JP7409580B1 - Furnace temperature control device, furnace temperature control method, and coke manufacturing method - Google Patents

Abstract

A furnace temperature control device (20) that controls the temperature of a coke oven in which a combustion chamber and a carbonization chamber are alternately connected is configured to control the actual temperature values of the combustion chamber and the carbonization chamber and the actual amount of heat supplied to the combustion chamber and the carbonization chamber. an input device (21) into which operational performance information including at least a value is input; time-series data of operational performance information that goes back a certain amount of time from a certain time; and a combustion chamber or carbonization after an arbitrarily determined predetermined time from a certain time. a temperature prediction unit (24) that calculates a temperature prediction value that is the temperature of the combustion chamber or carbonization chamber after an arbitrarily determined predetermined time based on a model expressing the relationship with the temperature of the chamber and operation performance information; The furnace temperature controller (25) calculates the amount of heat to be supplied to the combustion chamber and the carbonization chamber so as to reach the target temperature based on the predicted temperature value.

Description

The present disclosure relates to a furnace temperature control device, a furnace temperature control method, and a coke manufacturing method. The present disclosure particularly relates to a furnace temperature control device, a furnace temperature control method, and a coke manufacturing method in a coke oven in which combustion chambers and carbonization chambers are alternately connected to form a furnace bank.

In a coke oven, in which a plurality of combustion chambers and carbonization chambers are alternately connected to form a furnace group, coke is produced by carbonizing coal charged into the carbonization chamber using heat from an adjacent combustion chamber. In order to stabilize the quality of coke produced in a coke oven, improve production efficiency, and reduce the amount of heat of carbonization, it is important to control the temperature of the combustion chamber or carbonization chamber to a desired temperature. Coke ovens have a large heat capacity and a long response time constant to actions to manipulate the furnace temperature. Therefore, in order to control the temperature of the combustion chamber or carbonization chamber with high precision, it is necessary to accurately determine the future temperature. It is necessary to predict and take action. However, the temperature of a coke oven fluctuates in a complex manner due to a large number of variable factors such as the fuel gas supply status, the properties of the coal charged, the carbonization status, and the status of the adjacent coking chamber and combustion chamber. Therefore, it has been difficult to predict future temperatures with high accuracy.

As a method for predicting future temperature with high accuracy, for example, Patent Document 1 describes the behavior of temperature with respect to coke oven temperature variation factors as a step response of a first-order lag system that occurs time-series at regular intervals. We suggest modeling.

Furthermore, Patent Document 2 discloses that in order to reduce the uncertainty of the furnace temperature prediction formula model and the furnace temperature prediction error due to disturbance, the furnace temperature prediction error in the previous carbonization cycle is reflected in the furnace temperature prediction calculation at the current time. , proposes a method to improve the accuracy of future temperature prediction.

Japanese Patent Application Publication No. 6-158050 Japanese Patent Application Publication No. 5-255668

Here, in the technique of Patent Document 1, the parameters of the step response model are identified in advance and used by investigating the furnace temperature fluctuation when the furnace temperature fluctuation factor is varied individually. Therefore, the accuracy of the model decreases due to long-term changes in furnace conditions. Furthermore, in normal coke oven operation, a plurality of oven temperature variation factors continuously fluctuate, so it is difficult to identify the parameters of the step response model for each oven temperature variation factor during normal operation. Further, when the temperature with respect to the coke oven temperature variation factor deviates from the behavior of the first-order lag system, the temperature prediction accuracy decreases significantly.

The technique disclosed in Patent Document 2 improves future temperature prediction accuracy by reflecting the furnace temperature prediction error in the previous carbonization cycle in the furnace temperature prediction calculation at the current time. However, in general, coal properties, amount of heat supplied, or other furnace temperature variation factors may vary significantly from the previous carbonization cycle. In this case, it is difficult to explain the error of the furnace temperature prediction formula model only by the prediction error in the previous carbonization cycle, and the prediction accuracy deteriorates. Moreover, the variation pattern of the furnace temperature variation factor becomes a huge number of combinations when considering combustion chambers and carbonization chambers that are close to the target combustion chamber or carbonization chamber. Therefore, it is difficult to take into account the adjacent combustion chambers and carbonization chambers, and the prediction accuracy cannot be said to be sufficient.

As described above, it is a challenge to predict the future oven temperature with high accuracy in a situation where the condition of the coke oven changes from moment to moment and a plurality of oven temperature fluctuation factors continuously fluctuate.

The purpose of the present disclosure, which was made in view of the above circumstances, is to provide a furnace temperature control device, a furnace temperature control method, and a furnace temperature control device that enable highly accurate furnace temperature control by predicting the future temperature of a combustion chamber or a carbonization chamber with high accuracy. An object of the present invention is to provide a method for producing coke.

(1) A furnace temperature control device according to an embodiment of the present disclosure includes:
A furnace temperature control device for controlling the temperature of the combustion chamber or the carbonization chamber in a coke oven in which combustion chambers and carbonization chambers are alternately connected to form a furnace group,
an input device into which operational performance information including at least actual temperature values of the combustion chamber and the carbonization chamber and actual values of the amount of heat supplied to the combustion chamber and the carbonization chamber is input;
A model representing the relationship between time-series data of the operation performance information that goes back a certain time from a certain time and the temperature of the combustion chamber or the carbonization chamber after an arbitrarily determined predetermined time from the certain time, and the operation performance information. a temperature prediction unit that calculates a temperature prediction value that is the temperature of the combustion chamber or the carbonization chamber after an arbitrarily determined predetermined time based on the temperature prediction unit;
A furnace temperature control unit is provided that calculates an amount of heat to be supplied to the combustion chamber or the carbonization chamber so as to reach a target temperature based on the calculated temperature prediction value.

(2) As an embodiment of the present disclosure, in (1),
The operation performance information adjusts the actual weight of the coal charged into the coking chamber, the actual moisture content, the actual time elapsed from charging, and the amount of fuel gas supplied to the combustion chamber. This includes the actual value of the adjustment valve opening.

(3) As an embodiment of the present disclosure, in (1) or (2),
The temperature prediction unit is configured to calculate the predicted temperature value based on the operation performance information regarding the combustion chamber or the carbonization chamber, which is a target for calculating the temperature prediction value, as well as the operation performance information regarding the adjacent combustion chamber and the carbonization chamber. Then, the predicted temperature value is calculated.

(4) As an embodiment of the present disclosure, in any one of (1) to (3),
The temperature prediction unit calculates, for each cycle of temperature change of the combustion chamber or the carbonization chamber, the time series data of the operation performance information that goes back from a certain time by a period of time longer than the cycle of temperature change of the combustion chamber or the carbonization chamber. The characteristics of the time-series data of the operation performance information for each division, and the temperature of the combustion chamber or the carbonization chamber after a period of change in the temperature of the combustion chamber or the carbonization chamber from the certain time. A predicted temperature value, which is the temperature of the combustion chamber or the carbonization chamber after a period of change in the temperature of the carbonization chamber, is calculated based on the model representing the relationship and the operation performance information.

(5) A furnace temperature control method according to an embodiment of the present disclosure,
A furnace temperature control method for controlling the temperature of the combustion chamber or the carbonization chamber in a coke oven in which combustion chambers and carbonization chambers are alternately connected to form a furnace group, the method comprising:
an input step in which operational performance information including at least actual temperature values of the combustion chamber and the carbonization chamber and actual values of the amount of heat supplied to the combustion chamber and the carbonization chamber is input;
A model representing the relationship between time-series data of the operation performance information that goes back a certain period of time from a certain time and the temperature of the combustion chamber or the carbonization chamber after an arbitrarily determined predetermined time from the certain time, and the operation performance information. a temperature prediction step of calculating a temperature prediction value that is the temperature of the combustion chamber or the carbonization chamber after an arbitrarily determined predetermined time based on the temperature;
The method includes a furnace temperature control step of calculating an amount of heat to be supplied to the combustion chamber or the carbonization chamber so as to reach a target temperature based on the calculated temperature prediction value.

(6) As an embodiment of the present disclosure, in (5),
The operation performance information adjusts the actual weight of the coal charged into the coking chamber, the actual moisture content, the actual time elapsed from charging, and the amount of fuel gas supplied to the combustion chamber. This includes the actual value of the adjustment valve opening.

(7) As an embodiment of the present disclosure, in (5) or (6),
The temperature prediction step is based on the operation performance information about the combustion chamber or the carbonization chamber that is nearby, in addition to the operation performance information about the combustion chamber or the carbonization chamber for which the temperature prediction value is calculated. Then, the predicted temperature value is calculated.

(8) As an embodiment of the present disclosure, in any of (5) to (7),
The temperature prediction step is performed for each cycle of temperature change in the combustion chamber or the carbonization chamber, with respect to the time series data of the operation performance information that goes back from a certain time by a period longer than the cycle of temperature change in the combustion chamber or the carbonization chamber. The characteristics of the time-series data of the operation performance information for each division, and the temperature of the combustion chamber or the carbonization chamber after a period of change in the temperature of the combustion chamber or the carbonization chamber from the certain time. A predicted temperature value, which is the temperature of the combustion chamber or the carbonization chamber after a period of change in the temperature of the carbonization chamber, is calculated based on the model representing the relationship and the operation performance information.

(9) A method for producing coke according to an embodiment of the present disclosure includes:
The amount of fuel gas supplied to the combustion chamber or the fuel gas supply based on the amount of heat supplied to the combustion chamber or the carbonization chamber calculated by the furnace temperature control method according to any one of (5) to (8). Coke is manufactured by adjusting the opening degree of the regulating valve that adjusts the amount.

According to the present disclosure, there are provided a furnace temperature control device, a furnace temperature control method, and a coke manufacturing method that enable highly accurate furnace temperature control by predicting the future temperature of a combustion chamber or a coking chamber with high accuracy. be able to.

FIG. 1 is a schematic diagram showing the configuration of a furnace temperature control device according to an embodiment of the present disclosure. FIG. 2 is a flowchart showing a process flow of a furnace temperature control method according to an embodiment of the present disclosure. FIG. 3 is a plot diagram of actual values and predicted values in the conventional example. FIG. 4 is a plot diagram of actual values and predicted values in the example. FIG. 5 is a diagram showing the relationship between actual values and predicted values in the conventional example. FIG. 6 is a diagram showing the relationship between actual values and predicted values in the example.

Hereinafter, a furnace temperature control device, a furnace temperature control method, and a coke manufacturing method according to an embodiment of the present disclosure will be described with reference to the drawings.

(Configuration of furnace temperature control device)
FIG. 1 is a schematic diagram showing the configuration of a furnace temperature control device 20 according to this embodiment. A coke oven 1 shown in FIG. 1 includes N combustion chambers 2 (2-1 to 2-N) and N-1 carbonization chambers 3 (3-1 to 3-(N-1)). are connected alternately to form a furnace group. Here, N is an integer of 2 or more, but is not limited to a specific number. The coke oven 1 charges coal, which is a raw material, into a carbonization chamber 3 and supplies fuel gas G to a combustion chamber 2. The coke oven 1 heats the carbonization chamber 3 with heat generated by the combustion chambers 2 on both sides, carbonizes the coal in the carbonization chamber 3, and produces coke. The furnace temperature control device 20 according to this embodiment controls the temperature of the combustion chamber 2 or the carbonization chamber 3 in the coke oven 1.

The coke oven 1 comprises a gas main 4 connected at one end to a gas supply source. The other end of the N-1 branched gas main pipe 4 is piped to the combustion chamber 2, and fuel gas G is supplied to the combustion chamber 2. A regulating valve 5 is provided at one end of the gas main pipe 4 to adjust the flow rate of the fuel gas G supplied to the entire furnace group (the total flow rate of the fuel gas G supplied to the combustion chamber 2). Further, each of the other branched ends is provided with a kiln-specific adjustment valve 6 (6-1 to 6-(N-1)) for finely adjusting the distributed gas flow rate and supplying the gas to the combustion chamber 2. ing. The fuel gas G finely adjusted by the 1st to (N-1)th kiln adjustment valves 6 is supplied to the two combustion chambers 2 adjacent to the corresponding 1st to (N-1)th carbonization chambers 3. . The opening degree (adjustment valve opening degree) of the adjustment valve 5 and the individual kiln adjustment valve 6 is controlled by the control terminal 10.

The furnace temperature control system includes a control terminal 10, a furnace temperature control device 20, and a display device 30 as main components. The control terminal 10 is configured by an information processing device such as a personal computer or a workstation. The control terminal 10 monitors the states of the combustion chamber 2 and the carbonization chamber 3 and manages the combustion chamber 2 and the carbonization chamber 3 according to the furnace temperature control device 20 . Further, the control terminal 10 manages the operation of the coke oven 1. For example, the control terminal 10 adjusts the opening degree of the regulating valve 5 and adjusts the amount of fuel gas G to be supplied to the entire furnace group so that the average temperature value of the combustion chamber 2 and carbonization chamber 3 of the entire furnace group becomes the target temperature. Control the flow rate. In addition, the control terminal 10 controls the flow rate of the fuel gas G supplied to the combustion chamber 2 by adjusting the opening degree of the adjustment valve 6 for each kiln, so that the temperatures in the combustion chamber 2 and the carbonization chamber 3 can be adjusted respectively. The operation of the coke oven 1 is managed so as to reach the individually set target temperature.

The furnace temperature control device 20 is configured by an information processing device such as a personal computer or a workstation. The furnace temperature control device 20 includes an input device 21 , a database 22 , a model learning section 23 , a temperature prediction section 24 , a furnace temperature control section 25 , and an output device 26 .

The input device 21 is an input interface into which various measurement results and operation performance information regarding the coke oven 1 are input. The input device 21 includes a keyboard, a mouse, a pointing device, a data receiving device, a graphical user interface (GUI), and the like. The input device 21 receives operational performance information, parameter setting values, etc. from the outside, writes the information into the database 22, and transmits it to the temperature prediction unit 24. Operation performance information is input to the input device 21 from the control terminal 10 . The operation performance information includes at least actual values of the temperature of the combustion chamber 2 and the carbonization chamber 3 (temperature measurement results) and actual values of the amount of heat supplied to the combustion chamber 2 and the carbonization chamber 3. The actual value of the amount of heat supplied to the combustion chamber 2 and the carbonization chamber 3 may include the opening degree of the regulating valve 5 and the per-kiln regulating valve 6. The operation performance information may include the flow rate and components of the fuel gas G supplied to the entire furnace group. Moreover, it is preferable that the operation performance information includes the actual weight value, the actual water content value, and the actual value of the elapsed time from the charging of the coal charged into the carbonization chamber 3.

Here, as will be described later, the temperature prediction unit 24 is based on operation performance information regarding the adjacent combustion chamber 2 and carbonization chamber 3 in addition to operation performance information regarding the target combustion chamber 2 or carbonization chamber 3. , calculate the temperature prediction value. Therefore, the operation performance information inputted into the input device 21 includes operation performance information about the adjacent combustion chamber 2 and carbonization chamber 3 in addition to the operation performance information about the specific combustion chamber 2 or carbonization chamber 3. In this embodiment, for example, when inputting operation performance information regarding the combustion chamber 2-n, operation performance information regarding the adjacent carbonization chamber 3-(n-1) and carbonization chamber 3-n is also input. Furthermore, operation performance information regarding the adjacent combustion chamber 2-(n-1) and combustion chamber 2-(n+1) is also input. Further, the flow rate and components of the fuel gas G related to the amount of heat supplied to the combustion chamber 2-n may also be input. Further, the opening degree of the regulating valve 5, the opening degree of the individual kiln regulating valve 6-(n-1), and the opening degree of the individual kiln regulating valve 6-n may be input together. Here, the adjacent combustion chamber 2 and carbonization chamber 3 do not necessarily include the combustion chamber 2 and the carbonization chamber 3, and may be a plurality of combustion chambers 2 or a plurality of carbonization chambers 3.

The database 22 is a storage device that stores the operational performance information input to the input device 21 and the model generated by the model learning section 23. Various information input to the input device 21 and model formulas and parameters calculated by the model learning section 23 are transmitted to the database 22 . Further, the model formula and parameters stored in the database 22 are used by the temperature prediction unit 24.

The model learning section 23, the temperature prediction section 24, and the furnace temperature control section 25 are realized by an arithmetic processing device such as a CPU. The model learning section 23, the temperature prediction section 24, and the furnace temperature control section 25 are realized, for example, by executing a computer program in an arithmetic processing device. Here, as another example, the model learning section 23, the temperature prediction section 24, and the furnace temperature control section 25 may have a dedicated computing device or a computing circuit.

Based on the operation record information stored in the database 22, the model learning unit 23 uses time-series data of the operation record information going back a certain amount of time from a certain time, and the combustion chamber 2 or the carbonization chamber after an arbitrary time from the certain time. A model is generated that expresses the relationship between No. 3 and temperature. In other words, a model is generated that predicts the temperature of the target combustion chamber 2 or carbonization chamber 3 after an arbitrary period of time. Here, arbitrary time means that there is no limit to the setting of future time for predicting temperature. Since the state of the coke oven 1 changes from moment to moment, there is a practical limit to the amount of time that can be predicted with sufficient accuracy, although it also depends on the period of time-series data. However, in the generated model, there are no particular setting restrictions based on prediction accuracy. Therefore, the future time at which the temperature is predicted can be set arbitrarily. The model learning unit 23 transmits the generated model formula and its parameters to the database 22. Hereinafter, a model formula and its parameters may be referred to as a model.

The temperature prediction unit 24 calculates the temperature of the combustion chamber 2 or the carbonization chamber 3 after a predetermined time arbitrarily determined during model learning, based on the operation performance information transmitted from the input device 21 and the model stored in the database 22. Calculate the temperature prediction value. Here, the operation performance information transmitted from the input device 21 is the operation performance information regarding the combustion chamber 2 or the carbonization chamber 3 for which the temperature prediction value is calculated, and is the information that has just been measured (the latest). good. In addition, as described above, the operation performance information regarding the target combustion chamber 2 or the carbonization chamber 3 includes, in addition to the operation performance information regarding the target combustion chamber 2 or the carbonization chamber 3, the adjacent combustion chamber 2 or the carbonization chamber Contains operational performance information for 3. The temperature prediction unit 24 can calculate an accurate temperature prediction value in consideration of heat conduction from the adjacent combustion chamber 2 and carbonization chamber 3. The temperature prediction unit 24 transmits the calculated temperature prediction value of the combustion chamber 2 or the carbonization chamber 3 after an arbitrarily determined predetermined time to the furnace temperature control unit 25.

The furnace temperature control unit 25 calculates the amount of heat to be supplied to the combustion chamber 2 or the carbonization chamber 3 based on the temperature prediction value calculated by the temperature prediction unit 24 so as to reach the target temperature. Here, the calculation of the amount of heat supplied to the combustion chamber 2 or the carbonization chamber 3 may be the calculation of the opening degree of the adjustment valve 5 and the adjustment valve 6 for each kiln in order to satisfy the required amount of supply heat. The furnace temperature control unit 25 transmits the calculated amount of supplied heat or the adjustment valve opening to the output device 26.

The output device 26 transmits the supplied heat amount or the adjustment valve opening calculated by the furnace temperature control section 25 to the control terminal 10. The control terminal 10 adjusts the opening degrees of the regulating valve 5 and the kiln-specific regulating valve 6 based on the information transmitted from the output device 26. In this way, the furnace temperature control device 20 controls the temperature of the combustion chamber 2 or the carbonization chamber 3 in the coke oven 1 via the control terminal 10. In the coke manufacturing method in the coke oven 1, the amount of fuel gas supplied to the combustion chamber 2 or the amount of fuel gas supplied is adjusted based on the amount of heat supplied to the combustion chamber 2 or the carbonization chamber 3 calculated by the furnace temperature control device 20. Coke is produced by adjusting the opening degree of the regulating valve. The output device 26 also has a function of transmitting information calculated by the furnace temperature control device 20 to the display device 30, and can also display the calculation results output from the furnace temperature control device 20. . Here, the display device 30 is, for example, a liquid crystal display.

The furnace temperature control device 20 having such a configuration controls the temperature of the combustion chamber 2 or the carbonization chamber 3 with high accuracy by executing the furnace temperature control process described below.

(furnace temperature control process)
FIG. 2 is a flowchart showing the flow of the process of the furnace temperature control method (furnace temperature control process) executed by the furnace temperature control device 20 according to an embodiment of the present disclosure. The flowchart shown in FIG. 2 may be started at any timing when operation performance information is acquired or at a timing when a signal instructing to start processing is input from the outside.

In the process of step S1 (input step), the input device 21 acquires operation performance information. The input device 21 transmits the acquired operational performance information to the database 22 and the temperature prediction unit 24. Thereby, the furnace temperature control process proceeds to the process of step S2.

In the process of step S2 (model generation step), the model learning unit 23 generates a model based on the operation performance information stored in the database 22. The process of step S2 may be executed each time operation performance information is acquired in step S1, or may be executed when model generation is instructed by a signal input from the outside. That is, the furnace temperature control process may proceed to the process of Step S3 without executing the process of Step S2. If the process in step S2 is not executed, the most recent model generated by the model learning unit 23 and stored in the database 22 may be used.

In the process of step S2, if X is the time-series data matrix of operational performance information that is input to the model to be generated, and y is the temperature of the combustion chamber 2 or carbonization chamber 3 after an arbitrary time that will be the output, the model to be generated is It can be expressed by the following equations (1) and (2).

Here, the subscripts of X and y represent time. t is an arbitrary time in the past. u is an arbitrarily determined time, and (t+u) is the predicted time. i is an arbitrary retroactive time starting from a certain time t. j is the number of items of operation performance information. f is the relationship between the time-series data matrix of j pieces of operation performance information going back from a certain time t to i and the temperature of the combustion chamber 2 or the carbonization chamber 3 after a certain time t to a future time u. This is a function that indicates

Although one model input/output data set is created for one time t, it is generally known that the more data sets there are, the more accurate the model is with respect to unknown inputs. Therefore, it is preferable to create data sets for as many times t as possible.

The time u for setting the predicted future time and the backward time i can be appropriately set according to the time constant or settling time of the response to temperature change in the combustion chamber 2 or the carbonization chamber 3. preferable. Here, the time constant or settling time is a time constant or settling time for the amount of heat supplied to the combustion chamber 2 or the carbonization chamber 3 or the adjustment valve opening degree. For example, if the sum of u and i is extremely shorter than the settling time, temperature changes at times before i, which is the retroactive time, cannot be taken into account, and the accuracy of the model may deteriorate. In addition, in a coke oven, the steps from coal charging to extrusion are usually performed at regular intervals in one kiln. Therefore, during steady state, the temperature to be predicted fluctuates periodically in accordance with the operating cycle. When y, which is the temperature to be predicted, exhibits periodic behavior, the sum of u and i may be set to a multiple of the change period (fluctuation period) of y. By setting u and i in this way, it is possible to reduce prediction errors due to periodic changes in the temperature of the prediction target.

In addition, for the number of data from time t to (t- _i ), that is, the number of rows of Just set it to the number divided by the period. Furthermore, when the temperature to be predicted exhibits periodic behavior, the number of data from time t to (ti) may be set to the number obtained by dividing i by the temperature cycle. When the input cycle of operation performance information is shorter than the data interval from time t to (ti), a feature value including an average value or rate of change of operation performance information may be calculated and input. When the input cycle of operation performance information is longer than the data interval from time t to (ti), a value obtained by linearly complementing the operation performance information may be input. In this way, by classifying the input operation performance information by change period, that is, by processing it into categories corresponding to periodic changes in the temperature to be predicted, it is possible to It becomes possible to reduce prediction errors.

In relation to j above, the operational performance information items include the actual temperature values (temperature measurement results) of the combustion chamber 2 and the carbonization chamber 3 and the amount of heat supplied to the combustion chamber 2 and the carbonization chamber 3, as described above. Contains at least actual values. The actual value of the amount of heat supplied to the combustion chamber 2 and the carbonization chamber 3 may include the opening degree of the regulating valve 5 and the per-kiln regulating valve 6. The item of operation performance information also includes performance values for the adjacent combustion chamber 2 and carbonization chamber 3. Items of operation performance information may include the flow rate and components of the fuel gas G supplied to the entire furnace group. Moreover, the items of operation performance information preferably include the actual weight value, the actual water content value, and the actual value of the elapsed time from the charging of the coal charged into the carbonization chamber 3.

It is preferable that the function f is a model that can efficiently learn or explain the characteristics of the variation information in the direction of i, which is the backward time, that is, in the row direction of _Xt . For example, as a model generation method, a regression-type neural network may be used that uses the value of the output layer for the input one time ago as the input layer of the next time. Further, as a model generation method, a neural network having a convolution layer for two-dimensional _Xt may be used. Additionally, automatic machine learning logic may be used that calculates model accuracy for multiple models and selects the optimal model. Additionally, a method called ensemble learning that combines multiple models may be used.

The model learning unit 23 transmits the generated model to the database 22. Thereby, the furnace temperature control process proceeds to step S3.

In the process of step S3 (temperature prediction step), the temperature prediction unit 24 calculates (predicts) the temperature of the combustion chamber 2 or the carbonization chamber 3 at a future time (t+u) determined during model learning. The temperature prediction unit 24 processes the operation performance information acquired by the input device 21 into a format similar to the above-mentioned _Xt . The temperature prediction unit 24 calculates the temperature of the combustion chamber 2 or the carbonization chamber 3 by inputting the processed operational performance information into the model generated by the model learning unit 23. Here, when a plurality of models are generated, the temperature prediction unit 24 selects a model generated by a data set with similar values of processed operation performance information (x _t,1 to x _{t, j} ). It is preferable to do so. The temperature prediction unit 24 transmits the calculated temperature of the combustion chamber 2 or the carbonization chamber 3 at a future time to the furnace temperature control unit 25. Thereby, the furnace temperature control process proceeds to the process of step S4.

In the process of step S4, the furnace temperature control section 25 controls the temperature of the combustion chamber 2 or the carbonization chamber 3 to reach the target temperature based on the temperature of the combustion chamber 2 or the carbonization chamber 3 calculated by the temperature prediction section 24. Calculate the amount of heat supplied. The amount of heat to be supplied is determined so that the difference between the temperature (target temperature) of the combustion chamber 2 or the carbonization chamber 3 set at the predicted time (t+u) and the predicted temperature value is minimized. Here, in order to further improve accuracy, the relationship between the amount of heat supplied and the amount of temperature change is modeled based on the mass and specific heat of the fuel gas in the combustion chamber 2, the coal in the carbonization chamber 3, and the substances related to the furnace wall. may be used in calculations. Further, the relationship between the amount of heat supplied and the amount of temperature change may be modeled based on the relationship between the amount of change in the amount of heat supplied and the amount of change in the temperature of the combustion chamber 2 or the carbonization chamber 3 in past performance data.

The furnace temperature control unit 25 may calculate the opening degrees of the regulating valves 5 and each kiln regulating valve 6 to satisfy the required amount of heat to be supplied. Here, in order to further improve accuracy, the relationship between the regulating valve opening and the fuel gas flow rate may be modeled based on physical laws and used for calculation. In addition, the relationship between the regulating valve opening and the amount of heat supplied or the amount of change in temperature is determined based on the relationship between the opening of the regulating valve and the amount of heat supplied to the combustion chamber 2 or the carbonization chamber 3 or the amount of change in temperature in past performance data. may be modeled. The furnace temperature control unit 25 transmits the calculated amount of supplied heat or the adjustment valve opening to the output device 26. Thereby, the furnace temperature control process proceeds to the process of step S5.

In the process of step S5, the output device 26 transmits the amount of heat to be supplied or the opening degree of the regulating valve calculated by the furnace temperature control section 25 to the control terminal 10. The control terminal 10 adjusts the opening degrees of the regulating valve 5 and the kiln-specific regulating valve 6 based on the information transmitted from the output device 26. Further, the information calculated by the furnace temperature control device 20 is also transmitted to the display device 30 and displayed on the screen. Therefore, it is also possible for the operator to adjust the opening degrees of the adjustment valve 5 and the individual kiln adjustment valve 6 based on the displayed calculation results. This completes the furnace temperature control process. Here, step S4 and step S5 may be referred to as a furnace temperature control step.

As described above, the furnace temperature control device 20, the furnace temperature control method, and the coke manufacturing method according to the present embodiment have a plurality of furnace temperature fluctuation factors and the adjacent combustion chamber 2 and coking chamber 3 through the above configuration and steps. Using a model that takes into account the effects of Therefore, furnace temperature fluctuations due to continuous fluctuations of a plurality of furnace temperature fluctuation factors of the coke oven 1 can be accurately predicted, and highly accurate furnace temperature control is possible.

(Example)
Hereinafter, the effects of the present disclosure will be specifically explained based on Examples, but the present disclosure is not limited to the contents of the Examples.

Predictions will be made regarding the case where the temperature of the carbonization chamber 3 is predicted using the apparatus and method described in the above embodiment (example) and the case where the temperature of the carbonization chamber 3 is predicted using the conventional prediction method (conventional example). Accuracy was evaluated. In the example (invention example), the time for determining the predicted future time (u above) was 12 hours, and the retroactive time (i above) was 17 hours. Furthermore, the temperature of the target carbonization chamber 3 and the two adjacent carbonization chambers 3 and the amount of fuel gas supplied to the two combustion chambers 2 adjacent to the target carbonization chamber 3 were used as operational performance information. Operation performance information was entered every hour. A recurrent neural network model having long and short-term memory structures in the middle layer was used as a model in the example. In the conventional example, a model was used that predicts the temperature of the carbonization chamber 3 after 12 hours based on the past change rate for 17 hours regarding the temperature of the target carbonization chamber 3. 3 and 4 are diagrams plotting actual values and predicted values of the temperature of the carbonization chamber 3 after 12 hours. In the conventional example shown in FIG. 3, there are times when the upper and lower limit fluctuations in the actual value do not match the fluctuations in the predicted value. In the example shown in FIG. 4, it is possible to accurately track and predict fluctuations in actual values.

Further, FIGS. 5 and 6 are diagrams plotting the relationship between actual values and predicted values at each time. In the conventional example shown in FIG. 5, the error between the actual value and the predicted value was large, and the root mean square error was 1.62. In the example of FIG. 6, the error between the actual value and the predicted value was small, and the root mean square error was 1.06. It has been confirmed that the apparatus and method described in the above embodiments can predict future furnace temperature fluctuations with high accuracy, thereby improving furnace temperature control accuracy.

Although the embodiments of the present disclosure have been described based on the drawings and examples, it should be noted that those skilled in the art can easily make various changes or modifications based on the present disclosure. It should therefore be noted that these variations or modifications are included within the scope of this disclosure. For example, the functions included in each component or each step can be rearranged to avoid logical contradictions, and multiple components or steps can be combined or divided into one. It is. Embodiments according to the present disclosure can also be realized as a program executed by a processor included in the device or a storage medium recording the program. It is to be understood that these are also encompassed within the scope of the present disclosure.

1 Coke oven 2 Combustion chamber 3 Carbonization chamber 4 Gas main pipe 5 Adjustment valve 6 Adjustment valve for each kiln 10 Control terminal 20 Furnace temperature control device 21 Input device 22 Database 23 Model learning section 24 Temperature prediction section 25 Furnace temperature control section 26 Output device 30 Display device

Claims (9)

A furnace temperature control device for controlling the temperature of the combustion chamber or the carbonization chamber in a coke oven in which combustion chambers and carbonization chambers are alternately connected to form a furnace group,
an input device into which operational performance information including at least actual temperature values of the combustion chamber and the carbonization chamber and actual values of the amount of heat supplied to the combustion chamber and the carbonization chamber is input;
A model representing the relationship between time-series data of the operation performance information that goes back a certain period of time from a certain time and the temperature of the combustion chamber or the carbonization chamber after an arbitrarily determined predetermined time from the certain time, and the operation performance information. a temperature prediction unit that calculates a temperature prediction value that is the temperature of the combustion chamber or the carbonization chamber after an arbitrarily determined predetermined time based on the temperature;
A furnace temperature control device, comprising: a furnace temperature control unit that calculates an amount of heat to be supplied to the combustion chamber or the carbonization chamber so as to reach a target temperature based on the calculated temperature prediction value. The operation performance information adjusts the actual weight of the coal charged into the coking chamber, the actual moisture content, the actual time elapsed from charging, and the amount of fuel gas supplied to the combustion chamber. The furnace temperature control device according to claim 1, further comprising: an actual value of the opening degree of the regulating valve. The temperature prediction unit is configured to calculate the predicted temperature value based on the operation performance information regarding the combustion chamber or the carbonization chamber, which is a target for calculating the temperature prediction value, as well as the operation performance information regarding the adjacent combustion chamber and the carbonization chamber. The furnace temperature control device according to claim 1 or 2, wherein the temperature prediction value is calculated by using the temperature prediction value. The temperature prediction unit calculates, for each cycle of temperature change of the combustion chamber or the carbonization chamber, the time series data of the operation performance information that goes back from a certain time by a period of time longer than the cycle of temperature change of the combustion chamber or the carbonization chamber. The characteristics of the time-series data of the operation performance information for each division, and the temperature of the combustion chamber or the carbonization chamber after a period of change in the temperature of the combustion chamber or the carbonization chamber from the certain time. The furnace according to claim 1 or 2 , wherein a predicted temperature value that is a temperature of the combustion chamber or the carbonization chamber after a period of change in temperature of the carbonization chamber is calculated based on a model representing the relationship and the operation performance information. Temperature control device. A furnace temperature control method for controlling the temperature of the combustion chamber or the carbonization chamber in a coke oven in which combustion chambers and carbonization chambers are alternately connected to form a furnace group, the method comprising:
an input step in which operational performance information including at least actual temperature values of the combustion chamber and the carbonization chamber and actual values of the amount of heat supplied to the combustion chamber and the carbonization chamber is input;
A model representing the relationship between time-series data of the operation performance information that goes back a certain time from a certain time and the temperature of the combustion chamber or the carbonization chamber after an arbitrarily determined predetermined time from the certain time, and the operation performance information. a temperature prediction step of calculating a temperature prediction value that is the temperature of the combustion chamber or the carbonization chamber after an arbitrarily determined predetermined time based on the temperature;
A furnace temperature control method, comprising a furnace temperature control step of calculating an amount of heat to be supplied to the combustion chamber or the carbonization chamber so as to reach a target temperature based on the calculated temperature prediction value. The operation performance information adjusts the actual weight of the coal charged into the coking chamber, the actual moisture content, the actual time elapsed from charging, and the amount of fuel gas supplied to the combustion chamber. 6. The furnace temperature control method according to claim 5, further comprising: an actual value of the opening degree of the regulating valve. The temperature prediction step is based on the operation performance information about the combustion chamber or the carbonization chamber that is nearby, in addition to the operation performance information about the combustion chamber or the carbonization chamber for which the temperature prediction value is calculated. The furnace temperature control method according to claim 5 or 6, wherein the predicted temperature value is calculated. The temperature prediction step is performed for each cycle of temperature change in the combustion chamber or the carbonization chamber, with respect to the time series data of the operation performance information that goes back from a certain time by a period longer than the cycle of temperature change in the combustion chamber or the carbonization chamber. The characteristics of the time-series data of the operation performance information for each division, and the temperature of the combustion chamber or the carbonization chamber after a period of change in the temperature of the combustion chamber or the carbonization chamber from the certain time. The furnace according to claim 5 or 6 , wherein a predicted temperature value, which is the temperature of the combustion chamber or the carbonization chamber after a period of change in temperature of the carbonization chamber, is calculated based on a model representing the relationship and the operation performance information. Temperature control method. Adjustment of the amount of fuel gas supplied to the combustion chamber or the amount of fuel gas supplied based on the amount of heat supplied to the combustion chamber or the carbonization chamber calculated by the furnace temperature control method according to claim 5 or 6. A coke production method that produces coke by adjusting the valve opening.

Images