JP7699415B2 - PLANT OPERATION OPTIMIZATION DEVICE AND PLANT OPERATION OPTIMIZATION METHOD - Google Patents

Description

Embodiments of the present invention relate to a plant operation optimization device and a plant operation optimization method.

In a steam turbine power plant, a boiler generates steam, which drives a steam turbine, which in turn drives a generator to generate electricity. Specifically, the boiler boils water using thermal energy generated by burning fuel, generating steam from the water. In this case, it is desirable to efficiently recover energy using the boiler.

European Patent Publication No. 1921280 U.S. Pat. No. 7,389,151 U.S. Pat. No. 7,584,024 European Patent Publication No. 2924572

In order to recover energy efficiently using a boiler, it is desirable, for example, to appropriately adjust the boiler's combustion conditions and to remove soot accumulated in the boiler at the appropriate time by soot blowing.

Therefore, an embodiment of the present invention provides a plant operation optimization device and a plant operation optimization method that can achieve efficient energy recovery in a boiler.

According to one embodiment, the plant operation optimization device includes a receiving unit that receives data related to a boiler in a power plant. The device further includes a determining unit that determines a combustion state set value, which is a setting value for controlling the combustion state of the boiler, using a neural network based on the data related to the boiler. The device further includes an output unit that outputs the combustion state set value to a display device.

1 is a schematic diagram showing a configuration of a plant operation optimization system according to a first embodiment. FIG. 1 is a block diagram showing a configuration of a plant operation optimization system according to a first embodiment. 4 is a graph for explaining soot blowing in the first embodiment. 4 is another graph for explaining soot blowing in the first embodiment. FIG. 4 is a diagram illustrating an example of an operator operation screen according to the first embodiment. FIG. 4 is a diagram showing an example of a simulation screen according to the first embodiment;

Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In Figs. 1 to 6, the same components are designated by the same reference numerals, and duplicated descriptions will be omitted.

First Embodiment
FIG. 1 is a schematic diagram showing the configuration of a plant operation optimization system according to a first embodiment.

The plant operation optimization system in FIG. 1 includes a power plant 1, a plant operation optimization device 2, and a display device 3. The power plant 1 is a plant whose operation is to be optimized. The plant operation optimization device 2 is a device that provides information for optimizing the operation of the power plant 1. The display device 3 is a device that displays the information provided by the plant operation optimization device 2.

The power plant 1 is, for example, a steam turbine power plant, and includes a boiler 11, a high pressure (HP) turbine 12, a reheater 13, an intermediate pressure (IP) turbine 14, a low pressure (LP) turbine 15, a generator 16, a condenser 17, and a feed water heater 18. The power plant 1 may be any other type of plant that includes a boiler 11.

The plant operation optimization device 2 may be installed inside the power plant 1 or outside the power plant 1. The plant operation optimization device 2 of this embodiment is installed outside the power plant 1, for example, in the base of the company that designed the power plant 1. In this case, the plant operation optimization device 2 exchanges data with the power plant 1 via a network. The plant operation optimization device 2 is, for example, a PC (Personal Computer). As described later, the plant operation optimization device 2 of this embodiment performs, for example, processing for optimizing the combustion state of the boiler 11 and processing for optimizing soot blowing in the boiler 11.

The display device 3 may also be installed inside the power plant 1 or outside the power plant 1. In this embodiment, the display device 3 is installed, for example, in a building constructed in the power plant 1. In this case, the display device 3 exchanges data with other devices in the power plant 1 and the plant operation optimization device 2 via a network. The display device 3 is, for example, a PC. In this embodiment, the display device 3 is provided separately from the plant operation optimization device 2, but may be provided as part of the plant operation optimization device 2.

The boiler 11 generates steam as the working fluid of the power plant 1. Specifically, the boiler 11 of this embodiment boils water using thermal energy generated by burning fuel, and generates steam from the water. The fuel is, for example, a fossil fuel such as coal. The high-pressure turbine 12 is driven by the steam supplied from the boiler 11.

The reheater 13 reheats the steam discharged from the high-pressure turbine 12. The reheater 13 may be provided as part of the boiler 11, or may be provided separately from the boiler 11. When the reheater 13 is part of the boiler 11, the reheater 13 heats the steam discharged from the high-pressure turbine 12, for example, by using thermal energy generated by burning the above-mentioned fuel. The intermediate-pressure turbine 14 is driven by the steam supplied from the reheater 13. The low-pressure turbine 15 is driven by the steam discharged from the intermediate-pressure turbine 14.

The generator 16 can be connected to a rotating shaft to which the high-pressure turbine 12, the intermediate-pressure turbine 14, and the low-pressure turbine 15 are connected, and when connected to this rotating shaft, it is driven by the high-pressure turbine 12, the intermediate-pressure turbine 14, and the low-pressure turbine 15 to generate electricity.

The condenser 17 converts the steam discharged from the low-pressure turbine 15 back into water. The feedwater heater 18 heats the water discharged from the condenser 17 using the heat of the steam extracted from the high-pressure turbine 12, the intermediate-pressure turbine 14, or the low-pressure turbine 15. The water discharged from the feedwater heater 18 is supplied back into the boiler 11.

Figure 2 is a block diagram showing the configuration of the plant operation optimization system of the first embodiment.

As shown in FIG. 2, in addition to the boiler 11 and generator 16 described above, the power plant 1 includes a sensor 41, an actuator 42, an operator input device 43, a control system 44, and a soot blowing system 45.

The plant operation optimization device 2 includes a boiler combustion optimization unit 21, a mill flow rate distribution optimization unit 22, an air flow rate distribution optimization unit 23, a soot blow optimization unit 24, an efficiency calculation unit 25, a boiler model management unit 26, an NN (neural network) processing unit 27, an LR (linear regression) processing unit 28, and an EKF (extended Kalman filter) processing unit 29. The data receiving function, data processing function, and data output function of the boiler combustion optimization unit 21, the mill flow rate distribution optimization unit 22, the air flow rate distribution optimization unit 23, and the soot blow optimization unit 24 are examples of a receiving unit, a determining unit, and an output unit, respectively.

The display device 3 includes a dashboard processing unit 31, a database 32, a simulation UI (user interface) processing unit 33, an efficiency UI processing unit 34, an operator UI processing unit 35, and a proposal UI processing unit 36.

The details of the plant operation optimization system of the first embodiment will be described below with reference to FIG. 2. In this description, FIG. 3 to FIG. 6 will also be referred to as appropriate. FIG. 3 is a graph for explaining soot blowing of the first embodiment. FIG. 4 is another graph for explaining soot blowing of the first embodiment. FIG. 5 is a diagram showing an example of an operator operation screen of the first embodiment. FIG. 6 is a diagram showing an example of a simulation screen of the first embodiment.

(1) Power Plant 1
The sensor 41 detects various values in the power plant 1 and outputs the detection results of these values. Examples of these values include the temperature, pressure, and flow rate of steam in the power plant 1, the temperature in the boiler 11 and the reheater 13, the oxygen concentration, and the steam spray flow rate. Another example of these values is the difference in physical quantities measured between the left and right parts of the furnace of the boiler 11. Another example of these values is the physical quantities of fluids other than steam (e.g., exhaust gas) in the power plant 1, and values related to the fuel and air supplied to the boiler 11. The values detected by the sensor 41 may be output to other devices in the power plant 1, or may be output to the plant operation optimization device 2 or the display device 3.

The actuator 42 is used to control the operation of various devices in the power plant 1. An example of the actuator 42 is a damper used to adjust the exhaust gas temperature of the boiler 11. Data indicating the state of the actuator 42 may be output to other devices in the power plant 1, or may be output to the plant operation optimization device 2 or the display device 3.

The operator input device 43 is used by an operator who controls the operation of the power plant 1. For example, the operator can use the operator input device 43 to perform operations related to soot blowing to remove soot accumulated in the boiler 11. The operator input device 43 may have a function to display values output from the sensor 41 and data output from the actuator 42.

The control system 44 controls various operations of the power plant 1. Examples of the control system 44 include a processor, an electric circuit, and a PC. For example, the control system 44 controls the start-up, operation, and shutdown of the power plant 1, and the opening and closing of steam valves in the power plant 1. The control system 44 may also control the operation of the power plant 1 based on values output from the sensor 41 and data output from the actuator 42, and may control the power plant 1 using the actuator 42 at this time. The data acquired or generated by the control system 44 may be output to the plant operation optimization device 2 or the display device 3.

The soot blowing system 45 performs soot blowing to remove soot accumulated in the boiler 11. In this embodiment, soot blowing is performed by a soot blower in the soot blowing system 45 supplying steam for soot blowing into the boiler 11. The soot adhering to the boiler 11 is removed by this steam. The steam for soot blowing is, for example, steam generated by the boiler 11. The soot blowing system 45 may perform soot blowing in response to manual operation from a device such as the operator input device 43, or may perform soot blowing automatically without manual operation. Data acquired or generated by the soot blowing system 45 may be output to the plant operation optimization device 2 or the display device 3.

(2) Plant operation optimization device 2
The plant operation optimization device 2 receives various data related to the boiler 11 in the power plant 1 from the power plant 1 and determines various setting values for controlling the boiler 11 based on the received data. For example, the plant operation optimization device 2 calculates setting values that physical quantities related to the boiler 11 should take in order to optimize the combustion state of the boiler 11, and calculates setting values that physical quantities related to soot blow should take in order to optimize soot blow in the boiler 11. The plant operation optimization device 2 further outputs the determined setting values to the display device 3. In this embodiment, the setting values are transmitted from the plant operation optimization device 2 to the display device 3 via a network.

These processes within the plant operation optimization device 2 are performed by the boiler combustion optimization unit 21, the mill flow rate distribution optimization unit 22, the air flow rate distribution optimization unit 23, and the soot blow optimization unit 24. The boiler combustion optimization unit 21, the mill flow rate distribution optimization unit 22, the air flow rate distribution optimization unit 23, and the soot blow optimization unit 24 perform different types of optimization processes, as described below.

The plant operation optimization device 2 of this embodiment manages existing set values already being used to control the boiler 11 as set values for various physical quantities. When the plant operation optimization device 2 calculates a new set value for a certain physical quantity, it may calculate the difference between the existing set value and the new set value for that physical quantity (this is called a "bias value") and output this bias value as the set value to the display device 3. For example, with regard to the set value of the fuel flow rate, if the existing set value is R1 and the new set value is R2, the plant operation optimization device 2 may output R2 to the display device 3, or may output ΔR (= R2 - R1) instead.

The following describes each block in the plant operation optimization device 2 in detail.

(2.1) Boiler combustion optimization unit 21
The boiler combustion optimization unit 21 determines a combustion state set value, which is a setting value for controlling the combustion state of the boiler 11, and outputs the determined combustion state set value to the display device 3. At this time, the boiler combustion optimization unit 21 may output to the display device 3 a bias value between the existing combustion state set value and the new combustion state set value.

The values of variables and parameters that represent the combustion state of the boiler 11 may deviate from the values of the design state. Examples of such variables and parameters include physical quantities such as the temperature, oxygen concentration, and steam spray flow rate inside the boiler 11, as well as the difference in physical quantities measured between the left and right parts of the furnace of the boiler 11. If such deviations are left unchecked, the combustion efficiency of the boiler 11 may decrease, the exhaust gas discharge conditions from the boiler 11 may not be met, and so on, and the combustion state of the boiler 11 may deteriorate.

The boiler combustion optimization unit 21 then determines a preferred combustion state set value for controlling the combustion state of the boiler 11, and outputs the determined combustion state set value to the display device 3. The operator of the power plant 1 can improve the combustion state of the boiler 11 by referring to the combustion state set value displayed on the display device 3. The operator may, for example, actually use this combustion state set value to control the boiler 11, or may use this combustion state set value to simulate the operation of the boiler 11.

The boiler combustion optimization unit 21 of this embodiment determines the combustion state setting value so as to improve the combustion state of the boiler 11, for example, so as to optimize the combustion efficiency of the boiler 11. This makes it possible, for example, to reduce the waste of fuel used in the boiler 11 and to make the operation of the boiler 11 more suitable both technically and economically.

The mode in which the combustion state set value and other set values are displayed on the display device 3 is called the monitoring mode. In the monitoring mode, bias values of various set values may be displayed on the display device 3. For example, an operator of the power plant 1 can know that the operating condition of the boiler 11 is deteriorating because of the presence of a bias value that is far from zero. The operator may manually change the existing set value used to control the boiler 11 by this bias value, or alternatively, the existing set value may be automatically changed by this bias value in an automatic control mode.

The boiler combustion optimization unit 21 performs such optimization processing, for example, during the trial operation of the power plant 1. During the trial operation of the power plant 1, for example, test coal is used as fuel for the boiler 11. In this case, after the power plant 1 starts commercial operation, the operating condition of the boiler 11 may become unsuitable due to changes in the operating conditions of the power plant 1 or the properties of the coal. In this case, the boiler combustion optimization unit 21 may perform the above-mentioned optimization processing even after the power plant 1 starts commercial operation.

The boiler combustion optimization unit 21 of this embodiment can perform optimization processing in cooperation with the boiler model management unit 26, the NN processing unit 27, and the LR processing unit 28. The boiler model management unit 26 manages a model for simulating the operation of the boiler 11. The boiler combustion optimization unit 21 may perform optimization processing using this model. The NN processing unit 27 performs information processing using a neural network. The LR processing unit 28 performs information processing using a linear regression algorithm. The boiler combustion optimization unit 21 can perform optimization processing using a neural network or a linear regression algorithm by working together with the NN processing unit 27 and the LR processing unit 28.

In general, optimizing the operating state of the boiler 11 requires adjustment of many variables and parameters. When such optimization is performed manually by a human being, only a small number of variables and parameters can be adjusted, such as adjusting the flow rate of air supplied to the boiler 11. On the other hand, even when such optimization is performed automatically by a computer, although a larger number of variables and parameters can be adjusted compared to manual adjustment, there is still a possibility that only a small number of variables and parameters can be adjusted.

The boiler combustion optimization unit 21 of this embodiment therefore uses a neural network to determine the combustion state set values for controlling the combustion state of the boiler 11. As a result, even when a large number of variables and parameters are adjusted, the operating state of the boiler 11 can be appropriately optimized by having the neural network learn such adjustments. According to this embodiment, by applying a neural network to the boiler combustion optimization unit 21, it becomes possible to determine a large number of combustion state set values, for example, by adjusting a large number of variables and parameters. The boiler combustion optimization unit 21 of this embodiment may operate to constantly optimize each combustion state set value using a neural network.

For example, when it is desirable to increase the exhaust gas temperature of the boiler 11, the neural network is made to learn the relationship between the actuator 42 (damper) used to adjust the exhaust gas temperature and the change in the combustion efficiency of the boiler 11. This allows the boiler combustion optimization unit 21 to obtain an optimal solution for the combustion efficiency of the boiler 11 from the NN processing unit 27. The boiler combustion optimization unit 21 outputs this optimal solution and the combustion state setting value (which may be a bias value) for this optimal solution to the display device 3. The operator of the power plant 1 can manually or automatically adjust the actuator 42 using this optimal solution and the combustion state setting value, thereby making it possible to increase the exhaust gas temperature while maintaining good combustion efficiency of the boiler 11.

The temperature of the exhaust gas from the boiler 11 can be changed by various methods. For example, the temperature of the exhaust gas from the boiler 11 can be changed by changing the flow rate of the fuel, changing the flow rate of the air, selecting whether the fuel coal is supplied in chunks or powder form, or selecting whether the fuel is supplied to the upper or lower part of the boiler 11. The neural network can learn the optimal combination of the manner in which these methods are used. Note that in this embodiment, physical quantities other than the temperature of the exhaust gas from the boiler 11 may also be subject to change, and a neural network may be used for this process.

In this embodiment, the operating state of the boiler 11 is optimized, for example, by minimizing a predetermined nonlinear function. This nonlinear function can minimize, for example, the combustion efficiency of the boiler 11, the coal consumption of the boiler 11, and the operating cost of the boiler 11. In this optimization process, it is possible to handle constraints such as the carbon dioxide emission amount, the nitrogen oxide emission amount, and the metal temperature of the boiler 11, and it is also possible to handle nonlinear constraints. The boiler combustion optimization unit 21 of this embodiment can perform optimization using feedforward, neural network, linear regression algorithm, etc., and can calculate solutions to objective functions and other functions. The weighting coefficient of the objective function is calculated, for example, after a bias test performed in advance in the power plant 1. The results of the optimization process by the boiler combustion optimization unit 21 may be stored in the plant operation optimization device 2 or in the display device 3.

The neural network of this embodiment operates based on multiple input values and outputs multiple output values to the boiler combustion optimization unit 21. The input values are, for example, values detected by the sensor 41. The output values are, for example, used as objective functions and constraint conditions in the optimization process by the boiler combustion optimization unit 21.

(2.2) Mill flow rate distribution optimization unit 22
The mill flow rate distribution optimization unit 22 determines a fuel set value, which is a setting value for controlling the fuel supplied to the boiler 11, and outputs the determined fuel set value to the display device 3. At this time, the mill flow rate distribution optimization unit 22 may output a bias value between the existing fuel set value and the new fuel set value to the display device 3.

The mill flow rate distribution optimization unit 22, for example, determines a fuel setting value so that the flow rate of fuel supplied to the boiler 11 is optimized. Examples of fuel setting values in this case are the setting value of the fuel flow rate itself and the setting value related to the operation of the actuator 42 that adjusts the fuel flow rate.

In the boiler 11 of this embodiment, fuel is supplied to the boiler 11 from multiple mills. The mill flow rate distribution optimization unit 22 of this embodiment may receive data on the flow rate of fuel supplied from these mills and data on the health of these mills, and may determine the fuel setting value of each mill based on the received data. The mill flow rate distribution optimization unit 22 of this embodiment may perform optimization processing so that the fireball center of the fuel in the boiler 11 is optimized with minimal setting changes. The results of the optimization processing by the mill flow rate distribution optimization unit 22 may be stored in the plant operation optimization device 2 or in the display device 3.

The system architecture of the plant operation optimization device 2 of this embodiment is divided into a first layer and a second layer. The first layer is a higher-level optimization level that supplies calculation results to the second layer, and includes a boiler combustion optimization unit 21 and an EKF processing unit 29 for the soot blow optimization unit 24 described later. The second layer includes a mill flow rate distribution optimization unit 22 and an air flow rate distribution optimization unit 23 described later. The plant operation optimization device 2 of this embodiment may determine various set values by optimization processing of the first and second layers, or may determine various set values by optimization processing of only the first layer depending on the purpose of optimization. For example, when optimizing oxygen concentration or burner tilt without changing fuel flow rate or air flow rate, the set values may be determined by optimization processing of only the first layer.

(2.3) Air flow rate distribution optimization unit 23
The air flow rate distribution optimization unit 23 determines an air set value, which is a set value for controlling the air supplied to the boiler 11, and outputs the determined air set value to the display device 3. At this time, the air flow rate distribution optimization unit 23 may output a bias value between the existing air set value and the new air set value to the display device 3.

The air flow rate distribution optimization unit 23, for example, determines an air set value so that the flow rate of air supplied to the boiler 11 is optimized. Examples of air set values in this case are the set value of the air flow rate itself and a set value related to the operation of the actuator 42 that adjusts the air flow rate.

In the boiler 11 of this embodiment, it is desirable to maintain the fuel-to-air ratio in order to maintain proper combustion. The air flow rate distribution optimization unit 23 of this embodiment performs optimization processing to maintain this ratio by, for example, secondary air damper control (SADC). The air flow rate distribution optimization unit 23 of this embodiment may receive data on the opening degree of multiple dampers and data on the soundness of these multiple dampers, and determine the air setting value of each damper based on the received data. Like the mill flow rate distribution optimization unit 22, the air flow rate distribution optimization unit 23 of this embodiment may also perform optimization processing so that the fireball center of the fuel in the boiler 11 is optimized with minimal setting changes. The results of the optimization processing by the air flow rate distribution optimization unit 23 may be stored in the plant operation optimization device 2 or in the display device 3.

(2.4) Soot blow optimization unit 24
The soot blow optimization unit 24 determines a soot blow setting value, which is a setting value for controlling soot blow in the boiler 11, and outputs the determined soot blow setting value to the display device 3. At this time, the soot blow optimization unit 24 may output to the display device 3 a new soot blow setting value to replace the existing soot blow setting value.

The soot blow optimization unit 24, for example, determines soot blow setting values so that the timing of performing soot blowing in the boiler 11 is optimized. Examples of soot blow setting values in this case include the start time, duration, and end time of performing soot blowing. The soot blow optimization unit 24 may output new times to replace the existing times for these times to the display device 3.

As described above, the boiler combustion optimization unit 21 performs optimization processing to optimize the combustion state of the boiler 11. This allows the boiler 11 to recover energy efficiently. However, if soot accumulates in the boiler 11, efficient energy recovery in the boiler 11 is hindered. Furthermore, if soot accumulates in the boiler 11, an error may occur in the optimization of the combustion state of the boiler 11 by the neural network. For this reason, it is desirable to perform soot blowing in the boiler 11.

Soot blowing can be performed at fixed intervals, for example, once every eight hours. However, when soot blowing is performed using steam, if soot blowing is performed when there is little soot, the steam will be wasted. For example, when steam generated by boiler 11 is used for soot blowing, if soot blowing is performed when there is little soot, fuel will be wasted in boiler 11. On the other hand, if soot blowing is not performed when there is a lot of soot, efficient energy recovery in boiler 11 will be hindered. In addition, when coal powder is used as fuel, the tendency for soot to accumulate becomes a problem.

Therefore, the soot blow optimization unit 24 of this embodiment is configured to be able to carry out optimization processing in cooperation with the EKF processing unit 29. The EKF processing unit 29 performs information processing using an extended Kalman filter. The soot blow optimization unit 24 can carry out optimization processing using an extended Kalman filter by working with the EKF processing unit 29. The soot blow optimization unit 24 of this embodiment can present the appropriate timing for carrying out soot blowing by determining the soot blowing setting value using an extended Kalman filter so that the timing for carrying out soot blowing is optimized. This makes it possible to suppress the reduction in energy recovery efficiency due to soot while suppressing waste of steam. The soot blow optimization unit 24 of this embodiment may operate to constantly optimize each soot blowing setting value using an extended Kalman filter.

The soot blow optimization unit 24 of this embodiment uses a soft sensor to estimate the accumulation of soot in the boiler 11 for the optimization process. For example, this soft sensor estimates the accumulation of soot in the furnace of the boiler 11 and each superheater (heat exchanger). The soft sensor of this embodiment includes three components: a thermodynamic model or a first principle model (1), an extended Kalman filter (2), and a feedback signal (3) from the sensor 41, which is a temperature sensor. The soot blow optimization unit 24 can determine a soot blow setting value that optimizes the timing of soot blowing by estimating the accumulation of soot using a soft sensor such as an extended Kalman filter. In this optimization process, for example, physical quantities (such as temperature) measured in the furnace of the boiler 11 and each superheater are used as input values. In addition, in this optimization process, the timing of soot blowing may be determined for each of multiple parts in the boiler 11.

The graph in Figure 3 shows the change over time in the thermal conduction efficiency of the boiler 11. The soot blow optimization unit 24 of this embodiment can obtain the thermal conduction efficiency estimated by an extended Kalman filter. In this case, the soot blow optimization unit 24 determines the soot blow setting value so that soot blowing is performed at the timing when the thermal conduction efficiency has fallen to a threshold value. This makes it possible to perform soot blowing at an appropriate timing, and makes it possible to suppress excessive decline in the thermal conduction efficiency of the boiler 11.

The graph in Figure 4 shows the relationship between the soot blow interval of the boiler 11 and the annual operating cost. According to this graph, the operating cost of the boiler 11 is at its highest when the soot blow interval is approximately 9 hours. Therefore, the soot blow optimization unit 24 may determine the soot blow setting value taking into account such operating cost characteristics. This makes it possible to optimize the timing of soot blowing while also taking into account costs.

(2.5) Efficiency Calculation Unit 25
The efficiency calculation unit 25 receives data related to the boiler 11 from the power plant 1 or the like, and calculates the efficiency of the boiler 11 based on the received data. The efficiency calculation unit 25 calculates, for example, the combustion efficiency of the boiler 11. The efficiency calculated by the efficiency calculation unit 25 may be output to the display device 3, or may be used in processing (e.g., optimization processing) by other blocks in the plant operation optimization device 2.

(3) Display device 3
In this embodiment, the plant operation optimization device 2 provides information for optimizing the operation of the power plant 1, and the display device 3 displays the information provided from the plant operation optimization device 2. As shown in Fig. 2, the display device 3 includes a dashboard processing unit 31, a database 32, a simulation UI processing unit 33, an efficiency UI processing unit 34, an operator UI processing unit 35, and a proposal UI processing unit 36. A GUI (Graphical User Interface) displayed on the screen of the display device 3 may be created by the display device 3 or may be created by the plant operation optimization device 2. In the latter case, the display device 3 may display the GUI by a browser.

The dashboard processing unit 31 displays a dashboard on the screen of the display device 3, and displays information provided by the plant operation optimization device 2 in the dashboard. The information displayed in the dashboard is, for example, various setting values output from the boiler combustion optimization unit 21, the mill flow rate distribution optimization unit 22, the air flow rate distribution optimization unit 23, or the soot blow optimization unit 24, and the efficiency output from the efficiency calculation unit 25 (for example, the combustion efficiency of the boiler 11). The setting value displayed in the dashboard may be the above-mentioned bias value or a new soot blow setting value that replaces the existing soot blow setting value. The dashboard processing unit 31 may display the dashboard on the simulation UI, efficiency UI, operator UI, or proposal UI described below.

The database 32 stores various information used by the display device 3. For example, information provided by the plant operation optimization device 2 is stored in the database 32.

The simulation UI processing unit 33 displays a simulation UI for simulating the operation of the power plant 1. The simulation UI processing unit 33 can simulate the operation of the power plant 1, for example, the operation of the boiler 11, based on the setting values and efficiency output from the plant operation optimization device 2 and information input on the simulation UI by the operator of the power plant 1. The simulation UI processing unit 33 can simulate the combustion state of the boiler 11, for example, based on the combustion state setting value. The results of the simulation performed by the simulation UI processing unit 33 are displayed on the simulation UI. This allows the operator to determine how to operate the boiler 11. The simulation UI is an example of a simulation screen.

The efficiency UI processing unit 34 displays an efficiency UI for providing information on the efficiency of the power plant 1, such as the combustion efficiency of the boiler 11. The efficiency output from the efficiency calculation unit 25 is displayed, for example, on the efficiency UI.

The operator UI processing unit 35 displays an operator UI for the operator of the power plant 1 to operate the power plant 1 and the plant operation optimization device 2. The operator UI may display various bias values output from the plant operation optimization device 2 or a new soot blow setting value that replaces the existing soot blow setting value. In this case, the operator UI processing unit 35 may transmit the changed bias value (or changed setting value) or the new soot blow setting value that replaces the existing soot blow setting value to the power plant 1 and the plant operation optimization device 2 by the operator changing the bias value or the new soot blow setting value that replaces the existing soot blow setting value from one value to another on the operator UI. This makes it possible to change the setting value used by the power plant 1 to control the boiler 11, or to change the setting value used by the plant operation optimization device 2 as the existing setting value. The operator UI is an example of an operator operation screen.

The proposal UI processing unit 36 displays a proposal UI that displays various setting values output from the plant operation optimization device 2 as proposed values (recommended values) from the display device 3. The proposal UI processing unit 36 may display all setting values output from the plant operation optimization device 2 as proposed values, or may display some of the setting values output from the plant operation optimization device 2 as proposed values. The operator of the power plant 1 can learn of the existence of suitable setting values by checking the proposal UI. The setting value displayed in the proposal UI may be a bias value or a new soot blow setting value that replaces the existing soot blow setting value.

Figure 5 shows an example of an operator UI. The operator UI shown in Figure 5 can display various bias values for the three mills A, B, and C and the secondary air damper control (SADC) or new soot blow setting values that replace existing soot blow setting values, and the operator can change these bias values or new soot blow setting values that replace existing soot blow setting values on the operator UI.

Figure 6 shows an example of a simulation UI. In the simulation UI shown in Figure 6, it is possible to specify the bias value to be used in the simulation and to view the simulation results. The bias value may be specified, for example, by specifying an upper or lower limit for the bias value. Figure 6 shows the simulation results for reheat temperature, nitrogen oxides, reheater spray, and metal temperature.

As described above, the plant operation optimization device 2 of this embodiment determines the set value (combustion state set value) for controlling the combustion state of the boiler 11 by a neural network, and determines the set value (soot blow set value) for controlling soot blowing in the boiler 11 by an extended Kalman filter. Therefore, according to this embodiment, it is possible to optimize the combustion state of the boiler 11 by taking into account various factors, and to optimize the timing for removing soot that hinders the optimization of the combustion state of the boiler 11, thereby making it possible to efficiently recover energy in the boiler 11.

The plant operation optimization device 2 of this embodiment may perform either one of determining the combustion state set value using a neural network or determining the soot blow set value using an extended Kalman filter, or may perform both.

Although several embodiments have been described above, these embodiments are presented only as examples and are not intended to limit the scope of the invention. The novel apparatus and method described in this specification can be embodied in various other forms. In addition, various omissions, substitutions, and modifications can be made to the forms of the apparatus and method described in this specification without departing from the spirit of the invention. The appended claims and their equivalents are intended to include such forms and modifications that fall within the scope and spirit of the invention.

1: power plant, 2: plant operation optimization device, 3: display device,
11: boiler, 12: high-pressure turbine, 13: reheater, 14: intermediate-pressure turbine,
15: low pressure turbine, 16: generator, 17: condenser, 18: feed water heater,
21: boiler combustion optimization unit, 22: mill flow rate distribution optimization unit,
23: air flow rate distribution optimization unit, 24: soot blow optimization unit,
25: efficiency calculation unit, 26: boiler model management unit,
27: NN processing unit, 28: LR processing unit, 29: EKF processing unit,
31: dashboard processing unit, 32: database,
33: Simulation UI processing unit, 34: Efficiency UI processing unit,
35: operator UI processing unit, 36: proposal UI processing unit,
41: sensor, 42: actuator, 43: operator input device,
44: Control system, 45: Soot blowing system

Claims (10)

a receiving unit for receiving data relating to a boiler in the power plant;
A determination unit that determines a combustion state set value, which is a set value for controlling the combustion state of the boiler, by a neural network based on data related to the boiler;
an output unit that outputs the combustion state setting value to a display device;
Equipped with
The determination unit further determines a fuel set value, which is a set value for controlling fuel supplied to the boiler, or an air set value, which is a set value for controlling air supplied to the boiler, based on the data related to the boiler;
The output unit further outputs the fuel setting value or the air setting value to the display device,
The determination unit is
Operate to determine the combustion condition setpoint and the fuel setpoint or the air setpoint such that a combustion efficiency of the boiler and a flow rate of the fuel or air supplied to the boiler are both optimized;
Operate to determine the combustion condition setpoint such that the combustion efficiency of the boiler is optimized without changing the flow rate of fuel or air supplied to the boiler;
This is a plant operation optimization device. 2. The plant operation optimization device according to claim 1 , wherein the combustion state set value is a bias value between an existing set value and a new set value of a physical quantity representing the combustion state of the boiler. The determination unit further determines a soot blow setting value, which is a setting value for controlling soot blow in the boiler, by an extended Kalman filter based on data related to the boiler,
The output unit further outputs the soot blow setting value to the display device.
The plant operation optimization device according to claim 1 or 2 . The determination unit is
Operate to determine the soot blowing setpoint and the fuel setpoint or the air setpoint such that the timing of performing soot blowing in the boiler and the flow rate of fuel or air supplied to the boiler are both optimized ;
Operate to determine the soot blowing setpoint to optimize timing of soot blowing within the boiler without changing the flow rate of fuel or air supplied to the boiler;
The plant operation optimization device according to claim 3 , wherein the above is possible . 5. The plant operation optimization apparatus according to claim 3 , wherein the soot blowing set value is a new set value replacing an existing set value for the time for performing soot blowing in the boiler. The plant operation optimization device according to claim 1 , wherein the output unit transmits the combustion condition set value to the display device different from the plant operation optimization device. 7. The plant operation optimization device according to claim 1 , wherein the display device displays an operator operation screen for operating the power plant or the plant operation optimization device. 8. The plant operation optimization device according to claim 1 , wherein the display device displays a simulation screen for simulating an operation of the boiler based on the combustion condition set value. a receiving unit for receiving data relating to a boiler in the power plant;
A determination unit that determines a soot blow setting value, which is a setting value for controlling soot blow in the boiler, based on data related to the boiler by an extended Kalman filter;
An output unit that outputs the soot blow setting value to a display device;
Equipped with
The determination unit further determines a fuel set value, which is a set value for controlling fuel supplied to the boiler, or an air set value, which is a set value for controlling air supplied to the boiler, based on the data related to the boiler;
The output unit further outputs the fuel setting value or the air setting value to the display device,
The determination unit is
Operate to determine the soot blowing setpoint and the fuel setpoint or the air setpoint such that the timing of performing soot blowing in the boiler and the flow rate of fuel or air supplied to the boiler are both optimized;
Operate to determine the soot blowing setpoint to optimize timing of soot blowing within the boiler without changing the flow rate of fuel or air supplied to the boiler;
This is a plant operation optimization device. Receive data about boilers in the power plant;
Based on the data regarding the boiler, a combustion state set value which is a set value for controlling the combustion state of the boiler is determined by a neural network, or a soot blow set value which is a set value for controlling soot blow in the boiler is determined by an extended Kalman filter;
The combustion state setting value or the soot blow setting value is output to a display device.
Including,
determining a fuel set value, which is a set value for controlling fuel supplied to the boiler, or an air set value, which is a set value for controlling air supplied to the boiler, based on the data related to the boiler;
outputting the fuel setpoint or the air setpoint to the display device;
Further including:
Determining the combustion condition set value or the soot blowing set value and the fuel set value or the air set value so that the combustion efficiency of the boiler or the timing of performing soot blowing in the boiler and the flow rate of fuel or air supplied to the boiler are both optimized;
determining the combustion condition setpoint or the soot blowing setpoint so that the combustion efficiency of the boiler or the timing of performing soot blowing in the boiler is optimized without changing the flow rate of fuel or air supplied to the boiler;
The method for optimizing plant operations further comprises :

Images