JP7250206B1 - Information processing method, information processing device and program - Google Patents

Abstract

An object of the present invention is to present an input amount of fuel gas required to achieve an adjustment target. A computer obtains measured values of state variables of a furnace in which at least one of fuel and waste is combusted and a current fuel input, and corresponds to the obtained measured values and input. Provided is an information processing method for determining a recommended input amount for adjusting a variable to an adjustment target, and presenting the determined recommended input amount on an operation screen. [Selection drawing] Fig. 3

Description

The present invention relates to an information processing method, an information processing apparatus, and a program.

The operation and maintenance of industrial combustion furnaces has long been the responsibility of field workers. However, individual adjustment by workers causes problems in skill inheritance in addition to variations in accuracy. In addition, adjustment by workers is basically subjective. Therefore, it is not known whether the current adjustment is the optimum solution.

JP 2019-020066 A JP 2008-249214 A

Conditions in industrial combustion furnaces change from moment to moment. For example, the amount and type of waste (hereinafter referred to as "state variables") charged into the furnace may fluctuate depending on the emission source. In addition, the relationship between the fuel gas and air (hereinafter referred to as "control variables") and the state variables is complex, and the gas discharged from the combustion furnace as a result of combustion (hereinafter referred to as "exhaust gas"). It is difficult to accurately predict the concentration of , the amount of flame radiation, the surface temperature of the object to be heated, etc. (hereinafter referred to as "objective variables"). As a result, it is difficult to determine the amount of fuel gas required to achieve the adjustment target.

An object of the present invention is to present the amount of fuel gas required to achieve an adjustment target.

In the invention according to claim 1, a computer obtains measured values of state variables of a furnace in which at least one of fuel and waste is combusted and a current amount of fuel input, and the obtained measured values and the An information processing method for determining a recommended input amount for adjusting an objective variable corresponding to an input amount to an adjustment target, and presenting the determined recommended input amount on an operation screen, wherein the input amount and the measured value is input and the output is the predicted value of the objective variable corresponding to the input amount and the measured value. A relationship in which the predicted value of the objective variable and the measured value of the state variable are input, and the recommended input amount for adjusting the predicted value of the objective variable to the adjustment target is output. The predicted value and the measured value are input to the learned second learning model to determine the recommended input amount, and when a change in the combination of the fuel and the waste to be input to the furnace is detected, the change A method of information processing, wherein the first learning model trained for subsequent combinations is used to predict the predicted value of the objective variable.
According to the second aspect of the invention, the first learning model used for predicting the predicted value is based on the difference between the number of nozzles used for inputting the waste and the position of the nozzle used for inputting the waste. 2. The information processing method according to claim 1, wherein switching is performed according to at least one of them.
In the invention according to claim 3, the computer acquires the measured value of the state variable of the furnace in which at least one of the fuel and the waste is combusted and the current amount of fuel input, and the acquired measured value and the An information processing method for determining a recommended input amount for adjusting an objective variable corresponding to an input amount to an adjustment target, and presenting the determined recommended input amount on an operation screen, wherein the input amount of fuel and the using a search key given by the measured value of the state variable to retrieve the operating data of the furnace including the combination of the input amount, the measured value, and the measured value of the objective variable, and the measured value of the objective variable; A method for information processing, wherein the recommended input amount of the fuel is determined based on search results that adjust the value to the adjustment target.
In the invention according to claim 4, when there is no search result corresponding to the search key, a plurality of search results having a high degree of similarity with the search key are used to make the recommendation for adjusting to the adjustment target. 4. The information processing method according to claim 3, wherein an input amount is calculated.
In the invention according to claim 5, the computer acquires the measured value of the state variable of the furnace in which at least one of the fuel and the waste is combusted and the current amount of fuel input, and the acquired measured value and the and determining a recommended input amount for adjusting an objective variable corresponding to an input amount to an adjustment target, and presenting the determined recommended input amount on an operation screen, wherein the adjustment target is a plurality of variables. When described as a function by, the predicted values corresponding to the measured values of the plurality of variables are input and the recommended input amount for adjusting the predicted values to the adjustment target is learned. It is an information processing method of inputting the predicted value to the learning model and determining the recommended input amount.
The invention according to claim 6 comprises an acquisition unit that acquires measured values of state variables of a furnace in which at least one of fuel and waste is combusted and the current amount of fuel input, and the acquired measured values and the an information processing apparatus comprising: a determination unit that determines a recommended input amount for adjusting an objective variable corresponding to an input amount to an adjustment target; and a presentation unit that presents the determined recommended input amount on an operation screen. , the input amount and the measured value are input, and the input amount and the measured value are applied to the first learning model that has learned the relationship in which the output is the predicted value of the objective variable corresponding to the input amount and the measured value. predicting the predicted value of the objective variable by inputting, using the predicted predicted value of the objective variable and the measured value of the state variable as inputs, the recommendation for adjusting the predicted value of the objective variable to the adjustment target Inputting the predicted value and the measured value to a second learning model that has learned a relationship in which the input amount is the output, determines the recommended input amount, and combines the fuel and the waste to be input into the furnace. is detected, the first learning model learned about the combination after the change is used for predicting the predicted value of the objective variable.
The invention according to claim 7 comprises an acquisition unit for acquiring measured values of state variables of a furnace in which at least one of fuel and waste is combusted and the current amount of fuel input, and the acquired measured values and the an information processing apparatus comprising: a determination unit that determines a recommended input amount for adjusting an objective variable corresponding to an input amount to an adjustment target; and a presentation unit that presents the determined recommended input amount on an operation screen. , using a search key given by the input amount of fuel and the measured value of the state variable, the operating data of the reactor including the combination of the input amount, the measured value, and the measured value of the objective variable; The information processing device searches for and determines the recommended injection amount of the fuel based on a search result of adjusting the measured value of the objective variable to the adjustment target.
According to an eighth aspect of the invention, an acquisition unit acquires measured values of state variables of a furnace in which at least one of fuel and waste is combusted and the current amount of fuel input, and the acquired measured values and the an information processing apparatus comprising: a determination unit that determines a recommended input amount for adjusting an objective variable corresponding to an input amount to an adjustment target; and a presentation unit that presents the determined recommended input amount on an operation screen. , when the adjustment target is described as a function of a plurality of variables, the input is a predicted value corresponding to the measured values of the plurality of variables, and the recommended input amount for adjusting the predicted value to the adjustment target is output. It is an information processing device that determines the recommended input amount by inputting the predicted value to a third learning model that has learned the relationship of .
According to the ninth aspect of the invention, the computer is provided with a function of acquiring measured values of state variables of a furnace in which at least one of fuel and waste is combusted and the current amount of fuel input, and A program for realizing a function of determining a recommended input amount for adjusting an objective variable corresponding to the input amount to an adjustment target, and a function of presenting the determined recommended input amount on an operation screen. The input amount and the measured value are input, and the output is the predicted value of the objective variable corresponding to the input amount and the measured value for the first learning model that learned the input amount and the measured value to predict the predicted value of the objective variable, input the predicted predicted value of the objective variable and the measured value of the state variable, and adjust the predicted value of the objective variable to the adjustment target. The predicted value and the measured value are input to the second learning model that has learned the relationship of outputting the recommended input amount to determine the recommended input amount, and the fuel and the waste to be input to the reactor are determined. In the program, when a change in combination is detected, the first learning model learned for the combination after the change is used to predict the predicted value of the objective variable.
According to the invention of claim 10, the computer has a function of acquiring measured values of state variables of a furnace in which at least one of fuel and waste is combusted and the current amount of fuel input, and the acquired measured value A program for realizing a function of determining a recommended input amount for adjusting an objective variable corresponding to the input amount to an adjustment target, and a function of presenting the determined recommended input amount on an operation screen. and using a search key given by the fuel input amount and the measured value of the state variable, operating data of the reactor including a combination of the input amount, the measured value, and the measured value of the objective variable. and adjusting the measured value of the objective variable to the adjustment target to determine the recommended injection amount of the fuel based on the search results.
The invention according to claim 11 is characterized in that the computer has a function of acquiring measured values of state variables of a furnace in which at least one of fuel and waste is combusted and the current amount of fuel input, and the acquired measured value A program for realizing a function of determining a recommended input amount for adjusting an objective variable corresponding to the input amount to an adjustment target, and a function of presenting the determined recommended input amount on an operation screen. and when the adjustment target is described as a function of a plurality of variables, the predicted values corresponding to the measured values of the plurality of variables are input, and the recommended input amount for adjusting the predicted values to the adjustment target is It is a program for determining the recommended input amount by inputting the predicted value to the third learning model that has learned the relationship to be the output.
According to a twelfth aspect of the invention, the computer is a furnace in which waste is combusted using fuel gas, and the state variable of the furnace in which at least one of the fuel gas and waste is combusted is measured according to the period during combustion. and the current input amount of fuel gas , and set the recommended input amount of fuel gas for adjusting the objective variable corresponding to the obtained measured value and input amount to the adjustment target. is determined using a learning model learned using the predicted value of , the measured value, and the recommended input amount as teacher data, and the determined recommended input amount is presented on an operation screen.
The invention according to claim 13 is the information processing method according to claim 12 , further comprising a process of presenting the measured value used for determining the recommended input amount on the operation screen.
The invention according to claim 14 is the information processing method according to claim 12 , further comprising a process of presenting, on the operation screen, a predicted value of the objective variable after adjustment according to the presented recommended input amount.
The invention according to claim 15 is the information processing method according to claim 12 , wherein the waste includes at least one of liquid waste, waste oil, incineration matter, and process exhaust gas.
According to the sixteenth aspect of the invention, the state variable is one of flow rate of waste liquid, flow rate of waste oil, flow rate of process exhaust gas, amount of incinerated matter, oxygen concentration in the furnace, temperature in the furnace, and temperature of the furnace wall. 13. The information processing method according to claim 12 , comprising at least one of
In the invention according to claim 17 , the objective variables are CO concentration, CO ₂ concentration, NO _x concentration, SO _x concentration, dust concentration, dioxin concentration, unburned fuel gas concentration, greenhouse 13. The information processing method according to claim 12 , comprising at least one of the concentration of the effect gas, the amount of flame radiation, and the surface temperature of the object to be heated.
The invention according to claim 18 is the information processing method according to claim 12 , further comprising a process of controlling the opening degree of the valve corresponding to the fuel gas based on the determined recommended input amount.
In the invention according to claim 19 , the input amount and the measured value are input, and the predicted value of the objective variable corresponding to the input amount and the measured value is output. Predicting the predicted value of the objective variable by inputting the input amount and the measured value, inputting the predicted predicted value of the objective variable and the measured value of the state variable, and adjusting the predicted value of the objective variable 13. The recommended input amount is determined by inputting the predicted value and the measured value to the second learning model that has learned the relationship of outputting the recommended input amount for adjusting to the target. It is an information processing method.
The invention according to claim 20 is a furnace that burns waste using fuel gas, and the measured value of the state variable of the furnace in which at least one of the fuel gas and the waste is burned and the current and an acquisition unit for acquiring the input amount of fuel gas and the recommended input amount of fuel gas for adjusting the objective variable corresponding to the obtained measured value and input amount to the adjustment target , the objective variable A determination unit that determines the predicted value of, the measured value, and the recommended input amount using a learning model learned using teacher data, and a presentation unit that presents the determined recommended input amount on an operation screen. It is an information processing device.
The invention according to claim 21 is a furnace for burning waste using fuel gas, wherein at least one of the fuel gas and the waste is burned according to the period during combustion, and the state variable of the furnace is measured. and a function of acquiring the current input amount of fuel gas , and the recommended input amount of the fuel gas for adjusting the target variable corresponding to the obtained measured value and the input amount to the adjustment target . A function of determining the predicted value of the objective variable, the measured value, and the recommended input amount using a learning model learned using teacher data, and a function of presenting the determined recommended input amount on the operation screen. It is a program for

According to the first aspect of the invention, it is possible to improve the accuracy of determination of the recommended input amount of fuel gas required to achieve the adjustment target.
According to the second aspect of the invention, it is possible to improve the accuracy of determination of the recommended input amount of fuel gas required to achieve the adjustment target.
According to the third aspect of the invention, it is possible to improve the accuracy of determining the recommended input amount of fuel gas required to achieve the adjustment target without preparing a learning model.
According to the fourth aspect of the invention, it is possible to determine the recommended injection amount of fuel gas required to achieve the adjustment target even if there is no operation data corresponding to the search key.
According to the fifth aspect of the invention, even when the adjustment target is described as a function of a plurality of objective variables, it is possible to determine the recommended injection amount of fuel gas required to achieve the adjustment target.
According to the sixth aspect of the invention, it is possible to improve the accuracy of determination of the recommended input amount of fuel gas required to achieve the adjustment target.
According to the seventh aspect of the invention, it is possible to improve the accuracy of determining the recommended input amount of fuel gas required to achieve the adjustment target without preparing a learning model.
According to the eighth aspect of the invention, even if the adjustment target is described as a function of a plurality of objective variables, it is possible to determine the recommended injection amount of fuel gas required to achieve the adjustment target.
According to the ninth aspect of the invention, it is possible to improve the accuracy of determination of the recommended input amount of fuel gas required to achieve the adjustment target.
According to the tenth aspect of the invention, it is possible to improve the accuracy of determining the recommended input amount of fuel gas required to achieve the adjustment target without preparing a learning model.
According to the eleventh aspect of the invention, even when the adjustment target is described as a function of a plurality of objective variables, it is possible to determine the recommended injection amount of fuel gas required to achieve the adjustment target.
According to the twelfth aspect of the invention, it is possible to present the recommended input amount of fuel gas required to achieve the adjustment target.
According to the thirteenth aspect of the present invention, it is possible to check the current state of the furnace in addition to the recommended input amount of fuel gas required to achieve the adjustment target.
According to the fourteenth aspect of the invention, it is possible to confirm the result of adjustment in advance.
According to the fifteenth aspect of the invention, it is possible to improve the accuracy of determining the recommended input amount of fuel gas required to achieve the adjustment target.
According to the sixteenth aspect of the invention, it is possible to improve the accuracy of determining the recommended input amount of fuel gas required to achieve the adjustment target.
According to the seventeenth aspect of the invention, it is possible to improve the accuracy of determining the recommended input amount of fuel gas required to achieve the adjustment target.
According to the eighteenth aspect of the invention, it is possible to automatically control the injection of the fuel gas required to achieve the adjustment target.
According to the nineteenth aspect of the invention, it is possible to improve the accuracy of determining the recommended input amount of fuel gas required to achieve the adjustment target.
According to the twentieth aspect of the invention , it is possible to present the recommended input amount of fuel gas required to achieve the adjustment target.
According to the twenty-first aspect of the invention, it is possible to present the recommended input amount of fuel gas required to achieve the adjustment target.

1 is a diagram illustrating a conceptual configuration of a combustion furnace system assumed in Embodiment 1; FIG. 1 is a diagram illustrating a configuration example of an information presentation device used in Embodiment 1; FIG. 2 is a diagram illustrating a functional configuration example of an information presentation device used in Embodiment 1; FIG. 4 is a diagram for explaining the processing contents of an input amount determination unit used in Embodiment 1. FIG. 4 is a flowchart for explaining the procedure for generating a learning model that uses the concentration of exhaust gas as a predicted value; FIG. 10 is a diagram explaining a method of correcting a learning model using a gradient boosting decision tree; 4 is a flowchart for explaining a procedure for generating a learning model; FIG. 10 is a flowchart for explaining a method of correcting a learning model using teacher data generated by interpolation; FIG. It is a figure explaining the generation method of teacher data. (A) shows a method of increasing the number of teacher data samples using linear interpolation, and (B) shows a method of increasing the number of teacher data samples using a polynomial or regression model. 4 is a flowchart for explaining an example of an operation for presenting a recommended value of fuel gas that achieves an adjustment target; FIG. 11 is a flow chart illustrating another example of the operation of presenting a recommended value of fuel gas that achieves an adjustment target; FIG. It is an example of a screen presenting a recommended input amount of fuel gas when the adjustment target is to minimize the CO concentration. It is an example of a screen presenting a recommended input amount of fuel gas when the adjustment target is to adjust the temperature in the furnace. It is an example of a screen presenting a recommended injection amount of fuel gas when the adjustment target is to minimize the amount of fuel gas injected into the combustion furnace. It is an example of a screen presenting a recommended input amount of fuel gas when the adjustment target is maximization of heated efficiency. FIG. 11 is an example of a screen presenting a recommended input amount of fuel gas when the adjustment target is improvement of flame radiation; FIG. FIG. 4 is a diagram illustrating an example of variables used for generating a learning model; FIG. FIG. 10 is a diagram explaining another example of variables used for generating a learning model; 2 is a diagram illustrating combustion patterns assumed in one combustion furnace; FIG. 12 is a flow chart for explaining learning model generation processing according to Embodiment 3. FIG. FIG. 4 is a diagram for explaining a function of classifying teacher data; FIG. 10 is a diagram illustrating a learning function of a dedicated learning model corresponding to each pattern using classified teacher data; FIG. 12 is a flowchart for explaining learning model switching processing according to Embodiment 3. FIG. FIG. 4 is a diagram illustrating a case where one nozzle is provided for each combustible material on the ceiling of the combustion furnace; FIG. 4 is a diagram illustrating a case where two nozzles are provided for each combustible material on the ceiling of the combustion furnace; FIG. 10 is a diagram illustrating another example in which a total of two nozzles are provided for each combustible material distributed in the ceiling portion and the wall portion of the combustion furnace; 4 is a flowchart for explaining learning model switching processing when the number and mounting positions of nozzles used for throwing in waste change during combustion. 4 is a flowchart for explaining a learning model setting process when the number and mounting positions of nozzles used for throwing in waste do not change during combustion. FIG. 10 is a flowchart for explaining learning model setting processing in consideration of a control time lag; FIG. 4 is a chart for explaining an example in which measured values such as an input amount of combustible material input at a reference time (current time) and measured values of concentration of exhaust gas after a certain period of time from the reference time are used as teacher data. FIG. 4 is a chart for explaining an example in which a measured value of exhaust gas concentration at a reference time (current time) and a measured value of an input amount of combustible material, etc., input a certain time before the reference time are used as teacher data. FIG. 12 is a diagram illustrating a configuration example of an information presentation device used in Embodiment 6; It is a figure explaining the example of driving data. FIG. 21 is a diagram for explaining a functional configuration example of an information presentation device used in Embodiment 6; It is a flowchart explaining the determination processing of the input amount by an input amount determination part. FIG. 22 is a diagram for explaining the processing contents of an input amount determination unit used in Embodiment 8; FIG. 12 is a diagram for explaining a conceptual configuration of a combustion furnace system assumed in an eighth embodiment;

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described with reference to the drawings.
<Embodiment 1>
<System configuration>
In this embodiment, a combustion furnace for burning waste will be described. In particular, a combustion furnace in which the input amount of waste and the combination and ratio of the types of waste to be input change from moment to moment according to the convenience of the emission source will be described.

In this type of combustion furnace, the state inside the furnace changes from moment to moment. For this reason, it is much more difficult to predict the exhaust gas and the like discharged from the combustion furnace than in the case of the combustion furnace in which the amount of waste input is controlled to be constant. Of course, the concentration of the exhaust gas discharged from the combustion furnace can be confirmed by measurement, but adjustment based on the measured value is a follow-up adjustment, and the result of the adjustment cannot be predicted in advance. For this reason, adjustments must be made on a trial-and-error basis. Moreover, since the conditions inside the furnace change from moment to moment, it is difficult to determine the details of the next adjustment with reference to the results of past adjustments.

In this way, it is necessary to adjust the combustion state of the combustion furnace in which the amount of waste input and the combination and ratio of types of waste to be input change from moment to moment depending on the emission source, so as to achieve the adjustment target. is difficult.
Therefore, in the present embodiment, even in a combustion furnace in which the state inside the furnace changes from moment to moment, a mechanism will be described that can present the optimum input amount of the fuel gas required to achieve the adjustment target and the like.

<Example of adjustment target>
The adjustment target of the combustion furnace assumed in this embodiment varies depending on the operator. Examples of typical adjustment targets are shown below.
(1) Adjustment target 1:
Minimize the concentration of specific gases discharged from the combustion furnace. The adjustment goal is to minimize the combustion environment at each point in time, and the minimum concentration is not always given in advance.
In this embodiment, the specific gas contained in the exhaust gas discharged from the combustion furnace is called "specific gas". The specific gas here includes, for example, CO, _CO2 , _NOx , _SOx , dust, dioxin, unburned fuel gas, greenhouse gas, hydrocarbon gas, and the like.

(2) Adjustment target 2:
Control the furnace temperature to the target temperature. As for the temperature inside the furnace, there are cases where it is desired to maintain the temperature inside the furnace at a target temperature even in a combustion furnace in which the amount of material to be burned in the furnace changes from moment to moment. The target temperature here may be designated as a specific temperature value such as 500.degree. C., or may have a range such as 500.degree.
(3) Adjustment target 3:
Minimize fuel gas input. It should be noted that the minimum input amount is not necessarily given in advance because the input amount of fuel gas required for combustion differs if the input amount of wastes or the like is different. Such adjustments are necessary, for example, when minimizing fuel costs is a priority. The fuel gas here includes, for example, city gas, mixed gas of ammonia and hydrogen, and the like.

(4) Adjustment target 4:
Maximize the combustion efficiency of the combustion furnace. Maximizing combustion efficiency also helps reduce fuel costs and harmful emissions.
(5) Adjustment target 5:
Improves flame radiation. In a combustion furnace for indirectly heating an object to be heated, improvement in flame radiation is required.
In this way, there are five adjustment targets in the representative example alone, but the embodiment will be described from the viewpoint of adjustment target 1 below.

FIG. 1 is a diagram illustrating a conceptual configuration of a combustion furnace system 1 assumed in Embodiment 1. FIG.
In the case of FIG. 1, the outline indicated by the dashed line represents the outer edge of the building or site where the combustion furnace 10 and the monitoring room are installed. An information presentation device 20 and a monitor 30 are arranged in the monitoring room.
Although FIG. 1 shows only one combustion furnace 10, it is not intended to limit the number of combustion furnaces 10 in the building or site to one. Also, in FIG. 1, there is one information presentation device 20 and one monitor 30, but there may be a plurality of them.

In the combustion furnace 10 shown in FIG. 1, waste (e.g., process exhaust gas, waste oil, waste liquid) and fuel gas (city gas, air) are introduced as materials to be burned, and exhaust gas is discharged as a result of the combustion reaction. be.
City gas and air are fuels for burning waste. City gas is, for example, natural gas whose main component is methane.

Instead of city gas, for example, liquefied petroleum gas (so-called LPG) containing propane or butane as a main component may be introduced. City gas and LPG are representative examples of hydrocarbon fuels.
In addition, ammonia gas, a mixed gas of ammonia and city gas, a mixed gas of ammonia and hydrogen, or the like may be fed into the combustion furnace 10 as the fuel gas.
The city gas flow rate Q _N [Nm ³ /h] and the air flow rate Q _A [Nm ³ /h] are examples of control variables.

Waste refers to substances that are by-products of manufacturing processes and chemical reaction processes and that are subject to disposal. Waste may be discharged as a gas, as a liquid, or as a solid.
A representative example of waste as a gas is process exhaust gas. Typical examples of liquid waste include waste oil and waste liquid.

Waste oil refers to a liquid containing oil.
Wastewater is the liquid recovered from each process. In the present embodiment, liquid that does not contain oil is referred to as waste liquid. The waste liquid includes contaminated water after being used for cooling and cleaning, as well as organic liquids and the like. It also includes mixtures of liquids discharged from multiple processes separately.
In this embodiment, the liquid combusted in the combustion furnace 10 may be generically called "waste liquid". Liquid waste in this sense also includes oil-containing liquids.
Solid waste includes dust, shavings, scraps, and the like.

In addition, the flow rate of process exhaust gas Qp [Nm ³ /h], the flow rate of waste oil _QH [Nm ³ /h], the flow rate of waste liquid _QD [Nm ³ /h], the oxygen concentration in the furnace X _O [%], the furnace The internal temperature T [°C] and the furnace wall temperature T [°C] are examples of variables representing the state of the furnace (that is, “state variables”). The state variable here is an example of "state information" in the claims.
Incidentally, the weight of waste as a state variable is represented by [ton/h].
Each flow rate is measured in real time by a flow meter, oxygen concentration by an oxygen concentration meter, temperature by a temperature sensor, and weight by a weight scale.

Exhaust gas varies depending on the substance burned in the combustion furnace 10, but includes, for example, CO, _CO2 , NOx, _SOx , dust, _dioxin , unburned fuel gas, greenhouse gas, hydrocarbon gas, and the like.
In addition, the information representing the results of the combustion reaction includes the amount of flame radiation [W/m ² ] and the surface temperature T [°C] of the object to be heated. For example, the concentration of exhaust gas, the amount of flame radiation, and the surface temperature of the object to be heated are examples of objective variables. In addition, the objective variable includes the heating efficiency [%].

Concentration is measured in real time by a densitometer, flame radiation and heating efficiency are calculated, and surface temperature is measured by a temperature sensor. The heating efficiency is calculated using, for example, the amount of heat input to the combustion furnace 10 and the surface temperature of the object to be heated.
However, as far as the combustion furnace 10 assumed in the present embodiment is concerned, there is no object to be heated. Therefore, the surface temperature of the object to be heated is measured when the combustion furnace 10 is used as a heating furnace or a melting furnace. Objects to be heated include, for example, glass, iron, aluminum oxide, and ash.

<Device configuration>
FIG. 2 is a diagram illustrating a configuration example of the information presentation device 20 used in the first embodiment. The information presentation device 20 is an example of an “information processing device”.
The information presentation device 20 includes a processor 201 that controls the overall operation of the device, a ROM 202 that stores BIOS and the like, a RAM 203 that is used as a work area for the processor 201, an auxiliary storage device 204 that records various data, a monitor, and a monitor. 30 (see FIG. 1) and an input device (for example, a keyboard, a mouse) not shown, and an I/O interface 205 used for connection, and communication to communicate with sensors installed in each part of the combustion furnace 10 (see FIG. 1) and an interface 206 . Note that the processor 201 and other processing units are interconnected through a bus or other signal line 207 .

The processor 201, ROM 202, and RAM 203 function as a so-called computer. That is, the information presentation device 20 is configured by, for example, a server, a desktop computer, or a notebook computer.
The processor 201 implements various functions through execution of programs.

In this embodiment, as one of the programs, each measured value of the control variable, the state information, and the objective variable is used as teacher data, and the objective variable is predicted when the control variable and the status information are given as inputs. Consider a program that generates a predictive model that outputs values.
As one of the other programs, a program is assumed in which measured values of control variables and state information acquired from sensors provided in each part of the combustion furnace 10 are input to a prediction model, and predicted values of objective variables are output. .

As one of the other programs, each measured value of the control variable, the state information, and the objective variable is used as teacher data, and the objective variable is adjusted to the adjustment target when the state information and the objective variable are given as inputs. Consider a program that generates a control model that outputs recommended values for control variables.
As one of the other programs, it is assumed that the state information obtained from the sensors provided in each part of the combustion furnace 10 and the measured values of the objective variables are input to the control model, and the recommended values of the control variables are output. .

The auxiliary storage device 204 is, for example, a hard disk device or a semiconductor memory. Auxiliary storage device 204 in the present embodiment stores prediction model 204A and control model 204B.
The prediction model 204A is a learning model that learns an input-output relationship that outputs a predicted value of an objective variable when a measured value of a control variable and a measured value of a state variable are input. This prediction model 204A is an example of a "first learning model" in the claims.

The control model 204B is a learning model that learns the input/output relationship for outputting the recommended value of the control variable that achieves the adjustment target when the predicted value of the objective variable and the measured value of the state variable are input. This control model 204B is an example of a "second learning model" in the claims.
Prediction model 204A and control model 204B in the present embodiment are prepared in advance by a program provider or the like according to the embodiment.
Information presentation device 20 in the present embodiment uses these learning models to present recommended values of fuel gas as control variables to monitor 30 .

<Functional configuration of information presentation device>
FIG. 3 is a diagram illustrating a functional configuration example of the information presentation device 20 used in the first embodiment.
The processor 201 (see FIG. 2) functions as a measurement data acquisition unit 211, an input amount determination unit 212, and an input amount presentation unit 213 through execution of programs.
The measurement data acquisition unit 211 is a functional unit that acquires measurement data from sensors installed in each part of the combustion furnace 10 (see FIG. 1).

In this embodiment, the measured value of the state information and the current fuel gas input amount are acquired as the measured data. The state information here includes the flow rate of waste liquid, the flow rate of waste oil, the amount of incinerated matter, the flow rate of process exhaust gas, the temperature inside the furnace, the temperature of the furnace wall, and the oxygen concentration inside the furnace. In addition, the input amount of fuel gas is the flow rate of city gas and the flow rate of air.
The measurement data acquisition unit 211 here is an example of the “acquisition unit” in the claims.

The input amount determination unit 212 is a functional unit that determines the fuel gas input amount (that is, the recommended input amount) recommended for adjusting the objective variable corresponding to the measured value of the state information and the fuel gas input amount to the adjustment target. is.
The input amount determination unit 212 here is an example of the "determination unit" in the scope of claims.
FIG. 4 is a diagram for explaining the processing contents of the input amount determination unit 212 used in the first embodiment.
As shown in FIG. 4, the input amount determination unit 212 uses an objective variable prediction unit 212A that predicts the predicted value of the objective variable in the current combustion environment, and an adjustment target using the predicted value of the objective variable and the measured value of the state variable. and a control variable determination unit 212B that determines the recommended value of the control variable for realizing the above.

The objective variable prediction unit 212A inputs the measured value of the control variable and the measured value of the state variable to the prediction model 204A, and obtains the predicted value of the objective variable from the prediction model 204A.
The control variable determination unit 212B inputs the predicted value of the objective variable and the measured value of the state variable to the control model 204B, and obtains the recommended value of the control variable from the control model 204B.

Returning to the description of FIG.
The input amount presentation unit 213 is a functional unit that presents the recommended values of the control variables determined by the input amount determination unit 212 through the monitor 30 (see FIG. 1) or the like.
The input amount presenting unit 213 here is an example of the "presenting unit" in the scope of claims.

<Generation of learning model>
For reference, a method for generating the prediction model 204A and the control model 204B described above will be described.

<Generation method 1>
FIG. 5 is a flowchart for explaining a procedure for generating a learning model (prediction model 204A, control model 204B) that uses the exhaust gas concentration as a prediction value. The symbol S shown in the figure means a step.
The processing operations illustrated in FIG. 5 are used to generate both predictive model 204A and control model 204B. Note that the generation procedure shown in FIG. 5 will be described for a case where the information presentation device 20 also generates a learning model.

First, the processor 201 acquires teacher data (step 1). The training data here consist of the input amount of fuel (for example, city gas and air) as control variables, furnace status information (for example, the flow rate of waste liquid, the flow rate of process exhaust gas, the temperature inside the furnace, and the temperature of the furnace wall). , the measured value of the target variable (eg concentration of the exhaust gas) corresponding to the input amount and the status information.

Next, processor 201 learns teacher data (step 2).
In the case of the prediction model 204A, the processor 201 inputs the control variables and the state information of the furnace to the input layer, and advances the learning of the parameters of the intermediate layers so that the predicted values of the objective variables are output from the output layer.

In the case of the control model 204B, the processor 201 advances the learning of the intermediate layer parameters so that the recommended values of the control variables are output from the output layer when the input layer receives the target variable and the state information of the furnace.
The recommended value here is determined according to the adjustment target. For example, when the purpose is to minimize the concentration (objective variable) of the exhaust gas discharged from the combustion furnace 10, the control variable that minimizes the concentration of the exhaust gas among a plurality of teacher data that matches the current state variables. Advance the learning of the parameters of the hidden layer so that the values are output from the output layer.

After learning is performed for one teacher data, the processor 201 determines whether or not there is unprocessed teacher data (step 3).
A positive result is obtained in step 3 if there is unprocessed training data. In this case, processor 201 returns to step 1 and one of the unprocessed teacher data is selected for training.
If there is no unprocessed teacher data, a negative result is obtained in step 3. In this case, processor 201 terminates the process of generating prediction model 204A. As a result, a prediction model 204A is generated that inputs the flow rate of city gas, the flow rate of air, the flow rate of waste liquid, the flow rate of process exhaust gas, the temperature in the furnace, and the temperature of the furnace wall, and outputs the predicted value of the exhaust gas concentration. be done.

<Generation method 2>
Many samples are required to improve the learning accuracy of the learning model. However, collecting a large number of samples is time consuming.
Therefore, in generation method 2, a Gradient Boosting Decision Tree (GBDT) is applied as a method of increasing the learning accuracy of a prediction model using limited samples.

FIG. 6 is a diagram explaining a method of correcting a learning model using a gradient boosting decision tree.
In FIG. 6, the correction of the learning model proceeds in the order of numbers 1, 2, 3, . . .
As shown in FIG. 6, the difference between the objective variable (concentration of exhaust gas) and the predicted value by the learning model created so far is learned, and the learning model is sequentially added until the difference becomes small.
FIG. 7 is a flowchart for explaining the procedure for generating a learning model. In FIG. 7, parts corresponding to those in FIG. 5 are shown with reference numerals.

First, processor 201 acquires teacher data (step 1), and learns the acquired teacher data (step 2). The teacher data here differs depending on the learning model to be learned.
After learning is performed for one teacher data, the processor 201 determines whether or not there is unprocessed teacher data (step 3).
A positive result is obtained in step 3 if there is unprocessed training data. In this case, processor 201 returns to step 1 .

If there is no unprocessed teacher data, a negative result is obtained in step 3. In this case, processor 201 calculates the difference between the output value of the learning model and the correct value (step 4). The correct value here is the measurement data corresponding to the output value.
Processor 201 then determines whether the difference is less than or equal to a threshold (step 5).

If the difference is greater than the threshold, step 5 yields a negative result. In this case, processor 201 modifies the output value of the learning model by multiplying the difference by the learning rate (step 6). After modifying the learning model, the processor 201 returns to step 4. This process is repeated until step 5 yields a positive result.
A positive result is obtained in step 5 if the difference is less than or equal to the threshold. In this case, the processor 201 ends the learning model generation process. As a result, even if the number of samples is small, a learning model with high accuracy in output value is generated.

<Generation method 3>
In the case of generation method 2, a highly accurate learning model is obtained even with a small number of samples of teacher data by correcting the learning model using a gradient boosting decision tree. Pre-increase the number of samples in .
FIG. 8 is a flowchart for explaining a method of correcting a learning model using teacher data generated by interpolation calculation. In FIG. 8, parts corresponding to those in FIG. 5 are shown with reference numerals corresponding thereto.

First, when the processor 201 acquires the measured values (step 11), the acquired measured values are used to generate teacher data (step 12).
FIG. 9 is a diagram for explaining a method of generating teacher data. (A) shows a method of increasing the number of teacher data samples using linear interpolation, and (B) shows a method of increasing the number of teacher data samples using a polynomial or regression model.
In FIG. 9, for convenience of explanation, the vertical axis represents the concentration of the specific gas, and the horizontal axis represents the input amount of city gas. In the case of FIG. 9, the purpose is to increase the number of samples for the concentration of the specific gas and the input amount of city gas.

In FIG. 9A, the midpoint between the two measured values is calculated as the complementary value by linear interpolation, but two or more complementary values may be calculated between the two measured values.
In FIG. 9B, a function passing through the measured values is defined, and the midpoint between the two measured values is calculated as the complementary value. may
When the generation of complement values used as teacher data is completed, the same processing as generation method 1 described with reference to FIG. 5 is executed.

<Recommended value presentation operation>
Next, the recommended value presentation operation by the information presentation device 20 used in the first embodiment will be described. Two presentation operation examples will be described below.

<When the process ends by presenting the recommended values to the worker>
FIG. 10 is a flow chart for explaining an example of an operation for presenting a recommended fuel gas value that achieves an adjustment target. It should be noted that the processing operations shown in FIG. 10 are realized through execution of a program by the processor 201. FIG. This processing operation is an example of an information processing method.
First, the processor 201 acquires measurement data from sensors installed in various parts of the combustion furnace 10 (step 21). The measured data here are, for example, real-time values of control variables (fuel gas input amount) and state variables (waste input amount, etc.).

Next, processor 201 inputs the measured data to prediction model 204A (see FIG. 2) to predict the predicted value of the objective variable (step 22). The target variable here is, for example, the real-time value of the concentration of the specific gas discharged from the combustion furnace 10 .
Subsequently, the processor 201 inputs the predicted value of the objective variable and the measured value of the state variable to the control model 204B (see FIG. 2) to determine the injection amount of the fuel gas (step 23).

After that, the processor 201 presents the determined injection amount of the fuel gas on the operation screen (step 24). The input amounts here are recommended values. The operation screen is displayed on the monitor 30 (see FIG. 1).
The operator of the combustion furnace 10 manually adjusts the opening degree of the fuel gas supply valve based on the amount of fuel gas to be fed presented on the operation screen.
This processing operation is used when the supply valve installed in the combustion furnace 10 is not compatible with remote control.

<When processing ends with automatic control based on recommended values>
FIG. 11 is a flowchart for explaining another example of the operation of presenting the recommended value of the fuel gas that achieves the adjustment target. In FIG. 11, the parts corresponding to those in FIG. 10 are indicated by the reference numerals.
The processing operations shown in FIG. 11 are also realized through execution of the program by the processor 201. FIG.
Among the processing operations shown in FIG. 11, the processing operations up to step 24 are the same as the processing operations shown in FIG.

In the case of the processing operation shown in FIG. 11, the processor 201 automatically adjusts the opening degree of the fuel gas supply valve according to the determined input amount of fuel gas after executing step 24 (step 25). Automatic adjustment here is performed when the city gas or air supply valve is an electric valve and, in addition, communication is possible between the information presentation device 20 and the electric valve. It should be noted that even if communication with the supply valve is poor, the operator who has confirmed the input amount presented in step 24 can continue to adjust the combustion furnace 10 .
On the other hand, when automatically adjusting the opening of the supply valve, step 24 may be skipped. Even if step 24 is skipped, a mechanism is adopted in which the recommended input amount is always displayed on the monitor 30 when instructed by the operator.

<Work screen example>
Examples of work screens corresponding to the purpose of adjustment will be described below. In addition, the prediction model 204A (see FIG. 2) and the control model 204B (see FIG. 2) that have been learned according to the purpose of adjustment are used to present the recommended value for the fuel gas input amount displayed on the work screen. .

<Screen example 1>
FIG. 12 is an example of a screen presenting a recommended input amount of fuel gas when the adjustment target is to minimize the CO concentration. A screen example shown in FIG. 12 is displayed on the monitor 30 .
The screen shown in FIG. 12 is composed of a combustion environment column 301A, a recommended control value column 302A, and a control result presentation column 303A.

Measured values of state variables are displayed in the combustion environment column 301A.
The recommended control value column 302A displays the city gas flow rate and the air flow rate for minimizing the concentration of CO emitted under the current combustion environment.
The control result presentation column 303A displays that the CO concentration is estimated to be 4 [ppm] when the flow rate of the fuel gas is controlled. By displaying the estimated values after control in this way, the operator can specifically know the optimal solution under the current combustion environment.

<Screen example 2>
FIG. 13 is an example of a screen that presents a recommended input amount of fuel gas when the adjustment target is adjustment of the in-furnace temperature. The screen example shown in FIG. 13 is also displayed on the monitor 30 .
The screen shown in FIG. 13 is composed of a combustion environment column 301B, a recommended control value column 302B, and a control result presentation column 303B.
In the case of FIG. 13, the control result presentation column 303B displays that the furnace temperature after controlling the flow rate of the fuel gas is estimated to be 800.degree.

<Screen example 3>
FIG. 14 is an example of a screen that presents a recommended input amount of fuel gas when the adjustment target is to minimize the amount of fuel gas that is input to the combustion furnace 10 . The screen example shown in FIG. 14 is also displayed on the monitor 30 .
The screen shown in FIG. 14 includes a combustion environment column 301C, a recommended control value column 302C, and a control result presentation column 303C.
In the case of FIG. 14, the control result presentation column 303C displays that the reduction amount due to the control of the flow rate of the fuel gas is estimated to be 0.3 Nm ³ /h.

<Screen example 4>
FIG. 15 is an example of a screen presenting a recommended input amount of fuel gas when the adjustment target is maximization of heated efficiency. The screen example shown in FIG. 15 is also displayed on the monitor 30 .
The screen shown in FIG. 15 is composed of a combustion environment column 301D, a recommended control value column 302D, and a control result presentation column 303D.
In the case of FIG. 15, the control result presentation column 303D indicates that the heated efficiency is estimated to be 30% by controlling the flow rate of the fuel gas.

<Screen example 5>
FIG. 16 is an example of a screen presenting a recommended input amount of fuel gas when the adjustment target is improvement of flame radiation. The screen example shown in FIG. 16 is also displayed on the monitor 30 .
The screen shown in FIG. 16 is composed of a combustion environment column 301E, a recommended control value column 302E, and a control result presentation column 303E.
In the case of FIG. 16, the control result presentation column 303E displays that the radiation amount after controlling the fuel gas flow rate is estimated to be 50 W/m ² .

<Summary of Embodiment 1>
By using the information presentation device 20 described in the present embodiment, the combustion furnace 10, in which the amount of waste input and the combination and ratio of the types of waste to be input change from moment to moment depending on the emission source, can be controlled. Even in some cases, it is possible to present to the operator the optimum value of the control variable (fuel gas flow rate) for achieving any desired adjustment target.

<Embodiment 2>
<Learning focusing on the content of causative substances>
Here, a case where the adjustment target is minimization of the concentration of CO gas discharged from the combustion furnace 10 will be described.
Mechanisms 1 and 2 below are conceivable for the mechanism by which CO is generated from the combustion furnace 10 .

・ Mechanism 1:
The flame is cooled by the atomized waste liquid introduced into the combustion furnace 10, and the oxidation reaction that produces _CO2 is prematurely frozen and discharged as CO.
・Mechanism 2:
As a result of a large amount of waste liquid containing a large amount of carbon being put in, part of the carbon is not completely burned and is discharged as CO. The C content here is the causative substance of CO.

This embodiment focuses on the mechanism 2. FIG.
Even if the amount of the entire waste liquid is the same, the concentration of CO discharged from the combustion furnace 10 differs between when a waste liquid containing a large amount of C is burned and when a waste liquid containing little C is burned.
Therefore, in the present embodiment, in addition to the input amount of waste liquid and the like, information on the concentration and viscosity of C, which is the causative material of the exhaust gas, is used for learning of the prediction model 204A and the control model 204B. Incidentally, the viscosity of a waste liquid with a high content of C is high, and the viscosity of a waste liquid with a low content of C is low. Therefore, if the viscosity is known, the content of C in the waste liquid can be specified.

<Learning model generation processing 1>
FIG. 17 is a diagram illustrating an example of variables used to generate a learning model (that is, prediction model 204A and control model 204B).
It should be noted that the flow rate of city gas introduced into the combustion furnace 10 can be adjusted by the opening of the valve 101 . Also, the flow rate of the air introduced into the combustion furnace 10 can be adjusted by the opening of the valve 102 .

Waste liquid is introduced into the combustion furnace 10 from the waste liquid tank 103 , and waste liquid and process exhaust gas are introduced into the combustion furnace 10 from the production line 104 or the like. Exhaust gas is discharged from the combustion furnace 10 as a result of the combustion reaction of these combustible materials.
In the case of FIG. 17, the training data includes not only the input amount of the waste liquid, etc., but also the measured values of the concentration and viscosity of the C component contained in the waste liquid, etc. FIG. The concentration and viscosity measurements here are real-time values.

In the generation process 1, the difference in the content of C contained in the waste liquid or the like that is input at each point is reflected in the learning of the learning model.
Specifically, teacher data is prepared for each content of C, which is a causative substance of CO gas, and learning models for each content of C are learned using the prepared teacher data. By preparing a learning model for each C component content in this way, it is expected that the prediction accuracy of the concentration of CO gas will be improved more than a learning model focusing only on the input amount of waste liquid or the like. As a result, it is expected that the accuracy of predicting the recommended amount of fuel gas to achieve the adjustment target will be improved.

It should be noted that the content of C charged into the combustion furnace 10 may change with time or be constant regardless of time.
The change in the content of C content over time occurs, for example, when the concentration of C content in the waste liquid changes with time for each discharge source. In this way, when the C content changes over time, it is better to switch the learning model according to the C content at each point than to use the learning model learned only by the input amount. The accuracy of predicting the concentration of CO gas emitted at the time is improved. As a result, it is expected that the accuracy of predicting the recommended amount of fuel gas to achieve the adjustment target will be improved.

Even if the concentration of C content is constant regardless of time, the combustion furnace 10 of Company A that burns the waste liquid with the C content of C1, etc., and B that burns the waste liquid with the C content of C2, etc. The concentration of CO gas is different between the combustion furnace 10 of the company and the same amount of input waste liquid.
Therefore, the learning model learned with the content C1 is used for the combustion furnace 10 of Company A, and the learning model learned with the content C2 is used for the combustion furnace 10 of Company B, so that learning is performed only with the input amount. The accuracy of predicting the concentration of CO gas emitted at each point in time is improved as compared with the case of using the learning model.

Thus, in the generation process 1, the input amount of fuel (city gas and air), the input amount of furnace state information (process exhaust gas, waste liquid), and the causative substance of the exhaust gas are used as teacher data for learning the learning model. and the concentration (objective variable) of the exhaust gas discharged at the time of inputting each value are used.
The teacher data may use the viscosity of the waste liquid instead of the concentration of the C component in the waste liquid, or may use both the concentration of the C component in the waste liquid and the viscosity of the waste liquid.
Also, the concentration of C in the process exhaust gas may be included in the training data.

<Learning model generation processing 2>
FIG. 18 is a diagram illustrating another example of variables used to generate learning models (that is, prediction model 204A and control model 204B). In FIG. 18, the parts corresponding to those in FIG. 17 are indicated by the reference numerals.
In the case of FIG. 18, instead of measuring the concentration and viscosity of C contained in the waste liquid, the cumulative amount of C (time integral value) is included in the teaching data. The cumulative amount of C is calculated, for example, as a time integral value of the amount of C (=input amount×concentration) at each time point.

In the generation process 2, when the cumulative amount of C contained in the combustion furnace 10 is large, a temporary increase or decrease in the content of C contained in the waste liquid or the like does not appear in the change in the concentration of CO gas. Assume that when the cumulative amount of C charged into the combustion furnace 10 is small, a temporary increase or decrease in the content of C contained in the waste liquid or the like is immediately reflected in the change in the concentration of CO gas.
In such a case, as shown in FIG. 18, it is required to include the cumulative amount of C in the teacher data.

Thus, in the generation process 2, the input amount of fuel (city gas and air), the input amount of furnace state information (process exhaust gas, waste liquid), and the causative substance of the exhaust gas are used as teacher data for learning the learning model. and the measured value of the concentration (objective variable) of the exhaust gas discharged at the time of inputting each value.
When the viscosity of the waste liquid is given as information representing the content of C, the cumulative amount of C is calculated by the time integral value of the amount of C (=α×input amount×viscosity) at each time point. α is a coefficient.
In addition, the teacher data may include the cumulative amount of C derived from the process exhaust gas fed into the combustion furnace 10 .

<Summary of Embodiment 2>
Using the prediction model 204A described in the present embodiment, the difference in the concentration of the causative substance of the specific gas (that is, CO) contained in the waste (process exhaust gas, waste oil, waste liquid) thrown into the combustion furnace 10, It is possible to predict the concentration (objective variable) of the exhaust gas discharged from the combustion furnace 10 in real time with high accuracy regardless of the time change of the concentration.

In addition, since the difference in the concentration of the causative substance of the specific gas (that is, CO) introduced into the combustion furnace 10 and the temporal change in the concentration are reflected in the learning of the control model 204B, compared to the first embodiment, It becomes possible to determine the recommended value of the control variable that achieves the adjustment target with high accuracy.

17 and 18, CO is assumed to be the specific gas contained in the exhaust gas, and C is assumed to be the causative substance. Nitride can be considered as
Further, when SOx is assumed as the specific gas, sulfide may be considered as the causative substance. When dioxin is assumed as the specific gas, chlorine compounds and aromatic compounds may be considered as causative substances.

Incidentally, a method similar to that shown in FIGS. 20 to 22, which will be described later, may be applied to generate a learning model for each content of the causative substance contained in the waste liquid or the like.
Also, a method similar to that shown in FIG. 23, which will be described later, may be applied to predict the concentration of the exhaust gas and the recommended fuel gas input amount. That is, the learning model may be switched according to the content of the causative substance contained in the current waste liquid or the like.

<Embodiment 3>
<Control using a dedicated learning model that specializes in combining the contents of waste>
Here, a case will be described where the combination of the contents of the waste thrown into the combustion furnace 10 changes with the lapse of time.
In the embodiment described above, it is assumed that the combination of the contents of the wastes thrown into the combustion furnace 10 is constant.
However, the above-mentioned waste oil, waste liquid, and process exhaust gas are not always fed at the same time.

FIG. 19 is a diagram illustrating combustion patterns assumed in one combustion furnace 10. FIG. In FIG. 19, the materials to be burned in the combustion furnace 10 are assumed to be fuel and waste. Also, waste liquid, waste oil, incinerated materials, and process exhaust gas are assumed as waste.
However, the waste shown in FIG. 19 is an example, and it is not necessary to burn all of these in one combustion furnace 10 .

・Pattern #1:
Period/Pattern #2 in which only fuel (city gas and air) is supplied:
Period/Pattern #3 in which only the waste liquid is introduced:
Period/Pattern #4 in which only waste oil is supplied:
Period/Pattern #5 in which only incinerators are put in:
Period/Pattern #6 to #15 in which only process exhaust gas is introduced:
Period/Pattern #16 to #25 in which only any two of fuel, waste liquid, waste oil, incinerated matter, and process exhaust gas are input:
Period/pattern #26 to #30 in which only three of fuel, waste liquid, waste oil, incinerator, and process exhaust gas are input:
Period/Pattern #31 in which only four of fuel, waste liquid, waste oil, incinerated matter, and process exhaust gas are input:
The period during which all of the fuel, waste liquid, waste oil, incinerators, and process exhaust gas are input

In the present embodiment, a learning model (prediction model 204A, control model 204B) is generated for each of these patterns (for each period), and depending on the combination of combustibles put into the combustion furnace 10, the exhaust gas concentration The prediction accuracy is improved by switching the learning model (prediction model 204A, control model 204B) used for prediction.

<Learning model generation process according to combination of combustible materials>
20 to 22, the process of generating the learning model (prediction model 204A, control model 204B) according to the third embodiment will be described below.
FIG. 20 is a flowchart for explaining learning model generation processing according to the third embodiment.
The processor 201 classifies the teacher data according to the combination of combustible materials (step 31). The processor 201 classifies the teacher data according to the combination of combustible materials introduced into the combustion furnace 10 at each point in time, for example, based on information from sensors provided for each flow path assumed as combustible materials.

FIG. 21 is a diagram for explaining the function of classifying teacher data. The teacher data acquisition unit 221 realized through program execution by the processor 201 has a classification function. In Embodiment 3, the classification function is enabled, and the measured values and the like at each point of time are classified into teacher data for each combustible material combination.
For example, when the current combination is pattern #1, the teacher data acquisition unit 221 obtains, for example, the temperature inside the furnace, the temperature of the furnace wall, the oxygen concentration inside the furnace, the flow rate of city gas, the flow rate of air, and the concentration of exhaust gas. It is classified as teacher data for pattern #1. As a result, the teacher data acquisition unit 221 generates 31 types of teacher data.

Returning to the description of FIG. After classifying the teacher data, processor 201 uses the classified teacher data to generate prediction model 204A and control model 204B corresponding to each classification (step 32).
FIG. 22 is a diagram for explaining the learning function of a dedicated learning model (prediction model 204A, control model 204B) corresponding to each pattern using teacher data after classification.

The learning model learning unit 222 shown in FIG. 22 generates a dedicated learning model corresponding to each combination by making the learning model learn using teacher data for each combustible material combination. In this embodiment, 31 types of learning models (prediction model 204A, control model 204B) are generated.

<Switching learning models>
FIG. 23 is a flowchart for explaining learning model switching processing according to the third embodiment. As a premise, it is assumed that a dedicated learning model (prediction model 204A, control model 204B) corresponding to each combination is prepared.
First, the processor 201 detects the combination of wastes being introduced into the combustion furnace 10 (step 41).

Processor 201 then determines whether the combination has changed (step 42). A change in the combination is determined by comparing the current detection result with the previous detection result.
If there is no change in the sensing result, processor 201 obtains a negative result at step 42 . In this case processor 201 returns to step 41 .

On the other hand, if a change is recognized in the detection result, processor 201 obtains a positive result in step 42 . In this case, the processor 201 switches to the learning model corresponding to the current combination (step 43). For example, if the current combination is pattern #3, processor 201 switches to a dedicated model corresponding to pattern #3 as prediction model 204A and control model 204B.
This enables output values optimized for the current combustion environment to be output from the target variable prediction section 212A (see FIG. 4) and the control variable determination section 212B (see FIG. 4).

<Summary of Embodiment 3>
According to the present embodiment, when the combination of the contents of the combustible materials put into the combustion furnace 10 changes with the passage of time, the adjustment target is achieved compared to the other embodiments described above. It is possible to improve the accuracy of the amount of fuel gas recommended for
In other words, by using a dedicated learning model that is learned according to the combination of the contents of the combustibles, it is possible to improve the accuracy of control compared to using a highly versatile learning model.

<Embodiment 4>
<Control using a learning model specialized for the combination of the number of nozzles and the mounting position>
The piping structure of the combustion furnace 10 (see FIG. 1) may differ depending on the difference in the combustible materials, site restrictions, and the like. In this embodiment, attention is focused on the number of nozzles used for throwing in the waste and the difference in the combination of the mounting positions.
The number of nozzles used for throwing in the waste and the combination of mounting positions may depend on the structure or may change during combustion. The case where the number of nozzles and their mounting positions change during combustion is, for example, when the amount of fuel injected increases, the number of nozzles used for injection is increased, but when the amount of injection decreases, the number of nozzles used for injection is reduced. is.

In any case, the difference in the number of nozzles used for inputting fuel gas, wastes, etc. and their mounting positions affects the combustion of wastes, etc., and even if the amount of input to the combustion furnace 10 is the same, the concentration of exhaust gas, etc. Subject to change.
For example, even if the input amount is the same when the waste liquid is input using one nozzle and when the waste liquid is input using two nozzles, the concentration of the exhaust gas discharged from the combustion furnace 10 may differ. .
In addition, even if the input amount is the same when each nozzle is attached to the ceiling surface of the combustion furnace 10 and when each nozzle is attached to the side surface of the combustion furnace 10, the concentration of the exhaust gas discharged from the combustion furnace 10 is different. there is a possibility.

<Example of nozzle number and arrangement pattern>
FIG. 24 is a diagram illustrating a case where one nozzle is provided for each combustible material in the ceiling portion of the combustion furnace 10. As shown in FIG. In FIG. 24, parts corresponding to those in FIG. 17 are shown with reference numerals. Note that the wastes in FIG. 24 are waste liquid and waste oil.
In the case of FIG. 24, one nozzle for city gas, one nozzle for air, one nozzle for waste liquid, and one nozzle for waste oil, four nozzles in total are installed on the ceiling of the combustion furnace 10. provided in

FIG. 25 is a diagram illustrating a case where two nozzles are provided for each combustible material on the ceiling of the combustion furnace 10. FIG. In FIG. 25, the parts corresponding to those in FIG. 24 are indicated by the reference numerals. Wastes in FIG. 25 are also waste liquid and waste oil.
In the case of FIG. 25, two nozzles for city gas, two nozzles for air, two nozzles for waste liquid, and two nozzles for waste oil, total eight nozzles are installed on the ceiling of the combustion furnace 10. provided in

FIG. 26 is a diagram illustrating another example in which a total of two nozzles are provided for each combustible material dispersedly on the ceiling portion and the wall portion of the combustion furnace 10 . In FIG. 26, parts corresponding to those in FIG. 24 are shown with reference numerals corresponding thereto. Wastes in FIG. 26 are also waste liquid and waste oil.
In the case of FIG. 26, a total of four nozzles, one for city gas, one for air, one for waste liquid, and one for waste oil, are attached to the ceiling of combustion furnace 10. A total of four nozzles, one for city gas, one for air, one for waste liquid, and one for waste oil, are provided on the side of the combustion furnace 10 is provided.

In addition, the number of nozzles for each combustible material may be three or more, and the mounting position of the nozzles may be on the bottom side of the combustion furnace 10 . Also, the nozzles attached to the side may be attached near the top surface or near the bottom surface. When a plurality of nozzles are attached to one side surface, the nozzles may be arranged vertically, horizontally, or diagonally. Further, when a plurality of nozzles are attached to two side surfaces, the nozzles may be attached to two opposing side surfaces or adjacent side surfaces.

Further, in the case of FIGS. 24 to 26, the case where the same number of nozzles are attached to each combustible material put into the combustion furnace 10 has been described, but the number of nozzles attached to each combustible material may be different. . For example, two nozzles may be prepared for inputting waste liquid, and one nozzle may be prepared for inputting waste oil.
In this way, the combination of the number of nozzles for ejecting the combustible material to the combustion furnace 10 and the mounting position can be classified into many patterns, considering the difference in the internal volume and the shape of the internal space of the combustion furnace 10. .

<Switching learning models 1>
FIG. 27 is a flowchart for explaining learning model switching processing when the number and mounting positions of nozzles used for introducing waste change during combustion. As a premise, it is assumed that a dedicated learning model (prediction model 204A, control model 204B) is prepared for each combination of the number of nozzles and the mounting position.

First, the processor 201 detects a combination of the number of nozzles and their mounting positions (step 51).
Processor 201 then determines whether the combination has changed (step 52). A change in the combination is determined by comparing the current detection result with the previous detection result.
If there is no change in the sensing result, processor 201 obtains a negative result at step 52 . In this case, processor 201 returns to step 51 .

On the other hand, if a change is found in the detection result, processor 201 obtains a positive result in step 52 . In this case, the processor 201 switches to the learning model corresponding to the current combination (step 53).
For example, if the current combination is pattern #3, processor 201 switches to a dedicated model corresponding to pattern #3 as prediction model 204A and control model 204B.
This enables output values optimized for the current combustion environment to be output from the target variable prediction section 212A (see FIG. 4) and the control variable determination section 212B (see FIG. 4).

<Switching learning models 2>
FIG. 28 is a flowchart for explaining learning model setting processing when the number and mounting positions of nozzles used for throwing in waste do not change during combustion. As a premise, it is assumed that a dedicated learning model (prediction model 204A, control model 204B) is prepared for each combination of the number of nozzles and the mounting position.

First, the processor 201 acquires information on a combination of the number of nozzles and their mounting positions (step 61). The information here is registered, for example, when setting the information presentation device 20 (see FIG. 1).
Processor 201 then sets the use of the learning model corresponding to the obtained combination (step 62). This enables output values optimized for the current combustion environment to be output from the target variable prediction section 212A (see FIG. 4) and the control variable determination section 212B (see FIG. 4).

<Summary of Embodiment 4>
According to this embodiment, it is possible to use a dedicated prediction model prepared for each combination of the number of nozzles used for charging the combustible material and the mounting position. It is possible to improve the accuracy of the amount of fuel gas recommended to achieve the target.
In other words, by using a dedicated learning model that is learned according to the combination of the contents of the combustibles, it is possible to improve the accuracy of control compared to using a highly versatile learning model.

<Embodiment 5>
<Control using learning model considering control time lag>
Depending on the combustion furnace 10 (see FIG. 1), the time lag until the change in the input amount of fuel gas, waste, etc. appears as a change in the concentration of the exhaust gas discharged from the combustion furnace 10 may not be negligible. For the combustion furnace 10 with a large time lag, there is a possibility that the adjustment target cannot be achieved even if the learning model is learned without considering the time lag.

<Switching learning models>
FIG. 29 is a flowchart for explaining learning model setting processing in consideration of a time lag in control. As a premise, it is assumed that dedicated learning models (prediction model 204A, control model 204B) are prepared for each length of time lag in control.

First, the processor 201 acquires information on the length of the control time lag specific to the combustion furnace 10 (step 71). The length of the time lag here is the time difference until the change in the amount of fuel gas input is reflected in the concentration of the exhaust gas, and is measured, for example, when the combustion furnace 10 starts operating. Note that the time lag may be measured periodically. The acquired length of the time lag is registered in the information presentation device 20 (see FIG. 1).

Processor 201 then sets the use of the learning model corresponding to the obtained time lag (step 72). This enables output values optimized for the current combustion environment to be output from the target variable prediction section 212A (see FIG. 4) and the control variable determination section 212B (see FIG. 4).
If the latest value of the length of the time lag measured during inspection or the like differs from the current setting, the target variable prediction unit 212A (see FIG. 4) and the control variable determination unit 212B (see FIG. 4) refer to Change the learning model according to the latest value of the lag length.

<Example of control time lag>
FIG. 30 is a chart for explaining an example in which measured values such as the amount of combustible material put in at the reference time (current time) and measured values of the concentration of exhaust gas after a certain period of time from the reference time are used as teacher data. be. In FIG. 30, it is assumed that the time difference (time lag) between changes in the input amount of combustible material and changes in exhaust gas concentration is 3 minutes. Of course, 3 minutes is just an example.

The chart shown in FIG. 30 shows the input amount of fuel gas (city gas and air) at each time, and the state information (furnace temperature, furnace oxygen concentration, flow rate of waste liquid, flow rate of process exhaust gas, amount of waste material, flow rate of waste oil). ) and the measured value of the exhaust gas concentration (CO concentration).
In the case of FIG. 30, "2022/4/10:01" is the current time. In this case, the exhaust gas concentration associated with the input amount of fuel gas and the state information at the current time is 50 [ppm] at "2022/4/10:04" three minutes after the current time. That is, in FIG. 30, combinations of measured values that satisfy the relationship of the colored numerical values are acquired as teacher data and used for learning the prediction model.

FIG. 31 is a chart for explaining an example in which the measured value of the exhaust gas concentration at the reference time (current time) and the measured value of the input amount of the combustible material that was input a certain time before the reference time are used as teacher data. be.
In the case of FIG. 31 as well, it is assumed that the time difference (time lag) between the change in the input amount of the combustible material and the change in the exhaust gas concentration is 3 minutes.
However, in the case of FIG. 31, the time when the measured value of the concentration of the exhaust gas is obtained is set as the reference time (current time), and the input amount of the fuel gas (city gas and air) and the state information (furnace temperature , furnace oxygen concentration, waste liquid flow rate, process exhaust gas flow rate, waste amount, waste oil flow rate), and exhaust gas concentration (CO concentration).

Specifically, with respect to the exhaust gas concentration value (that is, 50 [ppm]) at "2022/4/10:04", the input amount of fuel gas at "2022/4/10:01" three minutes ago Associate the exhaust gas concentration associated with the state information.
That is, in FIG. 31, combinations of measured values that satisfy the relationship of colored numerical values are acquired as teacher data and used for learning the prediction model.
30 and 31, the combination of data recorded as teacher data is the same, but the time at which any measured value is obtained is different from the reference time.

<Summary of Embodiment 5>
According to the present embodiment, compared to the case where the learning model corresponding to the length of the time lag in the control of the combustion furnace 10 is not used, the accuracy of the amount of fuel gas recommended for achieving the adjustment target is increased. can be improved.
As for the length of the time lag used to generate the teacher data, for example, the time difference at which the matching rate between the measured value and the predicted value becomes high may be specified from experimental results.

<Embodiment 6>
<Control using history of operation data>
In the present embodiment, control of the fuel gas input amount without using the learning model will be described.

<Device configuration>
FIG. 32 is a diagram illustrating a configuration example of the information presentation device 20 used in the sixth embodiment. In FIG. 32, parts corresponding to those in FIG. 2 are shown with reference numerals. The information presentation device 20 here is also an example of an “information processing device”.
The hardware configuration of the information presentation device 20 shown in FIG. 32 is the same as the hardware configuration of the information presentation device 20 shown in FIG. The difference is that the operating data 204C is stored in the auxiliary storage device 204, and the learning models (prediction model 204A and control model 204B) are not stored.

<Example of operation data>
FIG. 33 is a diagram illustrating an example of operating data. Operational data is a collection (history) of measurement data. FIG. 33 shows only a part of the operating data regarding the combustion furnace 10. As shown in FIG. In the case of FIG. 33, the measurement data constituting the operation data is stored in units of one minute, but the units need not be one minute, and may be in units of 15 seconds, 30 seconds, or several minutes.

<Functional configuration of information presentation device>
FIG. 34 is a diagram illustrating a functional configuration example of the information presentation device 20 used in the sixth embodiment. In FIG. 34, parts corresponding to those in FIG. 3 are shown with reference numerals.
Processor 201 (see FIG. 2) in the present embodiment also functions as measurement data acquisition section 211, input amount determination section 212, and input amount presentation section 213 through program execution.

The difference from the functional configuration shown in FIG. 3 is that a search unit 212C is provided as a sub-function of the input amount determination unit 212, and that the data stored in the auxiliary storage device 204 is operation data 204C. .
The search unit 212C searches the operation data 204C using the input amount of fuel gas and the measured values of the state variables given from the measurement data acquisition unit 211 as search keys, and searches for operation data that satisfies the search keys. A search by the search unit 212C does not need to be a perfect match. For example, the driving data 204C whose distance (eg, Euclidean distance, cosine similarity) in the data space from the search key is within a threshold is output as a search result.

The input amount determination unit 212 identifies the operation data 204C that is highly consistent with the adjustment target from among the search results. In other words, the input amount determining unit 212 identifies the operating data 204C that is highly consistent with the adjustment target from among the operating data 204C that match the current combustion environment. When the operation data 204</b>C highly consistent with the tonality target is identified, the input amount determination unit 212 outputs the input amount of fuel gas in the identified operation data 204</b>C to the input amount presentation unit 213 . For example, when three pieces of operation data 204C are output as search results, the operation data 204C with the lowest specific gas (for example, CO) concentration among the three is specified as the operation data 204C with the highest consistency with the adjustment target. .

<Decision processing of input amount>
FIG. 35 is a flowchart for explaining processing for determining the input amount by the input amount determination unit 212. As shown in FIG.
First, the processor 201 acquires measurement data (step 81), and sets a search key from the acquired measurement data (step 82). The search key may be part of the acquired measurement data.

Next, processor 201 retrieves driving data using the retrieval key (step 83).
Processor 201 then determines whether there is any matching operating data (step 84). If there is matching operational data, step 84 yields a positive result. In this case, the processor 201 presents to the monitor 30 (see FIG. 1) the input amount of fuel gas used in the operating data with the lowest CO concentration (step 85).

On the other hand, if no matching operating data exists, step 84 yields a negative result. In this case, processor 201 determines whether or not there is driving data whose distance from the search key is less than a threshold (step 86). In other words, the processor 201 searches for driving data close to the search key.
If there is no driving data below the threshold, step 86 yields a negative result. In this case, processor 201 changes the threshold (step 87 ) and then returns to step 86 . Specifically, the numerical value of the threshold is increased.

A positive result is obtained at step 86 if there is driving data below the threshold. In this case, the processor 201 proceeds to step 85 and presents to the monitor 30 the input amount of the fuel gas used in the operating data that minimizes the CO concentration.
If multiple pieces of driving data whose distance from the search key is less than the threshold are found, the operation data is further narrowed down from the viewpoint of the adjustment target, and the average value of the fuel gas input amount in the multiple pieces of driving data after narrowing down. etc. may be determined as recommended values. Note that the recommended value is not limited to the average value.

<Summary of Embodiment 6>
By adopting a method of searching for operating data used under a combustion environment that matches or is similar to the current combustion environment, and identifying operating data that is highly consistent with the adjustment target from among the searched operating data, the learning model , it is possible to present the amount of fuel gas required to achieve the adjustment target.
Since the method of the present embodiment directly uses the operating data as actual values of the combustion furnace 10 to be controlled, the consistency of control with the combustion furnace 10 to be controlled is enhanced. In other words, expected control can be realized.

<Embodiment 7>
In this embodiment, a case in which there are a plurality of adjustment targets will be described.
The other embodiments described above assume a case where there is one adjustment target. For example, the goal is to minimize the CO concentration. On the other hand, there may be cases where it is desired to simultaneously minimize the CO concentration and maximize the combustion efficiency.

Therefore, in the present embodiment, a method of expressing an adjustment target as a function f of a plurality of objective variables and presenting a recommended input amount of fuel gas that minimizes or maximizes the function f will be described.
For example, if one of the objective variables is the concentration of CO and one of the other objective variables is the combustion efficiency P, the function f is defined by the following equation, for example.
f=w1*Xco-w2*P...Equation 1
However, Xco is the concentration value of CO. Also, w1 and w2 are adjustable weights.
Note that the combustion efficiency P is calculated by the amount of heat removed by the object to be heated/the amount of heat input.

The value of the function f shown in Equation 1 becomes small when two adjustment targets are achieved.
For example, even if the CO concentration value Xco is the same, the value of the function f with the lower combustion efficiency P is smaller than the value of the function f with the higher combustion efficiency P.
Also, even if the combustion efficiency P is the same, the value of the function f with the smaller CO concentration value Xco is smaller than the value of the function f with the higher combustion efficiency P.
It should be noted that, like the adjustment target 3, when the target is to minimize the input amount of fuel gas, the function f is defined as a function of the objective variable and the control variable.

<Functional configuration of information presentation device>
FIG. 36 is a diagram for explaining the processing contents of the input amount determination unit 212 used in the eighth embodiment.
As shown in FIG. 36, the input amount determination unit 212 includes a function value calculation unit 212D that calculates the value of the function f prepared for each adjustment target using the measured value representing the current combustion environment, and and a control variable decision unit 212E that decides the recommended value of the control variable for

A function f is prepared for each combination of a plurality of adjustment targets. The function f here may be defined for a combination of three or more adjustment targets. A variable corresponding to the adjustment target is used as the variable of the function f.
In the case of this embodiment, the value of function f is calculated based on the measured value of the sensor installed in combustion furnace 10 . Although the function value is calculated using the function f in the present embodiment, the function value corresponding to the current measurement value may be read out from a conversion table prepared for each combination of adjustment targets.

The control variable determining unit 212E receives the calculated function value and the measured value of the state variable as an input, and determines the calculated function value for the control model 204D that has learned the relationship in which the amount of fuel gas supplied to the combustion furnace 10 is the output. to determine the fuel gas input amount. This control model 204D is an example of a "third learning model" in the claims.
For the learning of the control model 204D here, teacher data composed of measured values of state variables, measured values of control variables, and function values corresponding to these are used.

The learning of the control model 204D is performed so that the recommended values of the control variables that minimize or maximize the function value (meaning that the adjustment target is achieved) are output. For example, if the combination of state variables is the same, the input amount of fuel gas in the operation data that more closely matches the adjustment target is selected as teacher data.
Incidentally, when learning the control model 204D, the prediction accuracy of the learning model may be improved by applying the gradient boosting decision tree or interpolation processing described above.

<Summary of Embodiment 7>
Even when the goal is to achieve a plurality of adjustment goals, by using the control model 204D that has been learned so as to minimize or maximize the value of the function defined according to the combination of the plurality of adjustment goals, It is possible to present the input amount of fuel gas required to achieve a plurality of adjustment targets.
As described in the sixth embodiment, the past operation data having numerical values lower or higher than the function value in the current combustion environment are searched, and the input amount of fuel gas required to achieve the adjustment target is presented. You may

<Embodiment 8>
<Example of providing recommended values from the server>
FIG. 37 is a diagram for explaining the conceptual configuration of the combustion furnace system 1A assumed in the eighth embodiment. In FIG. 37, parts corresponding to those in FIG. 1 are shown with reference numerals.
In the case of the combustion furnace system 1A shown in FIG. 37, the information processing server 40 existing on the network N executes the processing operations described above. Therefore, in FIG. 37, the information presentation device 20 is not provided. However, the terminal for uploading the operation data acquired from the sensor installed in the combustion furnace 10 to the information processing server 40 is arranged in the same building or site as the combustion furnace 10 .
The network N here is assumed to be, for example, a LAN (Local Area Network), the Internet, or a mobile communication system such as 4G or 5G.

The information processing server 40 executes, for example, acquisition of teacher data, generation of a learning model using the acquired teacher data, presentation of a recommended input amount of fuel gas, and the like. Incidentally, a learning model is not necessarily required for presenting the recommended input amount as described in the sixth embodiment.
The information processing server 40 is a server type computer. However, the information processing server 40 does not need to be composed of a single computer, and may be composed of a plurality of computers connected via a network N and cooperating to execute prediction model generation processing and the like. .

In addition, the training data acquisition process is executed by an information terminal (server, desktop computer, notebook computer, etc.) in the building or site where the combustion furnace 10 is installed, and the acquired training data is It may be uploaded to the processing server 40 .
Further, the input amount of fuel gas predicted by the information processing server 40 may be fed back to an information terminal in the building or site where the combustion furnace 10 is provided.

<Other embodiments>
(1) Although the embodiments of the present invention have been described above, the technical scope of the present invention is not limited to the scope described in the above-described embodiments. It is clear from the scope of claims that the technical scope of the present invention includes various modifications and improvements to the above-described embodiment.

(2) In the above-described embodiments, an industrial furnace having a manufacturing process and a chemical reaction process is assumed as a waste generation source, but the waste generation source is not limited to an industrial furnace.

(3) In the above-described embodiment, the flow rate of the process exhaust gas, the flow rate of the waste oil, the flow rate of the waste liquid, etc. were exemplified as the state information. ratio of city gas and air), and the opening degree of an exhaust damper provided in the exhaust path of the combustion furnace 10 may be included.

(4) In the above embodiment, the combustion furnace was explained, but an incinerator or a heating furnace may be used. An incinerator is a furnace that uses fuel gas to incinerate waste and other materials. A heating furnace is a furnace that heats an object to be heated using a fuel gas.

Reference Signs List 1, 1A Combustion furnace system 10 Combustion furnace 20 Information presentation device 30 Monitor 40 Information processing server 101, 102 Valve 103 Waste liquid tank 104 Production line, etc. 201 Processor 202 ... ROM, 203 ... RAM, 204 ... auxiliary storage device, 204A ... prediction model, 204B, 204D ... control model, 204C ... operation data, 205 ... I / O interface, 206 ... communication interface, 211 ... measurement data acquisition unit, 212 Input amount determination unit 212A Objective variable prediction unit 212B, 212E Control variable determination unit 212C Search unit 212D Function value calculation unit 213 Input amount presentation unit 221 Teacher data acquisition unit 222 …Learning model learning part

Claims (21)

the computer
obtaining measurements of state variables and current fuel input for a furnace in which at least one of fuel and waste is combusted;
Determining a recommended input amount for adjusting an objective variable corresponding to the obtained measured value and the input amount to an adjustment target;
presenting the determined recommended input amount on an operation screen;
A method of processing information,
Input the input amount and the measured value to a first learning model that has learned a relationship in which the input amount and the measured value are input, and the predicted value of the objective variable corresponding to the input amount and the measured value is output. to predict the predicted value of the objective variable,
Second learning for learning a relationship in which the predicted value of the objective variable and the measured value of the state variable are input, and the recommended input amount for adjusting the predicted value of the objective variable to the adjustment target is output. inputting the predicted value and the measured value to the model to determine the recommended input;
when detecting a change in the combination of the fuel and the waste introduced into the reactor, using the first learning model learned for the combination after the change to predict the predicted value of the objective variable ;
Information processing methods. The first learning model used for predicting the predicted value is switched according to at least one of the number of nozzles used for inputting the waste and the position of the nozzle used for inputting the waste.
The information processing method according to claim 1 . the computer
obtaining measurements of state variables and current fuel input for a furnace in which at least one of fuel and waste is combusted;
Determining a recommended input amount for adjusting an objective variable corresponding to the obtained measured value and the input amount to an adjustment target;
presenting the determined recommended input amount on an operation screen;
A method of processing information,
Using a search key given by the fuel input and the measured value of the state variable, retrieve the operating data of the reactor including the combination of the input, the measured value, and the measured value of the objective variable. death,
determining a recommended dosage of the fuel based on search results that adjust the measured value of the objective variable to the adjustment target ;
Information processing methods. If there is no search result corresponding to the search key,
Using a plurality of search results with high similarity to the search key, calculating the recommended input amount for adjusting to the adjustment target;
The information processing method according to claim 3 . the computer
obtaining measurements of state variables and current fuel input for a furnace in which at least one of fuel and waste is combusted;
Determining a recommended input amount for adjusting an objective variable corresponding to the obtained measured value and the input amount to an adjustment target;
presenting the determined recommended input amount on an operation screen;
A method of processing information,
If the adjustment goal is described as a function of multiple variables,
For a third learning model that has learned a relationship in which predicted values corresponding to the measured values of the plurality of variables are input and the recommended input amount for adjusting the predicted values to the adjustment target is output enter a value to determine the recommended dosage ;
Information processing methods. an acquisition unit for acquiring measurements of state variables and a current fuel input of a furnace in which at least one of fuel and waste is combusted;
a determination unit that determines a recommended input amount for adjusting an objective variable corresponding to the obtained measurement value and the input amount to an adjustment target;
a presentation unit that presents the determined recommended input amount on an operation screen;
An information processing device having
Input the input amount and the measured value to a first learning model that has learned a relationship in which the input amount and the measured value are input, and the predicted value of the objective variable corresponding to the input amount and the measured value is output. to predict the predicted value of the objective variable,
Second learning for learning a relationship in which the predicted value of the objective variable and the measured value of the state variable are input, and the recommended input amount for adjusting the predicted value of the objective variable to the adjustment target is output. inputting the predicted value and the measured value to the model to determine the recommended input;
when detecting a change in the combination of the fuel and the waste introduced into the reactor, using the first learning model learned for the combination after the change to predict the predicted value of the objective variable;
Information processing equipment. an acquisition unit for acquiring measurements of state variables and a current fuel input of a furnace in which at least one of fuel and waste is combusted;
a determination unit that determines a recommended input amount for adjusting an objective variable corresponding to the obtained measurement value and the input amount to an adjustment target;
a presentation unit that presents the determined recommended input amount on an operation screen;
An information processing device having
Using a search key given by the fuel input and the measured value of the state variable, retrieve the operating data of the reactor including the combination of the input, the measured value, and the measured value of the objective variable. death,
determining a recommended dosage of the fuel based on search results that adjust the measured value of the objective variable to the adjustment target;
Information processing equipment. an acquisition unit for acquiring measurements of state variables and a current fuel input of a furnace in which at least one of fuel and waste is combusted;
a determination unit that determines a recommended input amount for adjusting an objective variable corresponding to the obtained measurement value and the input amount to an adjustment target;
a presentation unit that presents the determined recommended input amount on an operation screen;
An information processing device having
If the adjustment goal is described as a function of multiple variables,
Predicted values for a third learning model that has learned a relationship in which the predicted values corresponding to the measured values of the plurality of variables are input and the recommended input amount for adjusting the predicted values to the adjustment target is output to determine the recommended dosage by entering
Information processing equipment. to the computer,
the ability to obtain measurements of state variables and current fuel input for a furnace in which at least one of fuel and waste is combusted;
A function of determining a recommended input amount for adjusting an objective variable corresponding to the obtained measurement value and the input amount to an adjustment target;
A function of presenting the determined recommended input amount on an operation screen;
It is a program to realize
Input the input amount and the measured value to a first learning model that has learned a relationship in which the input amount and the measured value are input, and the predicted value of the objective variable corresponding to the input amount and the measured value is output. to predict the predicted value of the objective variable,
Second learning for learning a relationship in which the predicted value of the objective variable and the measured value of the state variable are input, and the recommended input amount for adjusting the predicted value of the objective variable to the adjustment target is output. inputting the predicted value and the measured value to the model to determine the recommended input;
when detecting a change in the combination of the fuel and the waste introduced into the reactor, using the first learning model learned for the combination after the change to predict the predicted value of the objective variable;
program. to the computer,
the ability to obtain measurements of state variables and current fuel input for a furnace in which at least one of fuel and waste is combusted;
A function of determining a recommended input amount for adjusting an objective variable corresponding to the obtained measurement value and the input amount to an adjustment target;
A function of presenting the determined recommended input amount on an operation screen;
It is a program to realize
Using a search key given by the fuel input and the measured value of the state variable, retrieve the operating data of the reactor including the combination of the input, the measured value, and the measured value of the objective variable. death,
determining a recommended dosage of the fuel based on search results that adjust the measured value of the objective variable to the adjustment target;
program. to the computer,
the ability to obtain measurements of state variables and current fuel input for a furnace in which at least one of fuel and waste is combusted;
A function of determining a recommended input amount for adjusting an objective variable corresponding to the obtained measurement value and the input amount to an adjustment target;
A function of presenting the determined recommended input amount on an operation screen;
It is a program to realize
If the adjustment goal is described as a function of multiple variables,
Predicted values for a third learning model that has learned a relationship in which predicted values corresponding to the measured values of the plurality of variables are input and the recommended input amount for adjusting the predicted values to the adjustment target is output to determine the recommended dosage by entering
program. the computer
A furnace that burns waste using fuel gas, and obtains measured values of state variables and the current input amount of fuel gas for a furnace in which at least one of fuel gas and waste is burned, depending on the duration of the combustion. death,
The recommended input amount of the fuel gas for adjusting the objective variable corresponding to the acquired measured value and input amount to the adjustment target , and teaching the predicted value of the objective variable, the measured value, and the recommended input amount. Determined using a learning model trained on the data ,
presenting the determined recommended input amount on an operation screen;
Information processing methods. Further comprising a process of presenting the measured value used to determine the recommended input amount on the operation screen,
The information processing method according to claim 12 . further comprising a process of presenting, on the operation screen, a predicted value of the objective variable after adjustment according to the presented recommended input amount;
The information processing method according to claim 12 . The waste includes at least one of liquid waste, waste oil, incinerated materials, and process exhaust gas.
The information processing method according to claim 12 . The state variables include at least one of flow rate of waste liquid, flow rate of waste oil, flow rate of process exhaust gas, amount of incinerated material, oxygen concentration in the furnace, temperature in the furnace, temperature of the furnace wall,
The information processing method according to claim 12 . The target variables are the concentration of CO, the concentration of _CO2 , the concentration of _NOx , the concentration of _SOx , the concentration of dust, the concentration of dioxin, the concentration of unburned fuel gas, the concentration of greenhouse gases, the amount of flame radiation, at least one of the surface temperature of the heated effect;
The information processing method according to claim 12 . Further comprising a process of controlling the opening of the valve corresponding to the fuel gas based on the determined recommended input amount,
The information processing method according to claim 12 . Input the input amount and the measured value to a first learning model that has learned a relationship in which the input amount and the measured value are input, and the predicted value of the objective variable corresponding to the input amount and the measured value is output. to predict the predicted value of the objective variable,
Second learning for learning a relationship in which the predicted value of the objective variable and the measured value of the state variable are input, and the recommended input amount for adjusting the predicted value of the objective variable to the adjustment target is output. inputting the predicted values and the measured values to the model to determine the recommended inputs;
The information processing method according to claim 12 . A furnace that burns waste using fuel gas, and obtains measured values of state variables and the current input amount of fuel gas for a furnace in which at least one of fuel gas and waste is burned, depending on the duration of the combustion. an acquisition unit that
The recommended input amount of the fuel gas for adjusting the objective variable corresponding to the acquired measured value and input amount to the adjustment target , and teaching the predicted value of the objective variable, the measured value, and the recommended input amount. a determination unit that determines using a learning model learned using data ;
a presentation unit that presents the determined recommended input amount on an operation screen;
Information processing device having to the computer,
A furnace that burns waste using fuel gas, and obtains measured values of state variables and the current input amount of fuel gas for a furnace in which at least one of fuel gas and waste is burned, depending on the duration of the combustion. and the ability to
The recommended input amount of the fuel gas for adjusting the objective variable corresponding to the acquired measured value and input amount to the adjustment target , and teaching the predicted value of the objective variable, the measured value, and the recommended input amount. A function to determine using a learning model learned using data ;
A function of presenting the determined recommended input amount on an operation screen;
program to make it happen.

Images