KR102275377B1 - Apparatus and Method for Automatic Tuning of Gas Turbine Combustion System Based on Fuel Composition Ratio - Google Patents

Abstract

Provided is an apparatus for automatic tuning of a gas turbine combustion system. The apparatus comprises: a calorific value calculating unit for calculating a low calorific value from a fuel composition ratio of a fuel supplied to a combustion device; and a control unit which selects any one control model corresponding to the calculated low calorific value from among a plurality of control models, and controls the combustion device using the selected control model. An object of the present invention is to provide the apparatus for automatic tuning of the gas turbine combustion system based on the fuel composition ratio, and a method therefor.

Description

Apparatus and Method for Automatic Tuning of Gas Turbine Combustion System Based on Fuel Composition Ratio

The present invention relates to an automatic tuning technology, and more particularly, to an apparatus for automatic tuning of a gas turbine combustion system based on a fuel composition ratio and a method therefor.

A gas turbine is a power engine that mixes and burns compressed air and fuel compressed in a compressor, and rotates the turbine with high-temperature gas generated by combustion. Gas turbines are used to power generators, aircraft, ships, trains, and the like.

A gas turbine generally includes a compressor, a combustor and a turbine. The compressor sucks in the outside air, compresses it and delivers it to the combustor. The compressed air in the compressor is in a state of high pressure and high temperature. The combustor mixes and combusts the compressed air and fuel introduced from the compressor. The combustion gases generated by the combustion are discharged to the turbine. The combustion gas causes the turbine blades inside the turbine to rotate, which in turn generates power. The generated power is used in various fields such as power generation and driving of mechanical devices.

Korean Patent Publication No. 2010-0126773 published on December 02, 2010 (Title: Gas turbine control method and apparatus)

An object of the present invention is to provide an apparatus and method for automatic tuning of a gas turbine combustion system based on fuel composition ratio.

An apparatus for automatic tuning of a gas turbine combustion system based on a fuel composition ratio is provided. The device includes a calorific value calculating unit that calculates a lower calorific value from a fuel composition ratio of fuel supplied to the combustion device, and selects any one control model corresponding to the calculated lower calorific value among a plurality of control models, and uses the selected control model and a control unit for controlling the combustion device.

The control unit inputs target values of the fuel composition ratio, the low calorific value, and the output of the combustion device into a control model to derive a control variable from the control model, and controls the combustion device according to the control variable.

The control unit derives a measured value obtained by measuring the output of the combustion device, and when the difference between the target value of the output of the combustion device and the derived measured value is greater than or equal to a preset threshold, a control model using the measured value of the output of the combustion device characterized by updating.

The control variable is characterized in that it includes at least one of a total fuel amount, an inlet guide vane (IGV) opening degree, a fuel distribution ratio, a fuel nozzle priority, and a bias direction.

The apparatus further includes a learning unit for generating and storing the plurality of control models for controlling the combustion device according to the range of the lower calorific value.

The learning unit collects learning data including the fuel composition ratio of the supplied fuel, calculates the lower calorific value from the fuel composition ratio of the learning data, classifies the learning data according to the range of the lower calorific value, and uses the classified learning data It is characterized in that a plurality of control models are trained according to the range of the lower calorific value.

A method for automatic tuning of a gas turbine combustion system based on fuel composition ratio is provided. The method includes the steps of: a calorific value calculating unit calculating a lower calorific value from a fuel composition ratio of fuel supplied to the combustion device; selecting, by the control unit, any one control model corresponding to the calculated lower calorific value from among a plurality of control models; and controlling the combustion device using the selected control model.

The controlling of the combustion device may include: the control unit inputting target values of the fuel composition ratio, the low calorific value, and the output of the combustion device into a control model to derive a control variable from the control model; and controlling the combustion device accordingly.

The method includes, after controlling the combustion device according to the control variable, deriving, by the control unit, a measured value obtained by measuring the output of the combustion device, and the control unit: a target value of the output of the combustion device and the derived measurement value; The method further includes the step of updating the control model using the measured value of the output of the combustion device when the difference between is equal to or greater than a preset threshold.

The control variable includes at least one of a total fuel amount, an inlet guide vane (IGV) opening degree, a fuel distribution ratio, a fuel nozzle priority, and a bias direction.

The method further includes the step of generating and storing a plurality of control models for controlling the combustion device according to the range of the low calorific value by the learning unit before calculating the low calorific value.

The step of generating and storing the plurality of control models includes the steps of: collecting learning data including a fuel composition ratio of the fuel supplied by the learning unit; calculating the low calorific value from the fuel composition ratio of the learning data by the learning unit; Classifying the learning data according to the range of the additional low calorific value, and the learning unit learning a plurality of control models according to the range of the lower calorific value by using the classified learning data.

According to the present invention, operational variability for a gas turbine combustion system is improved with fuel and calorific fluctuations. This allows active combustion tuning for the gas turbine combustion system.

1 is a block diagram for explaining the configuration of a gas turbine combustion system according to an embodiment of the present invention.
2 is a block diagram illustrating a detailed configuration of a control device of a gas turbine combustion system according to an embodiment of the present invention.
3 is a view for explaining a control model according to an embodiment of the present invention.
4 is a flowchart illustrating a method for learning a control model according to an embodiment of the present invention.
5 is a flowchart illustrating a method for automatic tuning of a gas turbine combustion system based on a fuel composition ratio according to an embodiment of the present invention.
6 is a flowchart illustrating a method of updating a control model according to an embodiment of the present invention.
7 is a diagram illustrating a computing device according to an embodiment of the present invention.

Since the present invention can apply various transformations and can have various embodiments, specific embodiments are illustrated and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and it should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention.

The terms used in the present invention are only used to describe specific embodiments, and are not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise. In the present invention, terms such as 'comprising' or 'having' are intended to designate that the features, numbers, steps, operations, components, parts, or combinations thereof described in the specification exist, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

First, a configuration of a gas turbine combustion system according to an embodiment of the present invention will be described. 1 is a block diagram for explaining the configuration of a gas turbine combustion system according to an embodiment of the present invention. 2 is a block diagram illustrating a detailed configuration of a control device of a gas turbine combustion system according to an embodiment of the present invention. 3 is a view for explaining a control model according to an embodiment of the present invention.

First, referring to FIG. 1 , a combustion system according to an embodiment of the present invention includes a combustion device 100 and a control device 200 . The thermodynamic cycle of the combustion device 100 ideally follows the Brayton cycle. The Brayton cycle consists of four processes: isentropic compression (adiabatic compression), static pressure rapid heat, isentropic expansion (adiabatic expansion), and static pressure dissipation. In other words, it sucks air from the atmosphere, compresses it to a high pressure, burns fuel in a static pressure environment to release heat energy, expands this high-temperature combustion gas to convert it into kinetic energy, and then releases the exhaust gas containing the remaining energy into the atmosphere. . That is, the cycle consists of four processes: compression, heating, expansion, and heat dissipation.

The combustion apparatus 100 includes a compressor 110 , a combustor 120 , and a turbine 130 . The compressor 110 of the combustion device 100 is a part that sucks air and compresses it, and while supplying the combustion air to the combustor 120 , the cooling air in the high temperature region requiring cooling in the combustion device 100 . Its main role is to provide Since the sucked air undergoes an adiabatic compression process in the compressor 110 , the pressure and temperature of the air passing through the compressor 110 are increased.

The compressor 110 included in the combustion device 100 may be designed as centrifugal compressors or axial compressors, but the embodiment of the present invention will be described with an example of an axial compressor. The compressor 110 compresses air introduced from the outside and supplies it to the combustor 120 . The compressor 110 is driven using a part of the power output from the turbine 130 . To this end, the rotation shaft of the compressor 110 and the rotation shaft of the turbine 130 are directly connected. A portion of the output produced by the turbine 130 is consumed to drive the compressor 110 . Accordingly, improving the efficiency of the compressor 110 has a direct and profound effect on improving the overall efficiency of the combustion device 100 .

The combustor 120 is disposed downstream of the compressor 110 , mixes compressed air supplied from the outlet of the compressor 110 with fuel, and then performs isostatic combustion to produce combustion gas of high energy. At this time, the combustor 120 mixes the compressed air compressed in the compressor 110 with the fuel injected and burns it, and discharges the combustion gas generated through such combustion. In the combustion device 100 , gas fuel and liquid fuel, or a combination fuel thereof may be used.

The turbine 130 includes a plurality of turbine blades. The turbine 130 generates electric power by rotating the turbine blades by the combustion gas burned by the combustor 120 . The high-temperature, high-pressure combustion gas produced in the combustor 120 is supplied to the turbine 130 . In the turbine 130 , the thermal energy of the combustion gas is converted into mechanical energy in which the rotary shaft rotates by colliding with a plurality of turbine blades radially disposed on the rotary shaft of the turbine 130 while the combustion gas expands adiabatically and by giving a reaction force. A portion of the mechanical energy obtained from the turbine 130 is supplied as energy required for compressing air in the compressor, and the remainder is utilized as effective energy such as driving a generator to generate electric power.

Referring to FIG. 2 , the control device 200 according to an embodiment of the present invention includes a fuel analysis unit 210 , a calorific value calculation unit 220 , a learning unit 230 , a storage unit 240 , and a control unit 250 . include

The control device 200 according to an embodiment of the present invention controls the combustion device 100 using a plurality of control models generated in advance. As the control model, for example, as shown in FIG. 3 , representatively, an artificial neural network may be exemplified. As shown in FIG. 3, although the artificial neural network is illustrated as a multilayer perceptron, the present invention is not limited thereto, and a convolutional neural network (CNN), a recurrent neural network (RNN), a long short- Various types of artificial neural networks such as term memory models) and GRU (Gated Recurrent Unit) may be selected. A control model according to an embodiment of the present invention includes a plurality of layers, and the plurality of layers includes a plurality of operations. A plurality of layers are connected by weights, which means that the weights are applied to the calculation results of the previous layer and input to the calculation of the next layer. The plurality of layers includes a convolution layer (CVL), a pooling layer (PLL), a fully connected layer (FCL), and the like. Here, the operation means operation of the activation function, and the activation function is Sigmoid, Hyperbolic tangent (tanh), Exponential Linear Unit (ELU), Rectified Linear Unit (ReLU), Leakly ReLU, Maxout. , Minout, Softmax, etc. can be exemplified. For example, as shown in FIG. 3 , the control model according to the embodiment of the present invention includes a plurality of layers including a plurality of operations. At least one weight is applied to each of the plurality of operations. In particular, the sufficiently learned control model may receive the target output as the target value of the fuel composition ratio, the low-level calorific value (LHV), and the output of the combustion device 100 . In addition, according to an additional embodiment, the control model may receive the current control variable and the current output of the combustion device 100 in addition to the fuel composition ratio, the low-level calorific value (LHV), and the target output of the combustion device 100 . When the above-described parameters are input, the control model outputs a target control variable corresponding to the target output by performing a plurality of operations in which a plurality of layer weights are applied to the input parameter. Here, the current control variable is a variable currently input to control the combustion device 100 and causes the current output. In addition, the target control variable includes a plurality of variables input to control the combustion device 100 so that the combustion device 100 outputs a target output. The current control variable and the target control variable include at least one of a total fuel amount, an inlet guide vane (IGV) opening degree, a fuel distribution ratio, a fuel nozzle priority, and a bias direction.

Learning is required to use the control model as described above. The learning unit 230 is to generate a plurality of control models for controlling the combustion device 100 according to the range of the lower calorific value and store the generated control models in the storage unit 240 .

The learning unit 230 includes a control variable used whenever the control unit 250 controls the combustion device 100 , an output of the combustion device 100 according to the control variable, and a fuel of the fuel supplied to the combustion device 100 . Collect the composition ratio. Then, the learning unit 230 calculates a lower heating value (LHV) from the fuel composition ratio of the collected learning data, and classifies and stores the learning data according to the range of the lower heating value. That is, according to the embodiment of the present invention, a plurality of control models are generated according to the range of the lower calorific value. Therefore, the collected learning data is also classified and stored according to the range of the lower calorific value. The learning unit 230 performs learning on the control model through the collected learning data. At this time, the learning unit 230 individually learns a plurality of control models for each range of the lower calorific value, and when the learning is completed, it is stored in the storage unit 240 . Accordingly, the storage unit 240 may store a plurality of control models CM1, CM2, ?, and CMn. Each of the plurality of control models (CM1, CM2, ?, CMn) has its own range of its own lower calorific value (LHV) (minimum value < LHV < maximum value).

The control unit 250 controls the combustion apparatus 100 by selecting any one of a plurality of control models that have been trained. Since different control models for each range of the lower calorific value are learned and generated by the learning unit 230 , the controller 250 may select a control model corresponding to the range of the lower calorific value. For this selection, the fuel analyzer 210 analyzes the fuel supplied to the combustion device 100 to derive a fuel composition ratio of the fuel, and the calorific value calculator 220 determines the fuel derived from the fuel analyzer 210 . Calculate the lower calorific value from the fuel composition ratio of Then, the control unit 250 selects a control model corresponding to the calculated low calorific value from among the plurality of control models.

When the control model is selected, the control unit 250 inputs target values of the fuel composition ratio, the lower calorific value, and the output of the combustion device 100 to the control model. According to an additional embodiment, the control unit 250 may further input the state information of the combustion apparatus 100 in addition to the target values of the fuel composition ratio, the lower calorific value, and the output of the combustion apparatus 100 . The state information includes the control parameters of the current combustion device 100 and the output accordingly. Here, the control variables of the current combustion device 100 include a total fuel amount, an inlet guide vane (IGV) opening degree, a fuel distribution ratio, a fuel nozzle priority (Fuel Circuit Priority), a bias direction, and the like. In addition, the output includes exhaust gas, combustion vibration, and the like.

Then, the control model derives a control variable through a plurality of operations to which the learned weight is applied, and outputs the derived control variable. In this way, the control unit 250 may derive a control variable through the output of the control model. When the control variable is derived, the controller 250 controls the combustion device 100 through the derived control variable.

On the other hand, as described above, after controlling the combustion device 100 through the control variable derived through the control model, the controller 250 may update the control model using the output of the actual combustion device 100 . . That is, the control unit 250 controls the combustion device 100 through the control variable derived through the control model, and then derives a measured value of the output of the combustion device 100 , and determines the output of the combustion device 100 . If the difference between the target value and the measured value is equal to or greater than a preset threshold, the control model is updated using the measured value of the output of the combustion device 100 .

Next, a method for automatic tuning of a gas turbine combustion system based on a fuel composition ratio according to an embodiment of the present invention will be described. Automatic tuning according to an embodiment of the present invention uses a plurality of pre-generated control models. A typical control model may be an artificial neural network. Accordingly, learning of a plurality of control models is required. The learning unit 230 generates a plurality of control models for controlling the combustion device according to the range of the lower calorific value and stores the generated control models in the storage unit 240 . Then, a method for learning a control model according to an embodiment of the present invention will be described. 4 is a flowchart illustrating a method for learning a control model according to an embodiment of the present invention.

Referring to FIG. 4 , the learning unit 230 collects learning data in step S110 . Here, the learning data is used by the controller 250 to control the combustion device 100 . That is, the learning unit 230 includes the control variable used whenever the control unit 250 controls the combustion device 100 , the output of the combustion device 100 according to the control variable, and the fuel supplied to the combustion device 100 . of fuel composition is collected. At this time, the previously used control variable of the used control variable and the output of the combustion device 100 according to the previously used control variable are mapped to each other and stored. At this time, the previously used control variable of the used control variable and the output of the combustion device 100 according to the previously used control variable are used as state information.

The learning unit 230 calculates a lower heating value (LHV) from the fuel composition ratio of the learning data collected in step S120, and classifies and stores the learning data according to the range of the lower heating value. That is, according to the embodiment of the present invention, a plurality of control models are generated according to the range of the lower calorific value. Therefore, the collected learning data is also classified and stored according to the range of the lower calorific value.

Next, the learning unit 230 checks whether learning data greater than or equal to a predetermined value has been collected in step S130 . As a result of the check, if learning data equal to or greater than a predetermined value is collected and stored, the process proceeds to step S140.

The learning unit 230 performs learning on the control model through the learning data collected in step S140. At this time, the learning unit 230 individually learns a plurality of control models for each range of the lower calorific value. These learning methods will be described in detail as follows. For example, the learning unit 230 sets the control variable as an expected value, and inputs the fuel composition ratio, the lower calorific value, and the output of the combustion device 100 corresponding to the control variable set as the expected value to the control model. In this case, according to an additional embodiment, state information may be additionally input. Accordingly, the control model derives a control variable by performing a plurality of calculations to which a weight for the fuel composition ratio, the low calorific value and the output of the combustion device 100, and, optionally, the state information is applied, and the derived control variable output as an output value. Then, the learning unit 230 compares the output value of the control model with the previously set expected value, and corrects the weight of the control model through an optimization algorithm so that the difference between the expected value and the predicted value is minimized. Here, the optimization algorithm may exemplify a backpropagation algorithm. The learning unit 230 learns the control model by repeating the above-described procedure for each of the plurality of learning data. According to the above-described method, the learning unit 230 generates a plurality of control models for controlling the combustion device according to the range of the lower calorific value, and stores the generated control models in the storage unit 240 .

Then, the control unit 250 controls the combustion device 100 using the plurality of control models that have been trained. These methods will be described. 5 is a flowchart illustrating a method for automatic tuning of a gas turbine combustion system based on a fuel composition ratio according to an embodiment of the present invention.

Referring to FIG. 5 , the fuel analysis unit 210 analyzes the fuel supplied to the combustion device 100 in step S210 and derives a fuel composition ratio of the fuel. Then, the calorific value calculation unit 220 calculates the lower calorific value from the fuel composition ratio of the fuel derived by the fuel analysis unit 210 in step S220.

Next, the control unit 250 selects a control model corresponding to the previously calculated low calorific value from among the plurality of control models in step S230 . As described above, different control models were trained and generated for each range of the lower calorific value. Accordingly, the control unit 250 may select a control model corresponding to the range of the lower calorific value.

Then, the control unit 250 derives a control variable corresponding to the output of the target combustion device 100 through the previously selected control model in step S240. At this time, the control unit 250 inputs the target values of the fuel composition ratio, the low calorific value, and the output of the combustion device 100 to the control model. In this case, according to an additional embodiment, state information may be additionally input. Then, the control model derives a control variable through a plurality of calculations to which the fuel composition ratio, the low calorific value and the target value of the output of the combustion device 100, and, optionally, the weight learned for the state information are applied, the derived control print the variable In this way, the control unit 250 may derive a control variable through the output of the control model.

When the control variable is derived, the control unit 250 controls the combustion apparatus 100 through the control variable derived in step S250. These control variables include at least one of a total fuel amount, an inlet guide vane (IGV) opening degree, a fuel distribution ratio, a fuel nozzle priority (Fuel Circuit Priority), and a bias direction.

On the other hand, as described above, after controlling the combustion device 100 through the control variable derived through the control model, the controller 250 may update the control model using the output of the actual combustion device 100 . . These methods will be described. 6 is a flowchart illustrating a method of updating a control model according to an embodiment of the present invention.

Referring to FIG. 6 , in the embodiment with reference to FIG. 5 , the controller 250 derived control variables through the control model and then controlled the combustion apparatus 100 through the derived control variables. Whenever such control is performed, the control unit 250 may measure the output of the combustion apparatus 100 through a plurality of sensors attached to the combustion apparatus 100 in step S310 .

Then, the control unit 250 compares the target value of the output of the combustion device 100 and the actually measured value in step S320.

According to this comparison, the control unit 250 determines whether the difference between the target value and the measured value of the output of the combustion device 100 is equal to or greater than a preset threshold in step S330 .

As a result of the determination in step S330 , if the difference is greater than or equal to the threshold, the controller 250 may update the control model using the measured value of the output of the combustion device 100 in step S340 . That is, the control unit 250 sets the control variable to an expected value, and inputs the measured value of the fuel composition ratio, the lower calorific value and the output of the combustion device 100 , and optionally state information to the control model. Accordingly, the control model derives a control variable by performing a plurality of calculations in which a weight is applied to the measured value of the fuel composition ratio, the low calorific value and the output of the combustion device 100, and, optionally, the state information, and the derived control variable is output as an output value. Then, the learning unit 230 compares the output value of the control model with the previously set expected value and updates the weight of the control model through the optimization algorithm so that the difference between the expected value and the predicted value is minimized.

On the other hand, if the difference is less than the threshold as a result of the determination in step S330, the controller 250 maintains the control model as it is in step S350.

7 is a diagram illustrating a computing device according to an embodiment of the present invention. The computing device TN100 of FIG. 7 may be a device described herein (eg, a device for automatic tuning of a gas turbine combustion system based on fuel composition ratio, etc.).

In the embodiment of FIG. 7 , the computing device TN100 may include at least one processor TN110 , a transceiver device TN120 , and a memory TN130 . In addition, the computing device TN100 may further include a storage device TN140 , an input interface device TN150 , an output interface device TN160 , and the like. Components included in the computing device TN100 may be connected by a bus TN170 to communicate with each other.

The processor TN110 may execute a program command stored in at least one of the memory TN130 and the storage device TN140. The processor TN110 may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to an embodiment of the present invention are performed. The processor TN110 may be configured to implement procedures, functions, methods, and the like described in connection with an embodiment of the present invention. The processor TN110 may control each component of the computing device TN100 .

Each of the memory TN130 and the storage device TN140 may store various information related to the operation of the processor TN110 . Each of the memory TN130 and the storage device TN140 may be configured as at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory TN130 may include at least one of a read only memory (ROM) and a random access memory (RAM).

The transceiver TN120 may transmit or receive a wired signal or a wireless signal. The transceiver TN120 may be connected to a network to perform communication.

On the other hand, the various methods according to the embodiment of the present invention described above may be implemented in the form of a program readable by various computer means and recorded in a computer readable recording medium. Here, the recording medium may include a program command, a data file, a data structure, etc. alone or in combination. The program instructions recorded on the recording medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the art of computer software. For example, the recording medium includes magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floppy disks ( magneto-optical media), and hardware devices specially configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions may include not only machine language such as generated by a compiler, but also a high-level language that can be executed by a computer using an interpreter or the like. Such hardware devices may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

In the above, although an embodiment of the present invention has been described, those of ordinary skill in the art can add, change, delete or add components within the scope that does not depart from the spirit of the present invention described in the claims. The present invention may be variously modified and changed by, etc., and this will also be included within the scope of the present invention.

100: combustion device
110: compressor
120: combustor
130: turbine
200: control device
210: fuel analysis unit
220: calorie calculation unit
230: study unit
240: storage
250: control unit

Claims (12)

a calorific value calculating unit for calculating a lower calorific value from a fuel composition ratio of fuel supplied to the combustion device;
a control unit that selects any one control model corresponding to the calculated low calorific value from among a plurality of control models, and controls the combustion device using the selected control model; and
a learning unit for generating and storing the plurality of control models for controlling a combustion device according to the range of the lower calorific value;
characterized by comprising
Device for automatic tuning. According to claim 1,
the control unit
Input the target values of the fuel composition ratio, the low calorific value, and the output of the combustion device into a control model to derive control variables from the control model,
Controlling the combustion device according to the control variable
Device for automatic tuning. 3. The method of claim 2,
the control unit
Deriving a measured value measuring the output of the combustion device,
When the difference between the target value of the output of the combustion device and the derived measured value is equal to or greater than a preset threshold, the control model is updated using the measured value of the output of the combustion device
Device for automatic tuning. 3. The method of claim 2,
The control variable is
Characterized in that it includes at least one of a total fuel amount, an inlet guide vane (IGV) opening degree, a fuel distribution ratio, a fuel nozzle priority, and a bias direction.
Device for automatic tuning. delete According to claim 1,
the learning unit
Collect learning data including fuel composition ratio of supplied fuel,
Calculate the low calorific value from the fuel composition ratio of the learning data,
Classify the training data according to the range of low calorific value,
Using the classified learning data, it characterized in that the plurality of control models are trained according to the range of the lower calorific value.
Device for automatic tuning. generating and storing, by the learning unit, a plurality of control models for controlling the combustion device according to the range of the lower calorific value;
after generating and storing the plurality of control models, calculating a lower calorific value from the fuel composition ratio of the fuel supplied to the combustion device by a calorific value calculating unit;
selecting, by a control unit, any one control model corresponding to the calculated low calorific value from among the plurality of control models; and
controlling, by a controller, the combustion device using the selected control model;
characterized by comprising
A method for auto-tuning. 8. The method of claim 7,
The step of controlling the combustion device is
deriving control variables from the control model by inputting, by the control unit, target values of the fuel composition ratio, the low calorific value, and the output of the combustion device into a control model; and
controlling, by the control unit, the combustion device according to the control variable;
characterized by comprising
A method for auto-tuning. 9. The method of claim 8,
After controlling the combustion device according to the control variable,
deriving, by the control unit, a measured value obtained by measuring the output of the combustion device; and
If the difference between the target value of the output of the combustion device and the derived measured value is greater than or equal to a preset threshold, the control unit updating the control model using the measured value of the output of the combustion device; characterized by further comprising:
A method for auto-tuning. 9. The method of claim 8,
The control variable is
Characterized in that it includes at least one of a total fuel amount, an inlet guide vane (IGV) opening degree, a fuel distribution ratio, a fuel nozzle priority, and a bias direction.
A method for auto-tuning. delete 8. The method of claim 7,
The step of creating and storing the plurality of control models is
collecting learning data including a fuel composition ratio of fuel supplied by the learning unit;
calculating, by the learning unit, a lower calorific value from the fuel composition ratio of the learning data;
classifying the learning data according to the range of the lower calorific value by the learning unit; and
learning, by the learning unit, a plurality of control models according to a range of a lower calorific value by using the classified learning data;
characterized by comprising
A method for auto-tuning.

Images