System and method for optimizing the combustion of a boiler

Abstract

System for a combustion optimization operation for a boiler, wherein the system includes the following: a self-learning modeler (40) configured to generate and update a boiler combustion model; an optimizer (30) configured to receive the boiler combustion model from the modeler (40) and to perform the combustion optimization operation for the boiler using the boiler combustion model to calculate an optimal control value; and an output controller (50) which is configured to check the current operating state of the boiler in real time, to receive the optimal control value from the optimizer (40) and to control the operation of the boiler by reflecting the optimal control value back to a boiler control logic in real time.

Description

BACKGROUND 1. Area

[0001] The devices and methods consistent with the exemplary embodiments relate to a system for controlling a boiler device in a power plant to ensure combustion under optimized conditions, and a method for optimizing the combustion of the boiler device using the same, and in particular to a system and a method for calculating an optimal control value or setpoint for optimizing the combustion of a boiler. 2. Description of the state of the art

[0002] A thermal power plant contains a boiler to heat water using an exothermic reaction produced when a fuel, such as coal, is burned. This heat generates steam to drive a turbine. When combustion occurs in the boiler, emissions such as nitrogen oxides and carbon dioxide are produced. In recent years, the combustion environment has been regulated to reduce emissions because operating the power plant incurs considerable costs to manage these emissions, and efforts are being made to increase the combustion efficiency of the boilers.

[0003] In a state-of-the-art thermal power plant, boiler control, i.e., combustion control, is performed by a skilled person by setting the parameters of the boiler's combustion environment with respect to the performance test data during a trial run and then starting boiler operation. After boiler operation has started, combustion control is further performed by fine-tuning an offset value. Therefore, according to the state-of-the-art boiler operating procedure, maintaining stable combustion for boiler stability is given higher priority than optimal control, because it is not easy to control the boiler in an optimal combustion state while it is operating. A problem in the state of the art was that the optimal combustion environment of a boiler could not be properly implemented.

[0004] To solve the problem, investigations have been carried out to optimize the combustion control of a boiler by automatically capturing and analyzing the boiler's operating data in real time and automatically adjusting various control variables of the boiler according to the analyzed result.

[0005] US 2010 / 0049561 A1 discloses a method and a device for optimizing an FBC power plant.

[0006] US 2016 / 0091203 A1 discloses a method and a device for optimizing the combustion of a boiler.

[0007] US 5,740,033 A discloses a model predictive control system for a process control system, comprising an execution sequencer and an interactive modeler. The interactive modeler includes a process model. SUMMARY

[0008] The aspects of one or more exemplary embodiments provide a system and a method for controlling a boiler in a power plant to calculate an optimal setpoint for a control object in the boiler in order to maximize the combustion efficiency of the boiler while minimizing the generation of emissions including nitrogen oxides and carbon oxides.

[0009] The aspects of one or more exemplary embodiments provide a method for controlling the combustion environment of a boiler in a power plant by applying an artificial intelligence algorithm for controlling boiler combustion, such that through self-learning and modeling a most suitable model for boiler combustion is generated in order to calculate an optimal setpoint required for controlling the combustion environment with respect to the generated model.

[0010] Additional aspects are partly explained in the following description and partly become obvious from the description or can be learned through the practice of the exemplary embodiments.

[0011] The problem is solved by the subject matter of independent claims. Advantageous designs and further developments are the subject of dependent claims.

[0012] According to one aspect of an exemplary embodiment, a system for a combustion optimization operation for a boiler is provided, wherein the system includes: a modeler configured to generate a boiler combustion model; an optimizer configured to receive the boiler combustion model from the modeler and to perform the combustion optimization operation for the boiler using the boiler combustion model to calculate an optimal control value; and an output controller configured to receive the optimal control value from the optimizer and to control the operation of the boiler by reflecting the optimal control value back to a boiler control logic.

[0013] The optimizer can perform the combustion optimization operation using a combustion optimization algorithm.

[0014] The optimizer can calculate a setpoint for at least one control object in the boiler by performing the combustion optimization operation, whereby the combustion optimization operation uses different logics depending on a purpose received from a user.

[0015] The purpose may include cost optimization, which prioritizes cost reduction, emissions optimization, which prioritizes emission reduction, and plant protection optimization, which prioritizes plant protection.

[0016] The combustion optimization operation can be performed according to the following objective function f, f=Cobj1*(factor 1)+Cobj2*(factor 2)+Cobj3*(factor 3), where C is a weighted value for the purpose and factor is an equation for calculating a value for the purpose.

[0017] If the purpose is selected by the user, from among several weighted values ​​contained in the objective function, a weighted value corresponding to the selected purpose can be set to a value greater than the weighted values ​​corresponding to the other purposes not selected by the user.

[0018] The weighted values ​​corresponding to purposes not selected by the user may be greater than zero.

[0019] The optimizer can be configured to collect at least one of the operating data or the status data of the boiler in operation and, based on at least one of the operating data or the status data, to determine whether the combustion optimization operation should be performed for the boiler.

[0020] The operating data may contain at least one of a power generation output, a setpoint or an instantaneous value, wherein the state data may contain at least one of a fluctuation of a boiler output, a fuel fluctuation, a temperature or a pressure in each component of the boiler.

[0021] The optimizer can determine whether the combustion optimization operation for the boiler is to be performed using at least one of an analysis method based on boiler operating data, an analysis method based on a state binary value, or an analysis method based on previously recorded and stored data from the knowledge and experience of the operators.

[0022] According to one aspect of a further exemplary embodiment, a method for performing a combustion optimization operation on a boiler is provided, wherein the method includes: generating a boiler combustion model; performing the combustion optimization operation using the generated boiler combustion model to calculate an optimal control value; and controlling the operation of the boiler by reflecting the optimal control value to a boiler control logic.

[0023] The execution of the combustion optimization operation can include calculating a setpoint for at least one control object in the boiler, whereby the combustion optimization operation uses different logics depending on a purpose received from a user.

[0024] The purpose may include cost optimization, which prioritizes cost reduction, emissions optimization, which prioritizes emission reduction, and plant protection optimization, which prioritizes plant protection.

[0025] The combustion optimization operation can be performed according to the following objective function f, f=Cobj1*(factor 1)+Cobj2*(factor 2)+Cobj3*(factor 3), where C is a weighted value for the purpose and factor is an equation for calculating a value for the purpose.

[0026] If the purpose is selected by the user, from among several weighted values ​​contained in the objective function, a weighted value corresponding to the selected purpose can be set to a value greater than the weighted values ​​corresponding to the other purposes not selected by the user.

[0027] The weighted values ​​corresponding to purposes not selected by the user may be greater than zero.

[0028] The procedure may further include: collecting at least one of the operating data or the state data of the boiler in operation; and determining whether the combustion optimization operation is to be performed for the boiler, based on the at least one of the operating data or the state data.

[0029] The operating data may contain at least one of a power generation output, a setpoint or an instantaneous value, wherein the state data may contain at least one of a fluctuation of a boiler output, a fuel fluctuation, a temperature or a pressure in each component of the boiler.

[0030] According to one aspect of a further exemplary embodiment, a non-transient computer-readable storage medium is provided which stores instructions for executing a method for optimizing a combustion optimization operation on a boiler, wherein the method includes: generating a boiler combustion model; executing the combustion optimization operation using the generated boiler combustion model to compute an optimal control value; and controlling an operation of the boiler by reflecting the optimal control value to a boiler control logic.

[0031] According to one or more exemplary embodiments, the combustion efficiency of the boiler in a power plant can be improved and the emissions that cause environmental pollution can also be minimized, thereby significantly reducing the costs of treating the emissions and thus significantly reducing the operating costs of the power plant.

[0032] In addition, one or more exemplary embodiments can control the boiler in an optimized combustion state with respect to the results learned by the artificial intelligence, so that even unskilled workers can easily achieve a better combustion environment compared to that obtained by a person skilled in the art. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The above and other aspects become more apparent from the following description of the exemplary embodiments with regard to the accompanying drawings; they show: Fig. 1 a schematic graphical representation illustrating a general configuration of a thermal power plant; Fig. 2 a block diagram illustrating a configuration of a boiler control system according to an exemplary embodiment; Fig. 3 a view illustrating a function of an optimizer in the boiler control system according to an exemplary embodiment; and Fig. 4. The optimizer performs operations according to an exemplary embodiment. DETAILED DESCRIPTION

[0034] Various modifications can be made to the embodiments of the disclosure, and there may be different types of embodiments. Consequently, specific embodiments are illustrated in the drawings, and these embodiments are described in detail in the description. However, it should be noted that the various embodiments are not intended to limit the scope of protection of the disclosure to a specific embodiment, but rather should be interpreted to include all modifications, equivalents, or alternatives of the embodiments contained in the ideas and technical scopes disclosed herein. Meanwhile, in cases where it is determined that, in describing the embodiments, a detailed explanation of related known techniques may unnecessarily obscure the main point of the disclosure, the detailed explanation is omitted.

[0035] Unless otherwise defined, the terms used herein, including technical and scientific terms, have the same meanings as they would generally be understood by those skilled in the art in the relevant field. However, these terms may vary depending on the intentions of those skilled in the art, legal or technical interpretation, and the emergence of new techniques. In addition, some terms are arbitrarily chosen by the applicant. These terms may be interpreted according to the meaning defined or described herein, and, unless otherwise specified, may be interpreted on the basis of the entire content of this description and the usual technical knowledge in the field.

[0036] The functional blocks illustrated in the drawings and described below are merely examples of possible implementations. Other implementations may use different functional blocks without compromising the scope of protection provided in the detailed description. While one or more functional blocks of this disclosure are represented by separate blocks, one or more of the functional blocks may be a combination of different hardware and software configurations that perform the same function.

[0037] Furthermore, a "module" or "part" in the disclosure performs at least one function or operation, and these elements may be implemented as hardware or software, or as a combination of hardware and software. Furthermore, several "modules" or "parts" may be integrated into at least one module and implemented as at least one processor, with the exception of those "modules" or "parts" that must be implemented as specific hardware.

[0038] The terminology used here is solely for the purpose of describing specific embodiments and is not intended to limit the scope of protection of the disclosure. The singular forms "a," "an," and "the," as used here, are intended to include the plural forms as well, unless the context clearly indicates otherwise. Furthermore, the terms "comprises," "contains," or "indicate / shows" should be interpreted as indicating the presence of such features, regions, integers, steps, operations, elements, components, and / or a combination thereof in the description, and not as excluding the presence or possibility of adding one or more other features, regions, integers, steps, operations, elements, components, and / or combinations thereof.

[0039] Additionally, terms relating to fastenings, coupling and the like, such as "connected" and "coupled", refer to a relationship in which the structures are either directly or indirectly attached or connected to each other by intermediate structures.

[0040] Furthermore, terms such as "first," "second," etc., can be used to describe different elements, but these terms should not restrict the elements. The terms are simply used to distinguish one element from others. The use of such ordinal numbers should not be interpreted as limiting the meaning of the term. For example, the components assigned to such an ordinal number should not be restricted in terms of their order of use, arrangement, or the like. Any ordinal number can be used synonymously where appropriate.

[0041] Expressions such as "at least one of" when preceding a list of elements modify the entire list and not the individual elements of the list. For example, the expression "at least one of a, b, and c" should be understood to include only a, only b, only c, both a and b, both a and c, both b and c, all of a, b, and c, or any variations of the above examples.

[0042] The exemplary embodiments are described in detail below with respect to the accompanying drawings. To clearly illustrate the disclosure in the drawings, some elements that are not essential for a complete understanding of the disclosure may be omitted, with the same reference numerals referring to the same elements throughout the description.

[0043] Fig. Figure 1 illustrates a general configuration of a thermal power plant, specifically to demonstrate the position and function of a boiler. Thermal power plants generate steam from the power of burning coal or petroleum to drive steam turbines and produce electrical energy. A boiler in a thermal power plant serves to boil water by burning fuel, in order to supply high-temperature, high-pressure steam to the steam turbines. The boiler may include a boiler body containing water and steam, a combustion chamber for burning fuel, and a furnace. The combustion chamber, furnace, and other components are controlled by a control system to regulate temperature, pressure, and other parameters.

[0044] Boiler control is a crucial control operation in a power plant. Previously, a specialist was required for normal operation because boiler control was a very complex process. Recently, by applying an automated control method, a boiler control system without manual intervention has been implemented. This automated control method enables real-time boiler control. Furthermore, to increase the boiler's combustion efficiency, a control system can be implemented that gradually approaches real-time operation by continuously monitoring the boiler's current state, allowing the respective control operations to be performed according to the current state and target parameters.

[0045] The exemplary embodiment provides a boiler control system and procedure which, by adding to a currently available boiler control system and procedure (i) the generation and updating of a boiler combustion model using artificial intelligence and (ii) an optimization operation to find an optimal setpoint for each control objective with respect to the state of a boiler in operation, can improve combustion efficiency and reduce emissions.

[0046] Fig. Figure 2 illustrates a block diagram of a boiler control system according to an exemplary embodiment. Fig. 2 The boiler control system contains a task manager 10, a preprocessor 20, an optimizer 30, a modeler 40, and an output controller 50. Although the boiler control system according to Fig. Since the boiler control system contains two configuration blocks, each designated by a specific function or step, it is recognized that the boiler control system can be implemented as a device including a CPU for operation and memory capable of storing a program and the data for operation. Furthermore, the boiler control system configurations described above can be implemented in a program written in a computer-readable language and executed by the CPU. The boiler control system can also be implemented using hardware or firmware, software, or a combination thereof. If the boiler control system is implemented using hardware, it may include an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA), and the like.If the boiler control system is implemented using firmware or software, it may contain a module, procedure, or function that performs the above functions or operations.

[0047] Task Manager 10 is configured to check the current operating state of the boiler and determine whether boiler combustion optimization should be performed. Task Manager 10 can, for example, collect operating data and state data (such as a state binary value) from the boiler while it is running and, based on this data, determine whether boiler combustion optimization is possible (i.e., whether boiler combustion optimization should be performed). The boiler's operating data includes the measured values ​​received from various sensors installed in the currently operating boiler, or the control values ​​that can be monitored by the boiler control system. Examples of operating data include power output (MW), commands, and the like. The state data includes values ​​indicating fluctuations in boiler output, fuel fluctuations, temperature, and pressure in each component, etc.

[0048] Task Manager 10 determines, based on collected operational and status data, whether combustion optimization is necessary or possible. This determination process considers the integrity of the boiler system (e.g., the operating status of the hardware, the condition of the system equipment, the communication environment, etc.) and the integrity of the individual modules within the boiler system (e.g., the operating status of the software, the presence of the boiler combustion model, etc.). For example, Task Manager 10 may determine that the power plant is unstable during a period of drastic power output fluctuations. If, for instance, the power output (e.g., 500 MW) deviates by several tens of megawatts (e.g., 50 MW) above a reference value over 30 minutes, Task Manager 10 may conclude that this is an unstable state and therefore cannot proceed with combustion optimization.

[0049] There are three analysis methods to determine whether combustion optimization is necessary or possible. These methods include one based on boiler operating data, one based on a state binary value, and one based on the previously recorded and stored knowledge and experience of an expert. These methods can be used individually or in combination by Task Manager 10 to determine whether combustion optimization should be performed. The analysis method based on the previously recorded and stored knowledge and experience of an expert is configured to perform the analysis based on the previously stored data, including the operating states of a boiler and the corresponding influences from an expert operating the boiler. The expert's influence, such as...For example, a supply B of fuel to a combustion chamber of a boiler, if the temperature in the combustion chamber is A, and the influence of the expert, such as setting a damper angle to D, if the temperature in the combustion chamber is C, can be stored and later referred to for analysis.

[0050] The preprocessor 20 is configured to pre-filter only the data suitable for modeling, i.e., the data suitable for training by the modeler 40. While a significant portion of the input data that can be collected by the boiler and the output data corresponding to the input data is much larger, some data contains error values ​​that are unfitted or poorly correlated, thus reducing the accuracy of the modeling. The preprocessor 20 may have a configuration necessary to further improve accuracy by pre-filtering such unnecessary data when a model is generated in the modeler 40.

[0051] The preprocessor 20 performs signal recovery, filtering, and outlier processing functions. The signal recovery function is configured to restore signals collected by the boiler if there is any signal loss, or to restore the corresponding signal if the boiler exhibits an anomaly or fault. The filtering function is configured to filter out data outside the normal data range from the restored signals, or to remove signal noise, and furthermore to extract only the data used for modeling, optimization operations, and output control, using known knowledge-based logic. The outlier processing function is configured to process data outside the trend using data-based logic.

[0052] The preprocessor 20 can be implemented to execute a tag clustering function and a data sampling function according to a developer's intent or a user's needs. Here, the tag clustering function constructs a data group by deleting unnecessary tag information and extracting only the relevant tag information from among the tags that correspond to the respective control objectives of a boiler. The data sampling function acts on data patterns and samples data according to a sampling algorithm to output the training data required for modeling.

[0053] As described above, the preprocessor 20 serves to collect the data associated with the operation of the boiler and to process the data into a form suitable for future modeling.

[0054] Optimizer 30 is a configuration that computes an input for generating an optimal combustion state using a boiler combustion model generated by Modeler 40. Optimizer 30 performs a function of receiving an optimization purpose selection from a user, a function of receiving a boiler combustion model from Modeler 40, and a function of performing a boiler combustion optimization using the boiler combustion model.

[0055] The function of receiving the optimization purpose selection from a user, i.e., an operator of the boiler control system, allows the user to first select a purpose for combustion optimization before executing the combustion optimization. Optimizer 30 can receive the user's selection by providing an interface for selecting multiple purposes. These multiple purposes could include, for example, cost optimization, which prioritizes cost; emissions optimization, which prioritizes emission reduction; and device protection optimization, which prioritizes device protection. It is recognized that this is merely an example and that other optimization purposes are possible. Optimizer 30 performs an optimization operation by applying various logics according to the user's selected purposes.

[0056] Regarding the function of receiving the boiler combustion model from the modeler 40, the optimizer 30 requires a boiler combustion model to perform the optimization operation, where the boiler combustion model can consist of a combination of mathematical models including an artificial neural network, which can be generated by repeated learning by the modeler 40.

[0057] Regarding the function of performing the boiler combustion optimization operation, optimizer 30 calculates an optimal input value as a final output value by running a simulation using the user's purpose selection and the boiler combustion model. The algorithms or controllers used in this case may include proportional-integral-differential algorithms (PID algorithms), degrees-of-freedom algorithms (DOF algorithms), model predictive control algorithms (MPC algorithms), adaptive algorithms, fuzzy algorithms, H-infinity algorithms, model-based linear parameter variation algorithms (LPV algorithms), particle swarm optimization algorithms, genetic algorithms (GA), etc.

[0058] As described above, the optimizer 30 performs an optimization operation according to the user's purpose selection and the boiler combustion model received from the modeler 40 in order to calculate the optimal input value required for boiler combustion control.

[0059] The modeler 40 generates a boiler combustion model that can be used in the optimizer 30. According to an exemplary embodiment, it is characterized in that the modeler 40 generates the boiler combustion model using an artificial neural network.

[0060] An artificial neural network (ANN) is a data processing methodology that simulates inductive learning by mathematically modeling the information processing structure of a brain composed of neurons. This methodology's primary purpose is to correlate patterns between input and output values ​​and to predict an output value from a new input value based on the derived pattern. The ANN consists of parallel interconnected structures (layers) of nodes that represent the neurons. Generally, the neural network has a serial connection: input layer – hidden layer – output layer. Alternatively, the neural network can be implemented with multiple hidden layers to handle the complex correlation between input and output values.When the artificial neural network is used, it is possible to learn the correlation using only the input and output values, predict multiple outputs, and derive the correlation between the input and output values ​​without linear extrapolation for nonlinear behavior, even if the physical properties or the correlation are not clearly known.

[0061] The modeler 40 can receive an input and an output value related to boiler combustion from the preprocessor 20. Examples of input data can include the damper angles of primary and secondary air, the damper angle of a combustion air nozzle (OFA), the amount of coal supplied by a coal feeder, the ambient temperature, etc. Examples of output data can include boiler output, the temperature and pressure of combustion gas in the boiler, the amount of nitrogen oxides, carbon monoxide, and oxygen in the combustion gas, the spray flow rate of a reheater, and the like.

[0062] As described above, the modeler 40 generates a boiler combustion model similar to the actual operating state of the boiler using an artificial neural network, and the generated boiler combustion model is provided to the optimizer 30.

[0063] Output controller 50 is configured to execute the combustion control of the boiler. Output controller 50 includes a function for checking the current operating state of the boiler before controlling the boiler, and a function for reflecting the optimal control value calculated by optimizer 30 by applying the optimal control value to the existing boiler control logic.

[0064] Regarding the function of checking the boiler's operating state, the output controller 50 must check the current operating state of the boiler before actually controlling the boiler. This is because, even if the optimal control value calculated by the optimizer 30 is immediately reflected in the boiler's operating state, the boiler may be in an unstable state or a fault may have occurred. Therefore, the optimal control value should be appropriately divided and reflected according to the boiler's current operating state.

[0065] The output controller 50 can maximize the actual combustion efficiency of the boiler by inputting the optimal control value, previously calculated by the optimizer 30, into the operating boiler. Here, the output controller 50 dynamically tracks systematic errors in the calculated optimal control value, thereby reflecting the optimal control value in real time to the existing combustion logic of the boiler. Assuming, for example, that the optimal control value is a temperature value T100 in a combustion chamber of the boiler, the output controller 50 should execute a control operation to increase the temperature from T1 to T100 if the current temperature in the combustion chamber is T1. A sudden change in temperature can cause problems, so the temperature should be regulated in stages.In this case, the output controller 50 can change the temperature in stages by dividing a temperature range to be changed (up to T100) into several continuous sub-ranges. For example, the output controller 50 can regulate the temperature to gradually increase it in a first stage from T1 to T20, in a second stage from T20 to T40, in a third stage from T40 to T60, in a fourth stage from T60 to T80, and in a final fifth stage from T80 to T100. This period is shorter than the period in which the optimal control value is calculated by the optimizer 30. Assuming, for example, that the optimizer 30 calculates the optimal control value every 5 minutes, the output controller 50 can perform operational control of the boiler every 10 seconds.This means that the execution of the boiler's operating control in all short time periods is defined as the dynamic tracking of systematic errors, which is provided to check the boiler's operating state in real time and to simultaneously and stably reflect the optimal control value without a sudden change in the boiler's operation.

[0066] Fig. Figure 3 illustrates a function of the optimizer 30 in the boiler control system according to an exemplary embodiment. Fig. In section 3, the optimizer 30 obtains an optimal control value through a boiler combustion model and a combustion optimization algorithm. It is recognized that the combustion optimization algorithm can be a set of procedures, methods, and instructions for combustion optimization and can be replaced by another term, such as combustion optimization technology, combustion optimization control (combustion optimization controller), or the like.

[0067] Here, the boiler combustion model is generated by the modeler 40 based on the results learned by the artificial neural network.

[0068] Although the combustion optimization algorithm can consist of various types of algorithms, in another exemplary embodiment it can be a control system using a particle swarm optimization technique. However, it is recognized that the system does not necessarily use the particle swarm optimization algorithm as the combustion optimization algorithm; other types of algorithms, such as PID, DOF, MPC, or the like, can also be used as the combustion optimization algorithm.

[0069] Particle swarm optimization techniques are classified as swarm intelligence techniques, which are stochastic global optimization techniques developed by drawing inspiration from the social behavior of animals such as fish or birds. The particle swarm optimization algorithm replicates this behavior for a large number of entities, referred to as particles, to find an optimal solution within a given search domain based on information about each particle and the group of particles as a whole. Compared to other heuristic optimization techniques, the particle swarm optimization algorithm is easy to implement because it can perform a search using only four arithmetic operations. In particular, it is easy to analyze a natural phenomenon that cannot be differentiated because it does not use gradient information.

[0070] It is recognized that the optimizer 30 inputs multiple control variables into a single boiler combustion model and repeatedly performs a process of converging the control variable as a single particle towards an optimal control value using the particle swarm optimization algorithm.

[0071] Fig. Figure 4 illustrates the optimizer 30 in more detail according to an exemplary embodiment. Fig. Figure 4 states that the optimizer 30 contains a purpose selection section 301, a model reception section 303, and a computation section 305 for optimal control values. It is recognized that the purpose selection section 301, the model reception section 303, and the computation section 305 for optimal control values ​​can be implemented by a CPU that executes computer-readable code stored in memory.

[0072] Purpose Selection Section 301 can provide a list of multiple purposes through an interface to a user, i.e., a user operating a boiler, to allow the user to select the purpose for which an optimization operation should be performed. If a specific purpose is selected by the user, Purpose Selection Section 301 can receive the selected purpose.

[0073] Optimizer 30 is used to calculate a control value by the calculation section 305 for optimal control values. In this case, the calculation section 305 for optimal control values ​​performs an operation using different logic according to the purpose selected by the user.

[0074] Purpose Selection Section 301 can provide a user with a list of at least 3 purposes, including cost optimization (i.e., profit maximization), which considers cost as a top priority; emissions optimization (i.e., emissions minimization), which considers emissions reduction as a top priority; and asset protection optimization (i.e., asset durability), which considers asset protection as a top priority; and receive input from the user to select any of the purposes.

[0075] Model receiver section 303 receives a boiler combustion model generated by modeler 40. Model receiver section 303 can receive the boiler combustion model within a predefined period cycle or regardless of a period cycle. Modeler 40 can continuously generate new boiler combustion models and, after accumulating several, select one from among them that has the same operating state as, or most closely resembles, the current operating state of the boiler. Model receiver section 303 can then receive the selected boiler combustion model from modeler 40.

[0076] The calculation section 305 for optimal control values ​​calculates an optimal control value (i.e., a setpoint) for at least one control object in a boiler based on the purpose selected by the user and the boiler combustion model provided by the modeler 40.

[0077] The calculation section 305 for optimal control values ​​uses different logics depending on the purpose selected by the user. For example, if profit maximization is selected, the logic is chosen to minimize the total costs, including fuel and emission treatment costs, associated with improved plant efficiency. Here, improving plant efficiency can also include reducing both the total amount of reheat spray and the amount of oxygen in the combustion gas. If emission minimization is selected, the logic is chosen to minimize the amount of nitrogen oxides (NOx) and carbon monoxide (CO) in the combustion gas. If plant durability is selected, the logic is chosen to minimize variations in temperature for each chamber within the boiler and variations in the injection volume of the reheat spray for each chamber within the boiler.

[0078] On the other hand, even if an operating mode is selected by the user, calculation section 305 for optimal control values ​​performs one operation for the purpose corresponding to the selected operating mode and one operation for the other purposes that have not been selected. In other words, even if the profit maximization is selected by the user, calculation section 305 for optimal control values ​​can perform one operation on the operating variable corresponding to the selected purpose and one operation on the operating variable corresponding to the other purposes, i.e., emission minimization and plant durability, which have not been selected, in order to calculate the optimal control value.This is to prevent a situation in which, if an operation to calculate an optimal control value is performed only for the operating variable corresponding to any of the purposes, the operating state of a boiler is worsened due to other operating variables that are not taken into account in this operation.

[0079] It is recognized that the three types of purposes described above are merely examples and various other purposes may be provided according to the intention of a developer for a boiler control system.

[0080] The calculation section 305 for optimal control values ​​can perform the calculation according to the following generalized objective function f. f=Cobj1*(factor 1)+Cobj2*(factor 2)+Cobj3*(factor 3), where C represents weighted values ​​according to a specific purpose selected by a user, and the respective weighted values ​​can vary according to the purpose selected by the user. For example, if the user selects a first purpose, a weighted value C obj1 a relatively larger value than C obj2 or C obj3 If the user additionally selects a second or third purpose, a weighted value C is displayed. obj2 or C obj3each factor exhibits a relatively larger value than the others. Factor 1, Factor 2, and Factor 3 of the objective function f refer to equations for calculating actual values ​​according to their respective purposes. For example, Factor 1 might contain an equation for calculating the cost of the fuel to be supplied and an equation for calculating the cost of using an intermediate superheater spray or the like. Factor 2 might contain an equation for calculating pollutant emissions, while Factor 3 might contain an equation for predicting the service life of various components installed in a boiler.

Claims

[1] System for a combustion optimization operation for a boiler, wherein the system comprises: a self-learning modeler (40) configured to generate and update a boiler combustion model; an optimizer (30) configured to receive the boiler combustion model from the modeler (40) and to perform the combustion optimization operation for the boiler using the boiler combustion model to calculate an optimal control value; and an output controller (50) which is configured to check the current operating state of the boiler in real time, to receive the optimal control value from the optimizer (40) and to control the operation of the boiler by reflecting the optimal control value back to a boiler control logic in real time. [2] System according to claim 1, wherein the optimizer (30) performs the combustion optimization operation using a combustion optimization algorithm. [3] System according to claim 1 or 2, wherein the optimizer (30) calculates a setpoint for at least one control object in the boiler by performing the combustion optimization operation, wherein the combustion optimization operation uses different logics depending on a purpose received from a user. [4] System according to claim 3, wherein the purpose includes cost optimization which considers cost as a top priority, emission optimization which considers emission reduction as a top priority, and plant protection optimization which considers plant protection as a top priority. [5] System according to claim 3 or 4, wherein the combustion optimization operation is performed according to the following objective function f: f=Cobj1*(factor 1)+Cobj2*(factor 2)+Cobj3*(factor 3), where C is a weighted value for the purpose and factor is an equation for calculating a value for the purpose. [6] System according to claim 5, wherein, if the purpose is selected by the user, from among several weighted values ​​contained in the objective function, a weighted value corresponding to the selected purpose is set to a value greater than the weighted values ​​corresponding to the other purposes not selected by the user. [7] System according to claim 6, wherein the weighted values ​​corresponding to purposes not selected by the user are greater than zero. [8] System according to any of the preceding claims, wherein the optimizer (30) is configured to collect the operating data and / or the state data of the boiler in operation and to determine, based on the operating data and / or the state data, whether the combustion optimization operation is to be performed for the boiler. [9] System according to claim 8, wherein the operating data includes a power generation output and / or a setpoint and / or an instantaneous value and wherein the state data includes a fluctuation of a boiler output and / or a fuel fluctuation and / or a temperature and / or a pressure in each component of the boiler. [10] System according to claim 8 or 9, wherein the optimizer (30) determines whether the combustion optimization operation for the boiler is to be performed using an analysis method based on boiler operating data, and / or an analysis method based on a state binary value, and / or an analysis method based on previously recorded and stored data from the knowledge and experience of the operators. [11] Method for performing a combustion optimization operation on a boiler, the method comprising: Generating and updating a boiler combustion model using a self-learning modeler (40); Performing the combustion optimization operation using the boiler combustion model to calculate an optimal control value; and Checking the current operating status of the boiler in real time and controlling the operation of the boiler in real time by reflecting the optimal control value to a boiler control logic.

Images