicted. For example, the CO efficiency can be reduced and the operating efficiency can be improved in parallel, for example, the operating efficiency can be improved within a range where the CO concentration is below the reference value, and both environmental problems and energy saving can be achieved. It becomes possible.

Next, embodiments of the present invention will be described with reference to the drawings.
[First Embodiment]
First, with reference to FIG. 1, the combustion condition estimation apparatus concerning the 1st Embodiment of this invention is demonstrated. FIG. 1 is a block diagram showing the configuration of the combustion condition estimation apparatus according to the first embodiment of the present invention.
This combustion condition estimation device 1 is a fuel flow rate and an air flow rate supplied to the boiler 2 based on measurement data 5 obtained from a combustion system such as a boiler 2 by a measurement system 3 comprising various measuring instruments and sensors, or these This is an apparatus for estimating an optimum combustion condition 7 indicated by either one of the flow rate and the air-fuel ratio.

The combustion condition estimation device 1 according to the present embodiment estimates CO and O ₂ discharged from the boiler 2 based on the operation state indicated by the measurement data 5 generated based on the measurement data acquired in advance from the combustion device. The estimation model and the estimation model for estimating the operation efficiency of the boiler 2 based on the operation state are stored in the storage unit, and the optimum combustion condition 7 in the boiler 2 is estimated based on the evaluation index generated from these estimation models. To do.

[Combustion condition estimation device]
Next, with reference to FIGS. 1-3, the structure of the combustion condition estimation apparatus concerning the 1st Embodiment of this invention is demonstrated in detail. FIG. 2 is a block diagram showing the configuration of the estimation model generation means. FIG. 3 is a block diagram showing the configuration of the optimum condition estimating means.
The combustion condition estimation apparatus 1 is composed of an information processing apparatus such as a computer as a whole, and includes an estimation model generation means 11, an optimum condition estimation means 12, a database section 13, a storage section 14, and a measurement data acquisition section 15.

  Among these, the estimated model generation unit 11, the optimum condition estimation unit 12, and the measurement data acquisition unit 15 may be realized by an information processing unit, or may be realized by a dedicated logic circuit that performs predetermined information processing. . The information processing unit includes a microprocessor such as a CPU and its peripheral circuits, and reads and executes a program (not shown) stored in the storage unit 14 to cause the hardware and the program to cooperate with each other. To realize various functional means.

The database unit 13 includes a hard disk, a memory, and the like, and is a storage device that stores various data used for processing in the information processing unit. Main data stored in the database unit 13 includes a CO concentration estimation model database 13A, an O ₂ concentration estimation model database 13B, and an operation efficiency estimation model database 13C.
As will be described later, these databases are composed of a plurality of sets of input parameters and output parameters in the estimation model, which are generated based on measurement data acquired from the boiler 2 in advance. Prior to the generation, the measurement data acquisition unit 15 generates the data in advance and stores it in the database unit 13.

The storage unit 14 is a storage device that includes a hard disk, a memory, and the like, stores various data and programs used for processing in the information processing unit, and may be configured by the same storage device as the database unit 13. The main data stored in the storage unit 14 includes a CO concentration estimation model f1 (14A), an O ₂ concentration estimation model f2 (14B), and an operation efficiency estimation model f3 (14C).
These estimation models are models for estimating CO and O ₂ contained in the exhaust gas of the boiler ₂ , and CO concentration, O ₂ concentration, and operation efficiency according to arbitrary combustion conditions for the operation efficiency of the boiler 2, respectively. These estimation models are generated in advance by the estimation model generation means 11 and stored in the storage unit 14 prior to the estimation of the optimal combustion condition 7.

The estimation model generation means 11 refers to the CO concentration estimation model database 13A, the O ₂ concentration estimation model database 13B, and the operating efficiency estimation model database 13C of the database unit 13, respectively, to thereby determine the CO concentration estimation model f1 (14A). , The O ₂ concentration estimation model f2 (14B), and the operation efficiency estimation model f3 (14C). These estimation models are well-known correlation models composed of a function expression or a data set for estimating each value using any one of the measurement data 5 as an input parameter.

The estimation model generation unit 11 includes a CO concentration estimation model generation unit 11A, an O ₂ concentration estimation model generation unit 11B, and an operation efficiency estimation model generation unit 11C.
The CO concentration estimation model generation unit 11A refers to the CO concentration estimation model database 13A of the database unit 13 based on the measurement data 5, and discharges from the boiler 2 under an arbitrary combustion condition based on the operation state indicated by the measurement data 5. Has a function of generating a CO concentration estimation model f1 (14A) for estimating the CO to be generated.

The O ₂ concentration estimation model generation unit 11B refers to the O ₂ concentration estimation model database 13B of the database unit 13 based on the measurement data 5, and the boiler 2 under an arbitrary combustion condition based on the operation state indicated by the measurement data 5. Has a function of generating an O ₂ concentration estimation model f2 (14B) for estimating the O ₂ discharged from the fuel.
The operation efficiency estimation model generation unit 11C refers to the operation efficiency estimation model database 13C of the database unit 13 based on the measurement data 5 and operates to estimate the operation efficiency of the boiler 2 in the operation state indicated by the measurement data 5. It has a function of generating an efficiency estimation model f3 (14C).

  These estimated models are generated by each estimated model generation means 11 based on each set of database input parameters and output parameters corresponding to the estimated model, for example, RSM-S (Response Surface Method by Spline: multivariable spline). For example, a response surface method (see, for example, Patent Document 1).

The measurement data acquisition unit 15 has a function of acquiring various measurement data 5 indicating the operation state of the boiler 2 from a combustion device such as the boiler 2 via the measurement system 3 including various measuring instruments and sensors, and the measurement data 5 A function for generating a set of database 13A to 13C for generating each model by extracting a set of input parameters and output parameters for each estimated model, and newly obtained measurement data 5 during the estimation operation of optimum combustion condition 7 And a function of outputting to the optimum condition estimating means 12.
When generating each model generation database, the measurement data acquisition unit 15 extracts each model generation database from the measurement data 5 from, for example, the fuel state quantity 5A, the air state quantity 5B, the operation state quantity 5C, and the load state quantity 5D. Generate.

The fuel state quantity 5A is measurement data indicating the state of the fuel, such as the fuel flow rate Qf of the fuel supplied to the boiler 2.
The air state quantity 5 </ b> B is measurement data indicating an air state such as an air flow rate Qa of air supplied to the boiler 2 and an air humidity H.
The operation state quantity 5C is measurement data indicating the operation state of the boiler 2, such as the furnace temperature T of the boiler 2 and the feed water temperature Tw of water supplied to the boiler 2 to become steam.
The load state quantity 5D is measurement data indicating the state of the load applied to the boiler 2, such as the steam temperature Ts, steam pressure Ps, steam flow rate Qs, CO concentration Cco, and O ₂ concentration of steam supplied from the boiler 2 to the load side. It is.

The optimum condition estimation unit 12 includes an evaluation index generation unit 12A, an optimum condition search unit 12B, and a search condition acquisition unit 12C.
The evaluation index generation unit 12A acquires the CO concentration estimation model f1, the O ₂ concentration estimation model f2, and the operation efficiency estimation model f3 generated by the estimation model generation unit 11 from the storage unit 14, and the measurement data acquisition unit 15 Based on the measurement data 5, it has a function of generating an evaluation index 16 obtained by synthesizing these estimation models f1, f2, and f3, that is, D = g (f1, f2, f3).

As the evaluation index 16, for example, a weight parameter method or a regret function widely used in searching for a Pareto optimal solution in a multi-objective optimization problem, for example, in each item of CO concentration, O ₂ concentration, or operation efficiency to be optimized. A known evaluation method using a distance function indicating a distance from the ideal value can be applied (see Non-Patent Document 2 etc.).

  The optimum condition search unit 12B searches for the optimum combustion condition 7 from the evaluation index 16 generated by the evaluation index generation unit 12A based on the measurement data 5 newly acquired by the measurement data acquisition unit 15 and the predetermined search condition 6. And has a function of outputting. For this optimum condition search processing, for example, a known method such as a Newton method or a quasi-Newton method with constraints widely used in a nonlinear programming problem can be applied.

The search condition acquisition unit 12C has a function of acquiring the search condition 6 set and inputted in advance and outputting it to the evaluation index generation unit 12A and the optimum condition search unit 12B.
The main conditions of the search condition 6 include an optimization desired condition and a search constraint condition. The optimization desired condition is a condition for instructing which estimation model is to be emphasized, such as “maximize operation efficiency by minimizing O ₂ under the range where CO is equal to or less than x”. The search constraint condition is a condition for instructing a range for searching for an optimum condition, such as “a range where the fuel flow rate is 50 to 60% and the air flow rate is 50 to 100%”.

[Measurement system]
Next, a measurement system used in the combustion condition estimation apparatus according to the first embodiment of the present invention will be described with reference to FIG. FIG. 4 is an instrumentation diagram illustrating a configuration example of the measurement system 3 used in the combustion condition estimation device 1 according to the first embodiment of the present invention.
The measurement system 3 is composed of a boiler 2 and various measuring instruments and sensors installed around the boiler 2.

  The boiler 2 mixes and burns fuel 22 such as heavy oil, light oil, coal, and gas supplied to the furnace 21 and air 23, heats the water 24 supplied to the boiler 2 with the heat, and generates steam 25. It is a combustion device that generates and supplies the gas to the load side and exhausts the exhaust gas 26 generated by the combustion to the outside of the furnace 21. In the system on the load side, this steam 25 is used for driving a power generation turbine, air conditioning of building facilities, and the like.

The measurement system 3 includes a fuel flow meter 30, an air flow meter 31, an air hygrometer 32, a feed water thermometer 33, a furnace thermometer 34, a steam thermometer 35, a steam pressure gauge 36, steam as main measuring instruments and sensors. A flow meter 37, a CO sensor 38, and an O ₂ sensor 39 are provided.
The fuel flow meter 30 is a flow meter that measures the flow rate Qf of the fuel 22 supplied to the furnace 21. The air flow meter 31 is a flow meter that measures the flow rate Qa of the air 23 supplied to the furnace 21. The air hygrometer 32 is a humidity sensor that measures the humidity H of the air 23 supplied to the furnace 21.

The feed water thermometer 33 is a flow meter that measures the flow rate Qw of the water 24 supplied to the boiler 2. The in-furnace thermometer 34 is a thermometer that measures the temperature T in the furnace 21. The steam thermometer 35 is a thermometer that measures the temperature Ts of the steam 25 supplied from the boiler 2 to the load side. The steam pressure gauge 36 is a pressure gauge that measures the pressure Ps of the steam 25 supplied from the boiler 2 to the load side. The steam flow meter 37 is a flow meter that measures the flow rate Qs of the steam 25 supplied from the boiler 2 to the load side.
The CO sensor 38 is a sensor that detects the CO concentration Cco contained in the exhaust gas 26 discharged to the outside of the furnace 21. The O ₂ sensor 39 is a sensor that detects the concentration Co _{2 of} O ₂ contained in the exhaust gas 26.

  Of the measurement data 5 acquired by these measuring instruments and sensors, the fuel flow rate Qf is output to the combustion condition estimation device 1 as the fuel state quantity 5A, and the air flow rate Qa and the air humidity H are the combustion state conditions as the air state quantity 5B. It is output to the estimation device 1. Further, the feed water temperature Tw and the in-furnace temperature T are output to the combustion condition estimation device 1 as the operation state quantity 5C, and the steam temperature Ts, the steam pressure Ps, and the steam flow rate Qs are set as the load state quantity 5D to the combustion condition estimation apparatus 1. Is output.

[Operation of First Embodiment]
Next, the operation of the combustion condition estimation apparatus according to the first embodiment of the present invention will be described with reference to FIG. FIG. 5 is a flowchart showing combustion condition estimation processing in the combustion condition estimation apparatus according to the first embodiment of the present invention.
The combustion condition estimation apparatus 1 performs the combustion condition estimation process of FIG. 5 by the information processing unit at every input of new measurement data 5, an instruction from the operator, or every predetermined interval. Here, the CO concentration estimation model f1 (14A), the O ₂ concentration estimation model f2 (14B), and the operation efficiency estimation model f3 (14C) are generated in advance in the estimation model generation means 11 and stored in the storage unit 14. It shall be.

First, the information processing unit acquires new measurement data 5 from the operating boiler 2 via the measurement system 3 by the measurement data acquisition unit 15 and outputs the new measurement data 5 to the optimum condition estimation means 12 (step 100).
Next, the information processing unit obtains the CO concentration estimation model f1, the O ₂ concentration estimation model f2, and the operation efficiency estimation model f3 from the storage unit 14 by the evaluation index generation unit 12A of the optimum condition estimation unit 12 (step 101). ), These estimation models f1, f2, and f3 are synthesized based on the new measurement data 5 from the measurement data acquisition unit 15, and an evaluation index (evaluation function) 16, that is, D = g (f1, f2, f3) is generated. (Step 102).

Subsequently, the optimum condition search unit 12B changes the fuel flow rate and the air flow rate based on the new measurement data 5 from the measurement data acquisition unit 15 and the search condition 6 from the search condition acquisition unit 12C, and the evaluation index generation unit 12A. The optimum value Dopt consisting of the maximum value or the minimum value is searched from the evaluation index 16 generated in step (step 103).
Then, the optimum flow rate search means 12B obtains the fuel flow rate Qf ′ and the air flow rate Qa ′ corresponding to the obtained optimum value Dopt or one of these flow rates and the flow rate ratio, that is, the air-fuel ratio of the boiler 2 indicated by the measurement data 5. The optimum combustion condition 7 for the operating state is output (step 104), and the series of combustion condition estimation processing ends.

Thus, in this Embodiment, it is discharged | emitted from the boiler 2 based on the driving | running state of the boiler 2 produced | generated based on the measurement data previously acquired from the combustion apparatus by the memory | storage part 14 of the combustion condition estimation apparatus 1. FIG. An estimation model for estimating CO and O ₂ and an estimation model for estimating the operation efficiency of the boiler 2 based on the operation state are stored, and the optimum condition estimation means 12 uses the evaluation index generated from these estimation models as an evaluation index. Based on this, since the optimum combustion condition 7 in the boiler 2 is estimated, the optimum combustion condition can be easily estimated even when a load fluctuation occurs in the combustion apparatus.

Thereby, in order to suppress generation | occurrence | production of the exhaust gas from a combustion apparatus, the air which was supplied excessively can be decreased and it becomes unnecessary to supply excess air to a combustion apparatus. As a result, the amount of generated steam per unit fuel, that is, the operation efficiency can be improved, and the combustion time can be shortened and the CO ₂ emission amount can be reduced.

  In general, although the concentration of CO, which is a harmful exhaust gas, tends to increase due to a reduction in the amount of air supplied to the combustion device, according to the present embodiment, the change in the CO concentration due to the change in combustion conditions Therefore, it is possible to reduce CO concentration and improve operating efficiency in parallel, for example, by improving the operating efficiency within the range where the CO concentration is below the reference value. It is possible to achieve both.

[Second Embodiment]
Next, with reference to FIG. 6, the combustion condition estimation apparatus concerning the 2nd Embodiment of this invention is demonstrated. FIG. 6 is a flowchart showing an estimation model generation process used in the combustion condition estimation apparatus according to the second embodiment of the present invention.
The combustion condition estimation apparatus 1 performs the estimation model generation process of FIG. 6 by the information processing unit at an instruction from the operator or at predetermined intervals. Here, the CO concentration estimation model database 13A, the O ₂ concentration estimation model database 13B, and the operating efficiency estimation model database 13C are generated by the measurement data acquisition unit 15 and stored in the database unit 13. To do.

The information processing unit first acquires the CO concentration estimation model database 13A from the database unit 13 by the CO concentration estimation model generation unit 11A of the estimation model generation unit 11, and constitutes the CO concentration estimation model database 13A. A CO concentration estimation model f1 (14A) is generated using each set of parameters and output parameters (step 110).
Next, the O ₂ concentration estimation model generation unit 11B of the estimation model generation unit 11 acquires the O ₂ concentration estimation model database 13B from the database unit 13 and configures the O ₂ concentration estimation model database 13B. And an output parameter are used to generate an O ₂ concentration estimation model f2 (14B) (step 111).

Subsequently, the operation efficiency estimation model generation unit 11C of the estimation model generation unit 11 acquires the operation efficiency estimation model database 13C from the database unit 13, and configures the operation efficiency estimation model database 13C. The driving efficiency estimation model f3 (14C) is generated using each set consisting of (step 112).
Then, the estimated model generation means 11 stores the CO concentration estimation model f1 (14A), the O ₂ concentration estimation model f2 (14B), and the operation efficiency estimation model f3 (14C) thus generated in the storage unit 14. Then, a series of estimation model generation processing ends.

When generating these estimation models, the estimation model generation means 11 uses a known inference model generation method such as RSM-S described above, TCBM (Topological Case-Based Modeling) / see, for example, Patent Document 2 ), Multiple regression models, and inference model generation methods such as neural networks.
In particular, since there are many operating states in the boiler 2, by acquiring measurement data discretely for a specific operation state and interpolating the measurement data by applying an inference model generation method such as RSM-S described above Can generate a model with high accuracy.

FIGS. 7 to 9 are explanatory diagrams illustrating examples of generating the CO concentration estimation model f1 (14A), the O ₂ concentration estimation model f2 (14B), and the operation efficiency estimation model f3 (14C). In these figures, the X and Y axes indicate the fuel flow rate Qf and the reciprocal of the air-fuel ratio (fuel flow rate / air flow rate), and the Z axis indicates the CO concentration, O ₂ concentration, and operating efficiency, respectively.
In the above-described combustion condition estimation processing, data corresponding to the measurement data 5 is selected from the data group of each estimation model database, and a response surface model is generated as a desired estimation model from these data.

  In the present embodiment, the estimation model generation process is described as an example performed prior to the combustion condition estimation process separately from the combustion condition estimation process. However, the present invention is not limited to this. In parallel with the condition estimation process, the estimation model generation process may be sequentially performed, and the estimation accuracy can be improved by causing each estimation model to learn the current behavior of the boiler 2 in real time.

Specific input / output parameters used for generating the estimation model are as shown in FIG. 2 described above. Among these parameters, there are parameters that can be estimated from other parameters. FIG. 10 is an explanatory diagram showing a specific example of a process for generating an estimation model.
Generally, the operation efficiency (boiler efficiency) of a boiler is calculated | required by following Formula (1). At this time, the steam flow rate Qs is required.

  At this time, when factor analysis was performed on the steam flow rate Qs, it was found that the steam Qs can be estimated from other measurement data. Below, the case where what was estimated from other measurement data is used as the steam flow rate Qs will be described.

The CO concentration estimation model generation means 11A of the estimation model generation means 11 uses the fuel flow rate Qf, the air flow rate Qa, the furnace temperature T, and the steam pressure Ps as input parameters as specific measurement data, and estimates the CO concentration in the database unit 13. The model database 13A is referred to, and the CO concentration estimation model f1 (14A) is obtained as its output parameter.
Further, the O ₂ concentration estimation model generation means 11B refers to the O ₂ concentration estimation model database 13B of the database unit 13 using the fuel flow rate Qf, the air flow rate Qa, and the air humidity H as specific measurement data as input parameters. The O ₂ concentration estimation model f2 (14B) is obtained as the output parameter.

On the other hand, in the operating efficiency estimation model generation means 11C, first, as specific measurement data, fuel flow rate Qf (fuel state quantity), steam pressure Ps (load state quantity), furnace temperature T (operation state quantity), and air humidity. With H (air state quantity) as an input parameter, the steam flow rate estimation model 13D of the database unit 13 is referred to, and the steam flow rate Qs ′ is obtained as its output parameter.
As specific measurement data, the fuel flow rate Qf, the steam pressure Ps, the in-furnace temperature T, and the steam flow rate Qs ′ obtained from the steam flow rate estimation model 13D are used as input parameters, and the CO concentration estimation model database 13A in the database unit 13 is used. , And an operation efficiency estimation model f3 (14C) is obtained as an output parameter.

  Thus, the steam flow rate estimation model 13D for estimating the steam flow rate Qs ′ from the fuel state quantity, the load state quantity, the operation state quantity, and the air state quantity obtained as the measurement data 5 is stored in the database unit 13 in advance. When generating the operational efficiency estimation model, the steam flow rate Qs ′ is obtained by referring to the steam flow rate estimation model 13D based on the measurement data 5, so that it is not necessary to measure the steam flow rate Qs, and the data amount of the measurement data 5 And the configuration of the measurement system 3 can be simplified.

Here, the case where the steam flow rate Qs is estimated from other measurement data has been described as an example. However, the present invention is not limited to this and may be applied to other measurement data. It is done.
Further, the steam flow rate estimation model 13D may be generated by applying a known inference model generation method as described above.

[Third Embodiment]
Next, a combustion condition estimation apparatus according to the third embodiment of the present invention will be described with reference to FIG. 11 and FIG. FIG. 11 is a flowchart showing database generation processing in the combustion condition estimation apparatus according to the third embodiment of the present invention. FIG. 12 is an explanatory diagram illustrating a generation process of each database used in the combustion condition estimation apparatus according to the second embodiment of the present invention.

In the present embodiment, the measurement data acquisition unit 15 of the information processing unit is provided with a CO concentration estimation model database generation unit 15A, an O ₂ concentration estimation model database generation unit 15B, and an operation efficiency estimation model database generation unit 15C. The estimation model generation means 11 uses the CO concentration estimation model database 13A and the O ₂ concentration estimation model for generating the CO concentration estimation model f1, the O ₂ concentration estimation model f2, and the operation efficiency estimation model f3 from the measurement data 5. Generation processing of the database 13B and the operation efficiency estimation model database 13C will be described.

First, the measurement data acquisition unit 15 operates the boiler from the boiler 2 in various operating states, and acquires measurement data as input parameters and output parameters of each estimation model (step 120). At this time, the input parameter of the estimation model is specified in advance by a known data mining method as a factor having a large correlation with the estimation target, here CO, O ₂ , and the operation efficiency.

Next, a CO concentration estimation model database 13A, an O ₂ concentration estimation model database 13B, and an operating efficiency estimation model database 13C are generated from the set of these input parameters and output parameters.

Specifically, in the CO concentration estimation model database generation means 15A, the fuel flow rate Qf (fuel state quantity), the air flow rate Qa (air state quantity), the furnace temperature T (operation state quantity), and the steam pressure Ps (load) The state concentration) is used as an input parameter, and the CO concentration estimation model database 13A is generated using the CO concentration Cco as an output parameter (step 121).
The O ₂ concentration estimation model database generation means 15B uses the fuel flow rate Qf (fuel state quantity), the air flow rate Qa (air state quantity), and the air humidity H (air state quantity) as input parameters, and O _2. Using the concentration Co ₂ as an output parameter, the O ₂ concentration estimation model database 13B is generated (step 122).

Further, the fuel flow rate Qf (fuel state quantity), the steam temperature Ts (load state quantity), the feed water temperature Tw (operating state quantity), and the steam flow rate Qs (load state quantity) are generated by the operating efficiency estimation model database generation means 15C. The operation efficiency estimation model database 13C is generated using the operation efficiency E as an output parameter as an input parameter (step 123).
Then, the CO concentration estimation model database 13A, the O ₂ concentration estimation model database 13B, and the operation efficiency estimation model database 13C are stored in the database unit 13 (step 124), and a series of database generation processing is terminated.

In each of the above-described embodiments, the information processing unit of the combustion condition estimation device 1 has been described as an example in which each estimation model and further an estimation model database are generated. However, the present invention is not limited to this. The present embodiment may be applied to other information processing apparatuses to generate these estimation models and the estimation model database.
In addition, the measurement data 5 used for generating each estimation model database is not limited to the input parameters shown in FIG. 12 described above, and the fuel state quantity, the air state quantity, and the operation indicating the operation state of the boiler 2. For each parameter of the state quantity and the load state quantity, any one or more or all of the state quantity parameters may be arbitrarily selected and used.

In each of the above-described embodiments, the boiler is described as an example of the target for estimating the optimal combustion condition. However, the present invention is not limited to this, and a combustion apparatus that mixes fuel with air and burns the same as in a boiler. If so, each embodiment can be applied, and the same effect can be obtained.
In addition, the combustion control apparatus is configured by providing the combustion condition estimation apparatus 1 according to each embodiment with a control unit that outputs a control signal for controlling the boiler 2 based on the optimum combustion condition 7 obtained by the estimation. Alternatively, the combustion conditions of the combustion apparatus such as the boiler 2 can be controlled in real time.

It is a block diagram which shows the structure of the combustion condition estimation apparatus concerning the 1st Embodiment of this invention. It is a block diagram which shows the structure of the estimation model production | generation means used with the combustion condition estimation apparatus concerning the 1st Embodiment of this invention. It is a block diagram which shows the structure of the optimal condition estimation means used with the combustion condition estimation apparatus concerning the 1st Embodiment of this invention. It is an instrumentation figure which shows the structural example of the measurement system 3 used with the combustion condition estimation apparatus concerning the 1st Embodiment of this invention. It is a flowchart which shows the combustion condition estimation process in the combustion condition estimation apparatus concerning the 1st Embodiment of this invention. It is a flowchart which shows the estimation model production | generation process used with the combustion condition estimation apparatus concerning the 2nd Embodiment of this invention. It is explanatory drawing which shows the example of a production | generation of the database for CO concentration estimation models. Example of generation of O ₂ database for density estimation model is an explanatory view showing a. It is explanatory drawing which shows the example of a production | generation of the database for driving efficiency estimation models. It is explanatory drawing which shows the example of a specific production | generation process of an estimation model. It is a flowchart which shows the database production | generation process in the combustion condition estimation apparatus concerning the 3rd Embodiment of this invention. It is explanatory drawing which shows the production | generation process of each database used with the combustion condition estimation apparatus concerning the 3rd Embodiment of this invention. It is a graph which shows the combustion characteristic in a common boiler.

Explanation of symbols

1 ... combustion conditions estimating apparatus, 11 ... estimation model generation unit, 11A ... CO concentration estimation model generation unit, 11B ... O ₂ concentration estimation model generation means, 11C ... operating efficiency estimation model generation unit, 12 ... optimal condition estimating means, 12A ... evaluation index generation means, 12B ... optimum condition search means, 12C ... search condition acquisition means, 13 ... database section, 13A ... CO concentration estimation model database, 13B ... O ₂ concentration estimation model database, 13C ... operating efficiency estimation model use database, 14 ... storage unit, 14A ... CO concentration estimation model, 14B ... O ₂ concentration estimation model, 14C ... operating efficiency estimation model, 15 ... measurement data acquisition unit, 15A ... CO concentration estimation model database generation means, 15B ... O ₂ concentration estimation model database generation means, 15C ... operating efficiency estimation model database 16 ... evaluation index, 2 ... boiler, 21 ... furnace, 22 ... fuel, 23 ... air, 24 ... water, 25 ... steam, 26 ... exhaust gas, 3 ... measuring system, 30 ... fuel flow meter, 31 ... Air flow meter, 32 ... Air hygrometer, 33 ... Water feed thermometer, 34 ... In-furnace thermometer, 35 ... Steam thermometer, 36 ... Steam pressure gauge, 37 ... Steam flow meter, 38 ... CO sensor, 39 ... O ₂ Sensors, 5 ... measured data, 5A ... fuel state quantity, 5B ... air state quantity, 5C ... operating state quantity, 5D ... load state quantity, 6 ... search condition, 7 ... optimum combustion condition.

Claims (6)

Combustion conditions for obtaining measurement data indicating an operating state from a combustion apparatus that burns fuel by mixing it with air and estimating optimum combustion conditions indicating supply amounts of fuel and air to the combustion apparatus based on the measurement data An estimation device,
A CO concentration estimation model for estimating a CO concentration according to an arbitrary combustion condition for CO contained in the exhaust gas of the combustion device, which is generated based on measurement data acquired in advance from the combustion device, and the exhaust gas of the combustion device and O ₂ concentration estimation model for estimating the O ₂ concentration in accordance with any of the combustion conditions for the O ₂ contained in the operating efficiency estimation model for estimating the operating efficiency in accordance with any of the combustion conditions for the operation efficiency of the combustion device Storage means for storing
Combustion conditions comprising: optimum condition estimation means for estimating a desired optimum combustion condition in the combustion apparatus based on an evaluation index generated from the CO concentration estimation model, the O ₂ concentration estimation model, and the operation efficiency estimation model Estimating device. In the combustion condition estimation device according to claim 1,
A database that stores a plurality of sets of input parameters and output parameters of the estimation model, which are generated based on measurement data from the combustion device in advance, for each estimation model, as an estimation model generation database of the estimation model And
A combustion condition estimation device, further comprising: an estimation model generation unit that generates the estimation model from each set of the estimation model database. In the combustion condition estimation device according to claim 1,
Each estimation model includes, among the measurement data, a fuel state quantity indicating a state of fuel supplied to the combustion apparatus, an air state quantity indicating a state of air supplied to the fuel apparatus, and the combustion apparatus Combustion conditions characterized by comprising a model for estimating CO concentration, O ₂ concentration, and operation efficiency from an operation state amount indicating an operation state and a load state amount indicating a load state applied to the combustion device. Estimating device. A processing device that acquires measurement data indicating an operating state from a combustion device that mixes fuel with air and burns, and estimates an optimal combustion condition indicating a supply amount of fuel and air to the combustion device based on the measurement data A combustion condition estimation method used in
A CO concentration estimation model for estimating a CO concentration according to an arbitrary combustion condition for CO contained in the exhaust gas of the combustion device, which is generated based on measurement data acquired in advance from the combustion device, and the exhaust gas of the combustion device and O ₂ concentration estimation model for estimating the O ₂ concentration in accordance with any of the combustion conditions for the O ₂ contained in the operating efficiency estimation model for estimating the operating efficiency in accordance with any of the combustion conditions for the operation efficiency of the combustion device Storing in the storage unit;
An optimum condition estimating step for estimating a desired optimum combustion condition in the combustion apparatus based on an evaluation index generated from the CO concentration estimation model, the O ₂ concentration estimation model, and the operation efficiency estimation model. Estimation method. In the combustion condition estimation method according to claim 4,
For each of the estimation models, a plurality of sets of input parameters and output parameters of the estimation model generated in advance based on the measurement data from the combustion device are used as an estimation model generation database of the estimation model in the database unit. Memorizing step;
An estimation model generation step of generating the estimation model from each set of the estimation model database. In the combustion condition estimation method according to claim 4,
Each estimation model includes, among the measurement data, a fuel state quantity indicating a state of fuel supplied to the combustion apparatus, an air state quantity indicating a state of air supplied to the fuel apparatus, and the combustion apparatus Combustion conditions characterized by comprising a model for estimating CO concentration, O ₂ concentration, and operation efficiency from an operation state amount indicating an operation state and a load state amount indicating a load state applied to the combustion device. Estimation method.

Images