Distribution prediction device and method

Abstract

To provide a distribution prediction device which can predict at least one of a temperature distribution or a gas concentration distribution on a plant measurement cross section.SOLUTION: A distribution prediction device 22 includes a prediction part 57 and a correction part 58. The prediction part 57 predicts at least one of a temperature distribution or a gas concentration distribution of a measurement cross-section of a second plant from at least one of process data or operation data of the second plant by using a prediction model in which at least one of process data or operation data of a first plant is associated with at least one of a temperature distribution or a gas concentration distribution of a measurement cross-section of the first plant. Based on at least one of a temperature or a gas concentration of the measurement cross-section of the second plant measured by a sensor, the correction part 58 corrects at least one of the predicted temperature distribution or the gas concentration distribution.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present invention relates to a distribution prediction device and method for predicting at least one of a temperature distribution and a gas concentration distribution in a measurement cross section of a plant such as an incinerator. [Background technology]

[0002] The temperature and gas concentration in a plant such as an incinerator are non-uniform. To control combustion in the plant, it is useful to understand the temperature distribution and gas concentration distribution in the measurement cross section of the plant.

[0003] As an invention for grasping the temperature distribution and gas concentration distribution of a measurement cross section of a plant, Patent Document 1 discloses a method in which a number of sensors (laser measurement devices) are installed at the measurement cross section of the plant and the temperature distribution and gas concentration distribution of the measurement cross section are measured using the multiple laser measurement devices.

[0004] On the other hand, Patent Document 2 discloses an invention that uses a prediction model to predict the concentration of a gas discharged from a plant. That is, the invention discloses a method for predicting the concentration of a gas discharged from a plant from the process data of the plant using a prediction model in which the process data of the plant and the concentration of a gas discharged from the plant are associated with each other. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] International Publication No. 2017 / 119283 [Patent Document 2] Patent No. 6673800 Summary of the Invention [Problem to be solved by the invention]

[0006] However, the invention of Patent Document 1 has the problem that it is necessary to provide a large number of openings in the plant through which the laser passes. The large number of openings causes heat loss and requires purge air to prevent the openings from becoming clogged with ash.

[0007] In the invention of Patent Document 2, the concentration of gas discharged from a plant can be predicted using a prediction model, but the gas concentration distribution at a measurement cross section of the plant cannot be predicted.

[0008] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a distribution prediction device and method that can predict at least one of the temperature distribution and gas concentration distribution at a measurement cross section of a plant. [Means for solving the problem]

[0009] In order to solve the above problem, one aspect of the present invention is a distribution prediction device that includes: a prediction unit that predicts at least one of a temperature distribution or a gas concentration distribution at a measurement cross section of a second plant from at least one of process data or operation data of the second plant using a prediction model in which at least one of process data or operation data of the first plant is associated with at least one of a temperature distribution or a gas concentration distribution at a measurement cross section of the first plant; and a correction unit that corrects at least one of the predicted temperature distribution or gas concentration distribution based on at least one of the temperature or gas concentration at the measurement cross section measured by a sensor in the second plant.

[0010] Another aspect of the present invention is a distribution prediction method comprising the steps of: predicting at least one of a temperature distribution or a gas concentration distribution at a measurement cross section of a second plant from at least one of process data or operation data of the second plant using a prediction model in which at least one of process data or operation data of the first plant corresponds to at least one of a temperature distribution or a gas concentration distribution at a measurement cross section of the first plant; and correcting at least one of the predicted temperature distribution or gas concentration distribution based on at least one of the temperature or gas concentration at the measurement cross section measured by a sensor in the second plant. [Effects of the Invention]

[0011] According to the present invention, at least one of the temperature distribution or gas concentration distribution in the measurement cross section of the second plant is predicted using a prediction model created in the first plant, so that at least one of the temperature distribution or gas concentration distribution in the measurement cross section of the second plant can be predicted without providing a large number of openings in the second plant. Furthermore, at least one of the predicted temperature distribution or gas concentration distribution in the measurement cross section of the second plant is corrected, so that errors due to differences between the first plant and the second plant can be reduced. [Brief explanation of the drawings]

[0012] [Figure 1] FIG. 2 is a vertical cross-sectional view of a second plant in which the distribution prediction device of the present embodiment is installed. [Figure 2] FIG. 2 is a vertical cross-sectional view of the first plant. [Figure 3] FIG. 1 is a diagram showing a measurement cross section of the first plant. [Figure 4] FIG. 10 is a diagram showing a measurement cross section of the second plant. [Figure 5] FIG. 2 is a functional block diagram of the model generating device. [Figure 6] FIG. 2 is a functional block diagram of a distribution prediction device. [Figure 7] 1 is a flowchart of a distribution prediction method according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0013] Hereinafter, a distribution prediction apparatus and method according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings. However, the distribution prediction apparatus and method according to the present invention may be embodied in various forms and are not limited to the embodiments described herein. The present embodiment is provided with the intention of enabling those skilled in the art to fully understand the invention by providing sufficient disclosure in the specification.

[0014] As shown in Fig. 1, a distribution prediction device 22 of this embodiment is installed in the second plant 2. As shown in Fig. 2, a prediction model creation device 23 is installed in the first plant 1. The distribution prediction device 22 of the second plant 2 predicts the temperature distribution and gas concentration distribution of the measurement cross section 42 of the second plant 2 using the prediction model created by the prediction model creation device 23 of the first plant 1. It is only necessary for the distribution prediction device 22 to predict at least one of the temperature distribution and the gas concentration distribution of the measurement cross section 42 of the second plant 2.

[0015] The second plant 2 is, for example, an operational plant. The first plant 1 is, for example, a plant of the same model as the second plant 2, or a plant whose internal fluid flow is considered to be equivalent to that of the second plant 2, such as a plant of the same model as the second plant 2, an experimental plant, or a pilot plant. In the following, an example will be described in which the first plant 1 and the second plant 2 are incinerators, but they are not limited to incinerators and may be furnaces such as boilers, melting furnaces, or blast furnaces. (Configuration of Plant 1 and Plant 2)

[0016] In this embodiment, the first plant 1 and the second plant 2 are stoker-type waste incinerators. The first plant 1 and the second plant 2 have substantially the same configuration, except that (1) a prediction model creation device 23 is installed in the first plant 1 and a distribution prediction device 22 is installed in the second plant 2, and (2) the number of sensors 26 installed in the measurement cross section 42 of the second plant 2 (see FIG. 4) is fewer than the number of sensors 26 installed in the measurement cross section 41 of the first plant 1 (see FIG. 3). The configuration of the second plant 2 shown in FIG. 1 will be described below. The components of the first plant 1 shown in FIG. 2 are assigned the same reference numerals as the components of the second plant 2.

[0017] As shown in Fig. 1, the second plant 2 includes a hopper 3 into which fuel is fed, a furnace body 4 in which the fuel is combusted, and a boiler 5 located downstream of the furnace outlet 13. The fuel is municipal solid waste, industrial waste, biomass, etc.

[0018] The waste fed into the hopper 3 is sent to the grate 7 by the pusher of the fuel supply device 6. The grate 7 moves back and forth, stirring and moving the fuel. The fuel on the grate 7 is dried by blowing primary combustion air into it by the blower 8, and then primary combustion takes place, producing combustion exhaust gas and ash.

[0019] The grate 7 includes a drying grate 7a for drying and igniting the waste, a combustion grate 7b for burning the waste, and a post-combustion grate 7c for storing and burning the unburned waste. Ash after burning falls from an ash drop port 9 and is discharged outside the furnace.

[0020] The total amount of primary combustion air supplied to the primary combustion chamber 11 from below the grate 7 is adjusted by a primary combustion air damper 10a located immediately adjacent to the blower 8. The amount of primary combustion air distributed to each wind box is adjusted by under-grate combustion air dampers 10b, 10c, and 10d located on the pipes supplying combustion air to each wind box. Multiple wind boxes and multiple under-grate combustion air dampers are also provided across the furnace, i.e., in the direction perpendicular to the plane of the paper in Figure 1.

[0021] The combustible gases generated above the drying grate 7a and the combustion grate 7b and the combustion exhaust gases generated above the post-combustion grate 7c join and mix in the secondary combustion chamber 12 at the furnace outlet 13. Air for secondary combustion is supplied to the secondary combustion chamber 12 by a blower 14, and secondary combustion of unburned fuel after primary combustion is carried out.

[0022] The total amount of secondary combustion air supplied to the secondary combustion chamber 12 is adjusted by a secondary combustion air damper 16a provided immediately adjacent to the blower 14. A plurality of blowing nozzles 15 are provided in the secondary combustion chamber 12. The distribution amount of secondary combustion air supplied to each blowing nozzle 15 is adjusted by secondary combustion air dampers 16b and 16c provided in the piping that supplies secondary combustion air to the blowing nozzles 15.

[0023] A stirring gas is supplied to the secondary combustion chamber 12 of the furnace body 4 by a blower 17. The gas in the secondary combustion chamber 12 is stirred by the stirring gas. The amount of stirring gas supplied is adjusted by a stirring gas damper 19 provided in the pipe that supplies the stirring gas to the injection nozzle 18. The type of stirring gas is not particularly limited, and may be air or a circulating gas obtained by circulating the combustion exhaust gas discharged from the furnace body 4.

[0024] A boiler 5 is installed downstream of the furnace outlet 13. Thermal energy of the combustion exhaust gas is recovered by the boiler 5 and the recovered gas is discharged to the outside through a chimney (not shown).

[0025] Thermometers for measuring the gas temperature inside the furnace body 4 are provided at multiple positions inside the furnace body 4. For example, a primary combustion chamber gas thermometer for measuring the temperature of the primary combustion chamber 11 is provided in the primary combustion chamber 11, a main flue gas thermometer is provided below the furnace outlet 13, a furnace outlet lower thermometer is provided below the furnace outlet 13, a furnace outlet middle gas thermometer is provided at the middle position of the furnace outlet 13, and a furnace outlet gas thermometer is provided downstream of the furnace outlet 13.

[0026] A boiler outlet oxygen concentration meter that measures the concentration of oxygen in the combustion exhaust gas is installed on the outlet side of the boiler 5. A gas concentration meter 31 that measures the concentrations of nitrogen oxides and carbon monoxide in the combustion exhaust gas is installed in the exhaust gas duct 30 connected to the chimney. In addition, a gas flow meter that measures the flow rate of the combustion exhaust gas is installed in the exhaust gas duct 30.

[0027] Process data such as gas temperature, concentration, and flow rate measured by various thermometers, gas concentration meter 31, boiler outlet oxygen concentration meter, and gas flow meter are transmitted to a combustion control device 21. Based on the process data, the combustion control device 21 controls the manipulated variables (e.g., fuel supply device 6, grate 7, primary combustion air damper 10a, under-grate combustion air dampers 10b, 10c, and 10d, secondary combustion air dampers 16a, 16b, and 16c, agitation gas damper 19, etc.) (e.g., fuel supply device feed rate, grate feed rate, primary combustion air volume, distribution of primary combustion air volume, secondary combustion air volume, distribution of secondary combustion air volume, agitation gas volume, etc.). The combustion control device 21 also transmits the process data and operation data, which are the manipulated variables of the manipulated variables, to a distribution prediction device 22. (Sensor installed at the measurement cross section of Plant 1)

[0028] As shown in Figure 3, a number of sensors 26 are installed at the measurement cross section 41 of the first plant 1. The sensors 26 measure temperature distribution and gas concentration distribution. The temperature distribution is the temperature distribution of a gas whose temperature is to be measured, such as water vapor. The gas concentration distribution is the concentration distribution of a gas whose concentration is to be measured, which is at least one of O2, CO, N2, NH3, NO, NOx, etc. Note that the sensors 26 only need to measure either the temperature distribution or the gas concentration distribution at the measurement cross section 41.

[0029] The sensor 26 is a laser measurement device that measures the temperature distribution and gas concentration distribution using laser light absorption spectroscopy and CT (Computed Tomography) technology.

[0030] Absorption spectroscopy is a measurement method that utilizes the property of certain gas molecules contained in the target gas absorbing light of a specific wavelength when a laser beam of a certain wavelength is irradiated onto the target gas, and the temperature and concentration dependence of the amount of absorption. The ratio of the intensity of the incident light to the transmitted light (I λ / I λ0 ) can measure temperature and gas concentration.

[0031] The sensor 26 includes a laser irradiator 26a that irradiates the measurement cross section with laser light, and a light receiver 26b that receives the laser light that has passed through the measurement cross section 41. The laser light output from the laser transmitter 24 is branched into multiple optical paths by a branching filter 25 and reaches the laser irradiator 26a. The laser irradiator 26a is a collimator lens or the like. The light receiver 26b includes a light receiving element such as a photodiode or a phototransistor, and outputs an electrical signal according to the intensity of the received laser light.

[0032] Based on the electrical signals received from each photodetector 26b, the analyzer 27 determines the average temperature and average gas concentration along the path of the laser beam on the measurement cross section 41. Furthermore, based on the electrical signals received from the multiple photodetectors 26b, the analyzer 27 uses CT technology to determine the temperature distribution and gas concentration distribution on the measurement cross section 41. Figure 3 conceptually shows an example of the determined temperature distribution using contour lines. (Sensor installed at the measurement cross section of the second plant)

[0033] As shown in Fig. 4, sensors 26 for measuring temperature and gas concentration are installed at the measurement cross section 42 of the second plant 2. The number of sensors 26 is, for example, one or two (here, one), which is fewer than the number of sensors 26 installed at the measurement cross section 41 of the first plant 1. The sensor 26 is a laser measurement device that uses absorption spectroscopy technology that utilizes laser light to determine the average temperature and average gas concentration along the path of the laser light at the measurement cross section 42. The configuration of the sensor 26 is the same as that of the sensor 26 shown in Fig. 3, so the same reference numerals are used and a description thereof will be omitted.

[0034] As the sensor 26 installed at the measurement cross sections 41 and 42, a thermocouple or a suction type gas concentration meter may be used instead of the laser measurement device. (Prediction model creation device)

[0035] 5, the prediction model creation device 23 acquires process data and operation data of the first plant 1 from the combustion control device 21 of the first plant 1, and also acquires training data of the temperature distribution and gas concentration distribution of the measurement cross section 41 of the first plant 1 from the analysis device 27. Based on these acquired learning data, the prediction model creates a prediction model that predicts the temperature distribution and gas concentration distribution of the measurement cross section 41 of the first plant 1 from the process data and operation data of the first plant 1. Note that the prediction model only needs to predict at least one of the temperature distribution or gas concentration distribution of the measurement cross section 41 of the first plant 1 from at least one of the process data or operation data of the first plant 1.

[0036] The process data is preferably process data upstream of the measurement cross section 41. The measurement cross section 41 is located after the secondary combustion chamber 12 where the gas flow is unidirectional. The process data upstream of the measurement cross section 41 is causally related to the temperature distribution and gas concentration distribution at the measurement cross section 41. In addition to or instead of data on physical quantities such as temperature, pressure, flow rate, and concentration, the process data may include image data obtained by a camera or the like used to monitor the process, or may include numerical data obtained by processing the image data.

[0037] The predictive model creation device 23 is a computer and includes a processor such as a CPU or GPU (not shown), a main storage device such as a ROM or RAM, an auxiliary storage device such as an SSD or HDD, an I / O interface, etc. The predictive model creation device 23 realizes each function provided by the predictive model creation device 23 by the processor operating in accordance with the instructions of a program loaded into the main storage device.

[0038] As shown in FIG. 5, the prediction model creation device 23 includes a data acquisition unit 51, a teacher data acquisition unit 52, and a model creation unit 53.

[0039] The data acquisition unit 51 acquires process data and operation data from the combustion control device 21 of the first plant 1. The teacher data acquisition unit 52 acquires the temperature distribution and gas concentration distribution of the measurement cross section 41 from the analysis device 27 of the first plant 1. The teacher data acquisition unit 52 may acquire the temperature distribution and gas concentration distribution of the measurement cross section 41 at the same time that a sensor or the like measures the process data, or may acquire the temperature distribution and gas concentration distribution of the measurement cross section 41 after a predetermined set time has elapsed since the time that a sensor or the like measures the process data.

[0040] The model creation unit 53 creates a prediction model by machine learning learning data in which the process data and operation data acquired by the data acquisition unit 51 are associated with the temperature distribution and gas concentration distribution of the measurement cross section 41 acquired by the teacher data acquisition unit 52. Specifically, the model creation unit 53 performs regression analysis on multiple pieces of learning data from different times and creates a regression equation for calculating the temperature distribution and gas concentration distribution of the measurement cross section 41 from the process data and operation data. The model creation unit 53 may perform machine learning using a machine learning method such as a neural network, generalized linear regression, or support vector regression. (Distribution prediction device)

[0041] The distribution prediction device 22 of the second plant 2 obtains a prediction model from the prediction model creation device 23 of the first plant 1, and predicts the temperature distribution and gas concentration distribution of the measurement cross section 42 of the second plant 2 by inputting the process data and operation data of the second plant 2 into the prediction model.

[0042] The distribution prediction device 22 is a computer and includes a processor such as a CPU or GPU (not shown), a main storage device such as a ROM or RAM, an auxiliary storage device such as an SSD or HDD, an I / O interface, etc. The distribution prediction device 22 realizes each function provided in the distribution prediction device 22 by the processor operating in accordance with the instructions of a program loaded into the main storage device.

[0043] 6, the distribution prediction device 22 includes a prediction model acquisition unit 54, a prediction model correction unit 55, a data acquisition unit 56, a prediction unit 57, and a correction unit 58. The prediction model correction unit 55 is optional and may be omitted.

[0044] The prediction model acquisition unit 54 acquires the prediction model created by the prediction model creation device 23, for example, via a communication network or a portable recording medium. The acquired prediction model is stored in, for example, an auxiliary storage device.

[0045] The data acquisition unit 56 acquires process data and operation data from the combustion control device 21 of the second plant 2.

[0046] The prediction unit 57 predicts the temperature distribution and gas concentration distribution of the measurement cross section 42 of the second plant 2 from the process data and operation data of the second plant 2 using the acquired prediction model.

[0047] The correction unit 58 acquires the temperature and gas concentration (e.g., average temperature and average gas concentration) of the measurement cross section 42 measured by the sensor 26 of the second plant 2 from the analysis device 27 of the second plant 2. The correction unit 58 corrects the predicted temperature distribution and gas concentration distribution based on the temperature and gas concentration of the measurement cross section 42 measured by the sensor 26. For example, the correction unit 58 corrects the predicted temperature distribution and / or gas concentration distribution (increases or decreases the absolute value of the distribution as a whole) so that the average temperature and / or average concentration of the measurement cross section 42 calculated from the predicted temperature distribution and / or gas concentration distribution matches the average temperature and / or average concentration of the measurement cross section 42 measured by the sensor 26. The correction unit 58 may also correct the predicted temperature distribution and / or gas concentration distribution (increases or decreases the absolute value of the distribution as a whole) so that the temperature and / or concentration at a certain point on the measurement cross section 42 calculated from the predicted temperature distribution and / or gas concentration distribution matches the temperature and / or concentration at a certain point on the measurement cross section 42 measured by the sensor 26.

[0048] The correction unit 58 outputs the corrected temperature distribution and gas concentration distribution to the combustion control device 21 of the second plant 2. The combustion control device 21 of the second plant 2 controls the manipulated variables of the respective operating terminals based on the corrected temperature distribution and gas concentration distribution.

[0049] The correction unit 58 may delete the predicted temperature distribution and gas concentration distribution when the predicted temperature distribution and gas concentration distribution deviate from the temperature and gas concentration at the measurement cross section 42 measured by the sensor 26 of the second plant 2. This can improve the prediction accuracy and reduce the risk of inappropriate control of the control element when the prediction is incorrect.

[0050] If two or more sensors 26 are installed on the measurement cross section 42 of the second plant 2, the correction unit 58 will be able to perform more diverse corrections.

[0051] The prediction model correction unit 55 corrects the prediction model when there is a discrepancy between the predicted temperature distribution and gas concentration distribution and the temperature and gas concentration at the measurement cross section 42 measured by the sensor 26 of the second plant 2. As described above, the prediction model correction unit 55 is optional and may not be provided. (Effect of distribution prediction device)

[0052] The temperature distribution and gas concentration distribution of the measurement cross section 42 of the second plant 2 are predicted using a prediction model created in the first plant 1, so the temperature distribution and gas concentration distribution of the measurement cross section 42 of the second plant 2 can be predicted without providing a large number of openings in the second plant 2.

[0053] When using a prediction model created in the first plant 1 to predict the temperature distribution and gas concentration distribution of the measurement cross section 42 of the second plant 2, experience shows that the trend of the predicted distribution (which side will have larger values ​​and which side will have smaller values) often resembles the trend of the actual distribution, but the absolute values ​​of the predicted temperature distribution and gas concentration distribution (for example, the absolute value of the average temperature of the measurement cross section, the absolute value of the average concentration of the measurement cross section, the absolute value of the temperature at a certain point on the measurement cross section, the absolute value of the concentration at a certain point on the measurement cross section, etc.) often deviate.

[0054] By correcting the predicted temperature distribution and gas concentration distribution based on the temperature and gas concentration of the measurement cross section 42 measured by the sensor 26 of the second plant 2, for example, by correcting the predicted distribution so that the absolute value matches the measurement value of the sensor 26 without changing the shape of the distribution as described above, the predicted distribution can be made closer to the actual distribution, and errors due to differences between the first plant 1 and the second plant 2 can be reduced. (Distribution prediction method)

[0055] The distribution prediction method of this embodiment is a method of predicting the temperature distribution and gas concentration distribution of the measurement cross section 42 of the second plant 2 using a prediction model created in the first plant 1.

[0056] As shown in FIG. 7, the distribution prediction method includes steps S1 to S5. In step S1, a prediction model is acquired from the prediction model creation device 23 of the first plant 1. In step S2, process data and operation data are acquired from the combustion control device 21 of the second plant 2. In step S3, the acquired prediction model is used to predict the temperature distribution and gas concentration distribution of the measurement cross section 42 of the second plant 2 from the process data and operation data of the second plant 2. In step S4, the predicted temperature distribution and gas concentration distribution are corrected based on the temperature and gas concentration of the measurement cross section 42 measured by the sensor 26 of the second plant 2. In step S5, the corrected temperature distribution and gas concentration distribution are output to the combustion control device 21 of the second plant 2. [Explanation of symbols]

[0057] 1...Plant 1 2...Second Plant 21...Combustion control device 22...Distribution prediction device 26...Sensor (laser measurement device) 41...Measurement cross section of Plant 1 42...Measurement cross section of the second plant 57…Prediction Department 58...Correction section

Claims

1. a prediction unit that predicts at least one of a temperature distribution or a gas concentration distribution at a measurement cross section of a second plant from at least one of process data or operation data of the second plant using a prediction model in which at least one of process data or operation data of the first plant is associated with at least one of a temperature distribution or a gas concentration distribution at a measurement cross section of the first plant; a correction unit that corrects at least one of the predicted temperature distribution or gas concentration distribution based on at least one of the temperature or gas concentration at the measurement cross section measured by the sensor in the second plant.

2. 2. The distribution prediction device according to claim 1, wherein the number of sensors in the second plant is smaller than the number of sensors that measure at least one of the temperature distribution or the gas concentration distribution in the measurement cross section of the first plant to create the prediction model.

3. 3. The distribution prediction device according to claim 1, wherein the sensor of the second plant and the sensor of the first plant are laser measurement devices.

4. 3. The distribution prediction device according to claim 1, wherein the corrector outputs at least one of the temperature distribution and the gas concentration distribution after the correction to a combustion control device of the second plant.

5. 3. The distribution prediction device according to claim 1, wherein the process data of the first plant and the second plant are process data on an upstream side of a measurement cross section.

6. 3. The distribution prediction device according to claim 1, wherein the correction unit deletes at least one of the temperature distribution or gas concentration distribution predicted by the prediction unit when there is a discrepancy between at least one of the predicted temperature distribution or gas concentration distribution and at least one of the temperature or gas concentration at the measurement cross section measured by the sensor of the second plant.

7. a step of predicting at least one of a temperature distribution or a gas concentration distribution at a measurement cross section of a second plant from at least one of process data or operation data of the second plant using a prediction model in which at least one of process data or operation data of the first plant is associated with at least one of a temperature distribution or a gas concentration distribution at a measurement cross section of the first plant; and correcting at least one of the predicted temperature distribution or gas concentration distribution based on at least one of the temperature or gas concentration at the measurement cross section measured by a sensor in the second plant.

Images