JP5900027B2 - Method and apparatus for estimating furnace temperature distribution - Google Patents

Description

  The present invention relates to a method and an apparatus for estimating an in-furnace temperature distribution in a formed coke production facility that produces formed coke typified by ferro-coke.

Ferro-coke produced by dry distillation of a molded product formed by mixing carbonaceous material (coal) and iron ore increases the CO ₂ reactivity of coke in ferro-coke by the catalytic effect of reduced iron ore. It is possible to reduce the ratio of the reducing material by the accompanying decrease in the temperature of the heat preservation zone. As a technique for producing such ferro-coke, there is known a method of carbonizing a molded product of coal and iron ore in a vertical carbonization furnace.

  Molded coke production equipment that produces molded coke as typified by ferro-coke as described above is charged with uncooked coking coal that has been kneaded and molded from the top of a vertical dry distillation furnace. Formed coke is discharged from the furnace bottom through a dry distillation process and a cooling process. The temperature history of the coking coal during the preheating, dry distillation, and cooling processes of this forming coke production facility has a large effect on the quality of the forming coal, so that the temperature distribution in the furnace height direction is properly controlled. It is important to be kept.

  For example, Patent Document 1 discloses a ferro-coke manufacturing method and manufacturing equipment that appropriately controls the temperature distribution in a vertical dry distillation furnace while using the cooling gas used in the cooling process as a low-temperature gas used in the preheating process. Is described.

JP 2011-057970 A

  However, since the number of temperature sensors that can be installed is limited in the molded coke manufacturing facility as described above, the temperature inside the furnace where the temperature sensor is not installed is unknown. As a result, even if there was a means for appropriately controlling the temperature distribution in the furnace, it was not possible to accurately control the temperature distribution in the height direction of the dry distillation furnace.

  The present invention has been made in view of the above, and an object of the present invention is to provide an in-furnace temperature distribution estimation method and estimation apparatus for estimating the entire in-furnace temperature distribution from temperature measurement by a limited number of temperature sensors. It is to provide.

  In order to solve the above-mentioned problems and achieve the object, the method for estimating the temperature distribution in the furnace of the present invention is based on the temperature in the furnace that estimates the temperature distribution in the height direction of the furnace of the formed coke production facility based on a physical model. In the distribution estimation method, a measurement step of acquiring temperature information at a specific position in the furnace height direction of the molded coke production facility by a temperature sensor, and calculating a temperature distribution by changing parameters in the physical model, A population creating step for creating a population having the temperature distribution as a member, and a principal component analysis for calculating a principal component vector by performing principal component analysis on the population having the temperature distribution created by the population creating step as members. A temperature component estimated based on the physical model by a linear combination of a component analysis step and a principal component vector calculated in the principal component analysis step. Characterized in that it comprises a step of correcting an error between the temperature information of the specific position obtained by the measuring step.

  In order to solve the above-described problems and achieve the object, the in-furnace temperature distribution estimating apparatus of the present invention estimates the temperature distribution in the height direction of the in-furnace of the formed coke production facility based on a physical model. In the distribution estimation apparatus, by measuring the temperature distribution by changing the parameters in the physical model by measuring means for acquiring temperature information of a specific position in the furnace height direction of the molded coke production facility by a temperature sensor, A population creating means for creating a population having the temperature distribution as a member, and a principal component analysis for calculating a principal component vector by performing principal component analysis on the population having the temperature distribution created by the population creating means as members. The temperature distribution estimated based on the physical model by the linear combination of the component analysis means and the principal component vector calculated by the principal component analysis means and the measurement means Characterized in that it comprises a correcting means for correcting the error between the acquired temperature information of the specific position.

  The in-furnace temperature distribution estimation method and estimation apparatus according to the present invention can estimate the entire in-furnace temperature distribution from temperature measurements by a limited number of temperature sensors.

FIG. 1 is a schematic diagram showing a configuration of a formed coke manufacturing facility to which a method for estimating a furnace temperature distribution according to an embodiment of the present invention is applied. FIG. 2 is a block diagram showing a temperature distribution correction method using a physical model as a basis in the in-furnace temperature distribution estimation method according to the embodiment of the present invention. FIG. 3 is a diagram showing the relationship between the tuyere and temperature distribution of a vertical dry distillation furnace used for explaining data assimilation in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention. FIG. 4 is a graph showing an example of an installation position of a temperature sensor used for explaining data assimilation in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention. FIG. 5 is a graph showing an example of the change amount of the gas temperature distribution and the change amount of the gas temperature distribution corresponding to the first principal component to the third principal component. FIG. 6 is a block diagram illustrating a configuration example of an in-furnace temperature distribution estimation apparatus that realizes the in-furnace temperature distribution estimation method according to the embodiment of the present invention. FIG. 7 is a flowchart showing a method of estimating the furnace temperature distribution according to the embodiment of the present invention. FIG. 8 is a graph showing an estimation state of gas and solid temperature distributions when data assimilation is applied in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention. FIG. 9 is a graph comparing the degree of divergence from the true temperature distribution with and without applying data assimilation in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention.

  Hereinafter, an in-furnace temperature distribution estimation method and estimation apparatus according to an embodiment of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

[Molded coke production equipment]
FIG. 1 is a schematic diagram showing a configuration of a formed coke manufacturing facility to which a method for estimating a furnace temperature distribution according to an embodiment of the present invention is applied. As shown in FIG. 1, a formed coke production facility 1 to which a method for estimating an in-furnace temperature distribution according to an embodiment of the present invention is mainly used as a vertical dry distillation furnace 2 that produces dry coke by carbonizing formed coal. It is provided as a component, and further includes other components attached to the vertical distillation furnace 2.

  The vertical carbonization furnace 2 is separated into a low temperature carbonization zone, a high temperature carbonization zone, and a cooling zone in order from the top of the vertical carbonization furnace 2, and carbonized coal is carbonized by the low temperature carbonization zone and the high temperature carbonization zone. Cool the dry-formed coke. A low temperature gas blowing tuyere 3a is provided between the low temperature carbonization zone and the high temperature carbonizing zone, and the low temperature gas distillation zone and the high temperature carbonization zone are separated by blowing low temperature gas from the low temperature gas blowing tuyere 3a. High temperature gas blowing tuyere 3b and 3d are provided between the high temperature carbonization zone and the cooling zone. By blowing high temperature gas from the high temperature gas blowing tuyere 3b and 3d, the high temperature carbonization zone and the cooling zone are separated. The

  A cooling gas blowing tuyere 3c is provided in the cooling zone, and a cooling zone is formed in the vertical dry distillation furnace 2 by the cooling gas blown from the cooling gas blowing tuyere 3c.

  An in-furnace gas discharge port 4 for discharging the in-furnace gas is provided at the top of the vertical carbonization furnace 2. The exhaust gas pipe connected to the furnace gas discharge port 4 includes a spray tower 5 as a first gas cooling device, a gas cooler 6 as a second gas cooling device, and dust in the furnace gas (mainly tar mist). An electrostatic precipitator 7 is connected to remove the. The in-furnace gas that has passed through the spray tower 5, the gas cooler 6, and the electric dust collector 7 is recovered as generated gas and used as fuel after purification and desulfurization. Furthermore, a part of the recovered generated gas is reused as a low temperature gas, a high temperature gas, and a cooling gas blown into the vertical dry distillation furnace 2. The liquid component in the furnace gas is recovered by the spray tower 5 and the gas cooler 6, separated into safety water and tar, and stored in the safety water tank 8 a and the tar tank 8 b.

  Part of the generated gas recovered as described above is reused for the low temperature gas, the high temperature gas, and the cooling gas blown into the vertical distillation furnace 2. The generated gas heated by the high-temperature gas heating furnace 9 is guided to the high-temperature gas blowing tuyere 3b, 3d and blown into the vertical dry distillation furnace 2 from the high-temperature gas blowing tuyere 3b, 3d. The generated gas heated by the low temperature gas heating furnace 10 is guided to the low temperature gas blowing tuyere 3a and blown into the vertical dry distillation furnace 2 from the low temperature gas blowing tuyere 3a. Furthermore, a part of the recovered generated gas is led to the cooling gas blowing tuyere 3c without being heated, and is blown into the vertical dry distillation furnace 2 from the cooling gas blowing tuyere 3c.

  Furthermore, a coal charcoal inlet 12 is provided near the top of the vertical carbonization furnace 2, and a discharge port 13 for cutting out the formed coke is provided at the bottom of the vertical carbonization furnace 2. The coking coal carbonized in the vertical carbonization furnace 2 is charged from the inlet 12 and filled into the vertical carbonization furnace 2, and the molded coke that has been carbonized from the discharge port 13 is cut out at a constant speed. As a result, the coal is left in the vertical distillation furnace 2 for a predetermined time.

[Physical model of furnace temperature distribution]
As one example of a physical model of the temperature distribution in the vertical distillation furnace 2 of the forming coke production facility 1 to which the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention is applied, the physical model includes a material balance and a heat balance. A method for estimating the furnace temperature distribution according to the embodiment of the present invention will be described using a physical model.

The mass balance used in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention is a physical model in which a part of the solid is gasified as in the following formulas (1) and (2).

On the other hand, as shown in the following formulas (3) and (4), the heat balance is composed of heat exchange between the gas and the solid, heat exchange between the gas and the atmosphere around the facility, and reaction heat.

The following formula (5) is obtained by discretizing this by an appropriate method (for example, “Numerical analysis of heat transfer and flow by computer” by Suhas V. Patanker) (Morikita Publishing). In addition, when calculating | requiring following formula (5), the discretization by the complete implicit method is performed.

T _s (k) and T _g (k) are vector quantities representing the temperature distributions of the solid and gas in the vertical direction of the vertical distillation furnace 2 at time step k, respectively. The number of elements corresponds to the number of meshes for discretization. In the examples used below, the number of meshes was 120. U (k) means the manipulated variable, and in this model, is the temperature of the gas blown into the vertical dry distillation furnace 2 from each tuyere. In addition, the symbol used in said Formula (1)-(5) means the following.

[Model assimilation]
The method for estimating the in-furnace temperature distribution according to the embodiment of the present invention performs, for example, a virtual true value simulation under a condition different from the physical model based on the physical model of the in-furnace temperature distribution. And the temperature distribution by the basic physical model is correct | amended using the partial information of this virtual true value simulation. FIG. 2 is a block diagram showing a temperature distribution correction method using a physical model as a basis in the in-furnace temperature distribution estimation method according to the embodiment of the present invention. In the following description, correction of the temperature distribution based on this basic physical model is referred to as “data assimilation” in the sense of assimilating data and the model.

Hereinafter, in the explanation of data assimilation in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention, the vertical dry distillation furnace 2 is divided into four, and the vertical dry distillation furnace 2 is provided with six temperature sensors. Examples are used. FIG. 3 is a diagram showing four sections of the vertical dry distillation furnace 2 used for explaining data assimilation in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention. The four sections referred to here are the zones of the tuyere shown in FIG. 3 (or FIG. 1) (low temperature gas blowing tuyere 3a, high temperature gas blowing tuyere 3b, 3d, cooling gas blowing tuyere 3c). That is, the zones 1 to 4 are named from the bottom. The heat exchange coefficients between the gas solids in the zones 1 to 4 are α ₁ , α ₂ , α ₃ , and α ₄ , and the solid specific heat is C _s1 , C _s2 , C _s3 , and C _s4 . Incidentally, without distinguishing between each zone 1-4, overall gas solid thermal exchange coefficient of the vertical carbonization furnace 2 is alpha, the solid specific heat is C _s.

  FIG. 4 is a graph showing an example of an installation position of a temperature sensor used for explaining data assimilation in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention. That is, in the explanation of data assimilation in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention, it is assumed that six temperature sensors are provided at the positions indicated by circles in FIG. That is, in the following description, data assimilation is performed on the assumption that only the temperature at the position of the circle in FIG. 4 can be measured.

In this example of data assimilation, principal component analysis is used. Therefore, data serving as a population for performing principal component analysis is prepared. In order to prepare this data, data on the temporal change of the temperature distribution when the heat exchange coefficient α between the gas and solid and the solid specific heat C _s4 are changed as a physical model under standard operating conditions is created. In the data population used in this data assimilation example, the gas-solid heat exchange coefficient α is changed from 0.7 times to 1.3 times the original numerical value, and the solid specific heat C _s4 is changed to 0.8 from the original numerical value. It was changed from double to 1.2 times. In the population whose members are the temperature distribution data used in this data assimilation example, only the solid specific heat C _s4 in the lowest zone is changed. The reason is that the solid in the lowermost zone is coke after carbonization, and the specific heat is expected to vary greatly depending on the degree of carbonization.

In the case where the gas-solid heat exchange coefficient α and the solid specific heat C _s4 of the above parameters were changed with time in a sinusoidal manner with different periods, the simulation was performed for 1000 steps, and the temperature distribution of the gas and solid at each position was recorded. A data matrix X is generated. The number of rows of the data matrix X is the number of time steps (that is, 1000), and the number of columns is twice the number of meshes N in the simulation because it includes the temperature distribution of gas and solid.

A principal component analysis is performed on the data matrix X created as described above. Specifically, in this data assimilation example, three eigenvalue vectors of the covariance matrix X ′ * X of this data matrix are selected in order from the large eigenvalue. Since the selected eigenvector is suitable as a factor for explaining the change in the temperature distribution of the gas and the solid when the physical parameter is changed as described above, the three selected eigenvectors are a ₁ , a _2, is defined as _{a 3.} Note that the number of elements of the vectors a ₁ , a ₂ , and a ₃ is a vector in which elements corresponding to the change amount of the gas temperature distribution and the change amount of the solid temperature distribution are vertically arranged. It is equal to twice the number N. FIG. 5 is a graph showing an example of the change amount of the gas temperature distribution and the change amount of the gas temperature distribution corresponding to the first principal component to the third principal component.

Next, let us consider correcting the error between the observed value and the simulation value by the linear combination of a ₁ , a ₂ , and a ₃ defined as described above.

Which are estimates of the target, the temperature distribution of the true gas and solid vertical carbonization furnace 2, and x _tr as the vertical vector of twice as many elements in the same manner as described above the number of meshes. On the other hand, the temperature distribution of the gas and the solid as a result of the simulation by the physical model is similarly set to x _sim as a vertical vector having the number of elements twice the number of meshes. Also, let C be a matrix that represents which temperature can be observed in the temperature distribution of gas and solid. That is, the matrix C is a matrix in which the number of temperature sensors is the number of rows and 2N, which is twice the number of meshes, is the number of columns. Among the elements of the matrix C, the i-th temperature sensor indicates the temperature at the position j. When measuring, C [i] [j] = 1, and other elements are set to zero.

When defined as above, if the temperature distribution measured by the temperature sensor is y _obs and the temperature distribution of the gas and solid as a result of the simulation by the physical model at the position of this temperature sensor is y _sim , the following equation holds: .

Hereinafter, a method of correcting the error between y _obs and y _sim by a linear combination of a ₁ , a ₂ , and a ₃ will be described.

The coefficients of the terms in the linear combination of a ₁ , a ₂ , and a ₃ are w ₁ , w ₂ , and w ₃ , added to the gas and solid temperature distribution x _sim as a result of the simulation by the physical model as a correction term, and the matrix C A value obtained by multiplying by y is _a calculation correction value. Then, the calculated correction value and y _a is represented by the following expression. However, in the following formula, A is a matrix notation in which a ₁ , a ₂ , and a ₃ are arranged horizontally, and w is a vector notation in which w ₁ , w ₂ , and w ₃ are arranged vertically.

The coefficients w _1, w _2, w _3, such as the difference between the measured temperature distribution y _obs by the calculated correction value y _a temperature sensor is minimized, obtained by the least squares method. Then, the temperature distribution x _a after data assimilation is calculated by adding it as a correction term to the gas and solid temperature distribution x _{sim as} a result of the simulation by the physical model. The following formula expresses this calculation by a mathematical formula.

As can be seen from the above calculation, the temperature distribution x _a after data assimilation is the gas and solid temperature distribution x _{sim as} a result of the simulation by the physical model, the three eigenvectors a ₁ , a ₂ and a _3, and the gas and solid. Is calculated using a matrix C representing which temperature can be observed in the temperature distribution and temperature distribution _yobs measured by the temperature sensor. That is, based on the above calculation, it was shown that the temperature distribution by the physical model can be corrected by the physical model from the temperature measurement by a limited temperature sensor.

[Estimation device for furnace temperature distribution]
FIG. 6 is a block diagram showing a configuration example of an in-furnace temperature distribution estimation apparatus that realizes the in-furnace temperature distribution estimation method based on the data assimilation described above. However, the implementation of the present invention is not limited to the estimation apparatus having the configuration described below.

  As shown in FIG. 6, an in-furnace temperature distribution estimation device 100 that realizes the in-furnace temperature distribution estimation method according to the embodiment of the present invention mainly includes an information processing device 101, an input device 102, and an output device 103. Using the temperature information at a specific position in the furnace height direction obtained from the molded coke manufacturing facility 1 as a component, the entire furnace temperature distribution in the furnace height direction is estimated.

  The information processing apparatus 101 can be configured by a general-purpose information processing apparatus such as a personal computer or a workstation, and includes a RAM 111, a ROM 112, and a CPU 113. The RAM 111 temporarily stores control programs and control data related to processing executed by the CPU 113 and functions as a working area for the CPU 113. The ROM 112 stores an estimation program 112a that realizes the method for estimating the furnace temperature distribution according to the embodiment of the present invention, a control program that controls the operation of the entire information processing apparatus 101, and control data. The CPU 113 controls the overall operation of the information processing apparatus 101 according to the estimation program 112a and the control program stored in the ROM 112. The input device 102 includes input devices such as a keyboard, a mouse pointer, and a numeric keypad, and is operated when inputting various types of information to the information processing device 101. The output device 103 is configured by an output device such as a display device or a printing device, and outputs various processing information of the information processing device 101.

  The estimation program 112a stored in the ROM 112 is a series of commands configured to cause the information processing apparatus 101 to execute the method for estimating the furnace temperature distribution according to the above-described embodiment of the present invention. The estimation apparatus 100 for the in-furnace temperature distribution according to the command uses the temperature information of the specific position in the height direction of the furnace acquired from the molded coke manufacturing facility 1, and the entire in-furnace temperature distribution in the height direction of the furnace. To estimate.

[Method for estimating furnace temperature distribution]
FIG. 7 is a flowchart showing a method of estimating the furnace temperature distribution according to the embodiment of the present invention. Hereinafter, the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention will be described with reference to a block diagram showing a configuration example of the apparatus for estimating the temperature distribution in the furnace that realizes the method for estimating the temperature distribution in the furnace described above. However, the implementation of the in-furnace temperature distribution estimation method of the present invention is not limited by the in-furnace temperature distribution estimation apparatus.
FIG. 7 is a flowchart showing a method of estimating the furnace temperature distribution according to the embodiment of the present invention. As shown in FIG. 7, in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention, the information processing apparatus 101 uses the temperature sensor provided in the molded coke manufacturing facility 1 to increase the temperature inside the furnace of the molded coke manufacturing facility 1. The process starts from acquiring temperature information at a specific position in the height direction (step S1).

  Next, the information processing apparatus 101 creates a data population by changing the parameters in the physical model and calculating the temperature distribution (step S2). For example, the information processing apparatus 101 calculates the temperature distribution by changing the heat exchange coefficient between the gas solids and the solid specific heat, which are parameters of the physical model described in the above formulas (1) to (4), thereby calculating the data population. Can be created.

  On the other hand, the information processing apparatus 101 performs principal component analysis on the population of data created in step S2 (step S3). For example, in this principal component analysis, three vectors indicating the change amount of the gas temperature distribution and the change amount of the gas temperature distribution corresponding to the first principal component to the third principal component are selected.

Finally, the information processing apparatus 101 corrects the error between the temperature distribution estimated based on the physical model and the temperature information at the specific position acquired in step S1 by the linear combination of the principal component vectors calculated in step S3. (Step S4). This correction method expresses a correction term for an error between the estimated temperature distribution and the temperature information at the specific position by linear combination of the principal component vectors calculated in step S3 as in the above equation (6). by calculation according to equation (7) is carried out by calculating the temperature distribution x _a later data assimilation.

〔simulation〕
Next, a simulation for verifying the effect of data assimilation in the method for estimating the furnace temperature distribution according to the embodiment of the present invention will be described. In the simulation described below, an example in which the gas-to-atmosphere heat exchange coefficient is multiplied by 0.7 at t = 100 and the solid specific heat is multiplied by 0.8 at t = 200 is assumed as a true value simulation (time t is time). This is the number of steps. In this simulation, one step is 20 minutes). On the other hand, the physical model remains the original value. Naturally, under this simulation condition, there is a discrepancy between the true temperature distribution and the temperature distribution based on the physical model unless correction is made.

  FIG. 8 is a graph showing the estimation status of the temperature distribution of gas and solid when data assimilation is applied in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention described above. In FIG. 8, the dotted line shows the temperature distribution of the gas and solid when the data assimilation is applied in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention, and the broken line shows the gas and solid when the data assimilation is not applied. The temperature distribution is shown, and the solid line shows the temperature distribution of gas and solid by true value simulation. According to FIG. 8, the case where the data assimilation in the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention is applied is closer to the true value than the case where the data assimilation is not applied. I understand.

  Further, FIG. 9 shows the degree of deviation from the true temperature distribution (average estimation error of gas and solid temperature) with and without data assimilation in the method for estimating the furnace temperature distribution according to the embodiment of the present invention. ). As shown in FIG. 9, when the data assimilation in the method for estimating the in-furnace temperature distribution according to the embodiment of the present invention is applied, the temporal change in the true temperature distribution is not as compared with the case where the data assimilation is not applied. It can be understood that estimation can be performed more accurately with respect to.

  As mentioned above, the estimation method of the temperature distribution in the furnace which concerns on embodiment of this invention is the measurement step which acquires the temperature information of the specific position of the height direction in the furnace of the said shaping | molding coke manufacturing equipment 1 with a temperature sensor, and a physical model. By calculating the temperature distribution by changing parameters, a population creation step for creating a population having these temperature distributions as members, and a population having the temperature distribution created by the population creation step as members The principal component analysis step for calculating the principal component vector by performing principal component analysis and the temperature distribution estimated based on the physical model and the measurement step are obtained by linear combination of the principal component vector calculated by the principal component analysis step. It includes a correction step that corrects errors from temperature information at a specific position. It has been shown that it is possible to estimate the furnace temperature distribution of the body.

  Furthermore, the method for estimating the temperature distribution in the furnace according to the embodiment of the present invention can create a population of data by calculating the temperature distribution by changing the heat exchange coefficient between the gas solid and the solid specific heat in the physical model. preferable.

DESCRIPTION OF SYMBOLS 1 Molded coke manufacturing equipment 2 Vertical-type carbonization furnace 3a Low temperature gas blowing tuyere 3b High temperature gas blowing tuyere 3c Cooling gas blowing tuyere 3d High temperature gas blowing tuyere 4 Furnace gas discharge 5 Spray tower 6 Gas cooler 7 Electric dust collector 8a Aqueous tank 8b Tar tank 9 High-temperature gas heating furnace 10 Low-temperature gas heating furnace 12 Loading port 13 Discharge port 100 Estimating device for furnace temperature distribution 101 Information processing device 102 Input device 103 Output device 111 RAM
112 ROM
112a estimation program 113 CPU

Claims (4)

The temperature distribution in the vertical direction of a vertical carbonization furnace in a molded coke production facility is shown in terms of the material balance related to gasification of solids, heat exchange between gas and solids, heat exchange between gas and air around the equipment, and heat related to reaction heat. In the estimation method of the furnace temperature distribution estimated based on the physical model consisting of the balance ,
A measurement step of acquiring temperature information of a specific position in the furnace height direction of the molded coke manufacturing facility by a temperature sensor;
A population creation step of creating a population having the temperature distribution as a member by calculating a temperature distribution by changing a heat exchange coefficient between the gas solid and the solid specific heat, which are parameters in the physical model,
A principal component analysis step of calculating a principal component vector by performing principal component analysis on the population having the temperature distribution created by the population creation step as a member;
Based on the physical model, a correction term obtained by linear combination that determines a coefficient for each principal component vector calculated by the principal component analysis step and multiplies each coefficient and the principal component vector and adds them. By calculating the corrected temperature distribution by adding to the estimated temperature distribution, the temperature at the position corresponding to the specific position in the temperature distribution estimated based on the physical model and the measurement step acquired And a correction step of correcting an error from temperature information at a specific position. The correction step calculates a calculation correction value obtained by adding the correction term to the temperature distribution estimated based on the physical model, and an error between the calculation correction value and the temperature information at the specific position acquired by the measurement step. 2. The method for estimating an in-furnace temperature distribution according to claim 1, wherein each of the coefficients is determined so as to minimize. The population creation step, claim 1, characterized in that to create a population of the temperature distribution as a member by calculating the temperature distribution by changing the heat exchange coefficient and solid specific heat between the gas solids in the physical model or The method for estimating a furnace temperature distribution according to claim 2 . The temperature distribution in the vertical direction of a vertical carbonization furnace in a molded coke production facility is shown in terms of the material balance related to gasification of solids, heat exchange between gas and solids, heat exchange between gas and air around the equipment, and heat related to reaction heat. In the furnace temperature distribution estimation device that estimates based on the physical model consisting of the balance ,
Measuring means for acquiring temperature information of a specific position in the furnace height direction of the molded coke production facility by a temperature sensor;
A population creating means for creating a population having the temperature distribution as a member by calculating a temperature distribution by changing a heat exchange coefficient between the gas solids and a solid specific heat which are parameters in the physical model;
Principal component analysis means for calculating principal component vectors by performing principal component analysis on a population whose members are the temperature distribution created by the population creation means;
Based on the physical model, a correction term obtained by linear combination that determines a coefficient for each principal component vector calculated by the principal component analysis means and multiplies each coefficient and the principal component vector for addition. By calculating the corrected temperature distribution by adding to the estimated temperature distribution, the temperature at the position corresponding to the specific position in the temperature distribution estimated based on the physical model and the measurement unit Correction means for correcting an error from temperature information at a specific position;
A furnace temperature distribution estimation apparatus comprising:

Images