Concentration estimation method, concentration control method, continuous crystallization process, and concentration estimation apparatus
A model-based concentration estimation method addresses the challenge of continuous crystallization by calculating concentration estimates from measured physical quantities, ensuring accurate and real-time monitoring and control in continuous processes.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- UBE CORPORATION
- Filing Date
- 2022-08-26
- Publication Date
- 2026-06-23
Smart Images
Figure 0007878667000002 
Figure 0007878667000003 
Figure 0007878667000004
Abstract
Description
[Technical Field]
[0001] This disclosure relates to a concentration estimation method, a concentration control method, a continuous crystallization process, and a concentration estimation apparatus. [Background technology]
[0002] Crystallization processes have been conventionally known as methods for creating desired crystals by bringing a raw material liquid into a supersaturated or supercooled state through evaporation or cooling. For example, Patent Document 1 describes a batch cooling crystallization process.
[0003] In general, in crystallization processes, it is important to appropriately control the concentration of the raw materials in the raw material solution. This is because, for example, if the concentration of raw materials in the raw material solution is too low, sufficient precipitation may not occur, while if the concentration is too high, excessive precipitation may occur, making it difficult to continue the process. [Prior art documents] [Patent Documents]
[0004] [Patent Document 1] Japanese Patent Publication No. 2015-87781 [Overview of the project] [Problems that the invention aims to solve]
[0005] However, even if a concentration sensor is installed in the apparatus into which the raw material liquid is introduced, if the crystallization process is a continuous process, it may not be possible to accurately measure the raw material concentration. This is because if crystals adhere to the concentration sensor, accurate concentration measurement becomes impossible, and in a continuous process, the raw material liquid is introduced and processed continuously, so the crystals attached to the concentration sensor cannot be removed unless the process is stopped. In contrast, with a batch process, the raw material liquid is introduced and processed intermittently, so it is possible to remove the crystals attached to the concentration sensor at regular intervals.
[0006] The present disclosure has been made in view of the above points, and provides a technique for accurately estimating the concentration of a continuous process involving precipitation.
Means for Solving the Problems
[0007] A concentration estimation method according to an aspect of the present disclosure includes an acquisition procedure for acquiring a measurement value obtained by measuring a predetermined physical quantity in a continuous process involving precipitation due to evaporation of a raw material liquid, and an estimation procedure for calculating an estimated value of the concentration from the measurement value using a model for estimating the concentration of the raw material in the raw material liquid, which are executed by a computer.
Effects of the Invention
[0008] A technique for accurately estimating the concentration of a continuous process involving precipitation is provided.
Brief Description of the Drawings
[0009] [Figure 1] It is a diagram showing an example of the overall configuration of the concentration estimation system according to the present embodiment. [Figure 2] It is a diagram schematically showing an example of a continuous plant involving precipitation due to evaporation. [Figure 3] It is a diagram showing an example of the hardware configuration of the concentration estimation device according to the present embodiment. [Figure 4] It is a diagram showing an example of the functional configuration of the concentration estimation device according to the present embodiment. [Figure 5] It is a flowchart showing an example of the offline processing according to the present embodiment. [Figure 6] It is a flowchart showing an example of the online processing according to the present embodiment. [Figure 7] It is a diagram showing an example of the concentration estimation result.
Modes for Carrying Out the Invention
[0010] The following describes one embodiment of the present invention. The following describes a concentration estimation system 1 that includes a concentration estimation device 10 capable of accurately estimating the raw material concentration from various physical quantities measured in a continuous process involving precipitation (particularly a continuous process involving precipitation due to heating and evaporation of the raw material liquid).
[0011] <Example of the overall configuration of concentration estimation system 1> Figure 1 shows an example of the overall configuration of the concentration estimation system 1 according to this embodiment. As shown in Figure 1, the concentration estimation system 1 according to this embodiment includes a concentration estimation device 10, a plant 20, a control device 30, a sensor group 40, a concentration measuring instrument 50, a database 60, and a monitoring device 70. The concentration estimation device 10, the control device 30, the database 60, and the monitoring device 70 are connected to each other via a communication network 80, such as a LAN (Local Area Network).
[0012] Plant 20 is a type of plant that performs a continuous process involving precipitation by evaporation. In other words, Plant 20 is a plant facility or apparatus that produces a target substance (for example, a concentrated liquid in which the raw materials are supersaturated) by heating a continuously fed raw material liquid, accompanied by evaporation and precipitation. Specific examples of such a continuous process Plant 20 (a continuous plant involving precipitation by evaporation) will be described later. However, this embodiment can target any process as long as it involves a continuous process involving precipitation by evaporation. Typical examples of continuous processes involving precipitation by evaporation include crystallization processes in which a raw material liquid is continuously fed in, and the target crystals are obtained by heating and evaporating the raw material liquid.
[0013] Sensor group 40 is a collection of sensors that measure various physical quantities of the process (a continuous process involving precipitation by evaporation) performed by plant 20. Hereinafter, let n be the total number of sensors included in sensor group 40, and each sensor will be referred to as "sensor 401", "sensor 402", ..., "sensor 40 nThis is expressed as "[...]". Examples of physical quantities measured by the sensor group 40 include temperature, pressure, flow rate, and ambient temperature. However, these are just examples, and the physical quantities that can be measured by the sensor group 40 are not limited to these.
[0014] Below, sensor 40 i Let x be the variable representing the physical quantity being measured in (i=1,···,n). i , the measured value of that physical quantity at time t is x i We will write it as (t). Also, for simplicity, in the following, each sensor 40 i Assuming that the measurement interval (sensing period) for (i=1,···,n) is the same, we will denote it as ΔT1. ΔT1 is generally, for example, a few seconds to a few minutes.
[0015] Note that while t represents time, time is not limited to hours, minutes, and seconds; for example, it may also include the year, month, and day. Furthermore, t may not represent the time itself, but rather an index representing time (i.e., a time index that takes a non-negative integer value).
[0016] The control device 30 displays the measured values x1(t), x2(t), ..., x at each time t. n (t) is collected from sensor group 40, and the measured values x1(t), x2(t), ..., x n (t) is a device or apparatus that stores (t) in the database 60. The control device 30 can be, for example, a distributed control system (DCS) or a programmable logic controller (PLC).
[0017] The concentration measuring device 50 is a device such as a concentration meter used by the operator of the plant 20 or the like. The operator of the plant 20 or the like can sample the raw material liquid from the plant 20 as needed and measure its concentration. Hereinafter, the variable representing the concentration will be written as y, and the concentration at time t measured by the concentration measuring device 50 (that is, the measured concentration value at time t) will be written as y(t). Also, hereinafter, for simplicity, it is assumed that the operator or the like periodically samples the raw material liquid from the plant 20 and measures its concentration, and this sampling interval will be written as ΔT2. ΔT2 is arbitrarily determined by the operator of the plant 20 or the like and may be about several hours to several days, or may be about several weeks. The concentration y(t) measured by the concentration measuring device 50 at each time t is stored in the database 60.
[0018] The database 60 stores the measured values x1(t), x2(t), ···, x n (t) at each time t and the measured concentration value y(t) at each time t. That is, the database 60 stores, for example, {(x1(t), x2(t), ···, x n (t))|t = t1 + ΔT1, t1 + 2ΔT1, t1 + 3ΔT1, ···, t1 + N1ΔT1} and {y(t)|t = t2 + ΔT2, t2 + 2ΔT2, t2 + 3ΔT2, ···, t2 + N2ΔT2}. However, t1 is the measurement start time of the sensor group 40, N1 is the number of measurements of the sensor group 40, t2 is the measurement start time by the concentration measuring device 50, and N2 is the number of measurements by the concentration measuring device 50.
[0019]
[0020] As described above, ΔT1 < ΔT2 (for example, ΔT1 is about several seconds to several minutes, and ΔT2 is about several hours to several days or several weeks). Therefore, when t1 = t2 and t1 + N1ΔT1 = t2 + N2ΔT2, N1 > N2, and in the database 60, compared with the measured concentration value y(t), the measured values x1(t), x2(t), ···, x n (t) of the sensor group 40 are stored in a very large number.
[0020] The concentration estimation device 10 stores the measured concentration values y(t) for past time periods and the measured values (x1(t), x2(t), ..., x) stored in the database 60. n Using (t), we create a model f (hereinafter referred to as the concentration estimation model) for estimating the concentration at any given time. The concentration estimation model f is expressed as some function with parameters.
[0021] Below are the measured concentration values y(t) at past time points and the measured values (x1(t), x2(t), ..., x n (t)) and the corresponding (x1(t), x2(t), ..., x n Let (t), y(t)) be called the actual data, and the set of these data is {(x1(t), x2(t), ..., x n Let (t), y(t))|t∈T}. Here, T is the set of past time points for which the observed cardinality y(t) was obtained.
[0022] Note that for the measured concentration value y(t) at past time points, the measured values (x1(t), x2(t), ..., x) at the same time point are... n It is possible that (t) does not exist. In this case, the measured values (x1(t), x2(t), ..., x) are considered to be the same time as the measured concentration value y(t) (for example, a time within a certain error range). n We just need to associate (t) with its measured concentration value y(t).
[0023] Furthermore, the concentration estimation device 10 uses the current time measurement values (x1(t), x2(t), ..., x) stored in the database 60. n (t)) and concentration estimation model f are used to determine the concentration at the current time.
[0024]
number
[0025] In other words, the concentration estimation device 10 estimates the measured values (x1(t), x2(t), ..., x) at each time t for which the concentration is to be estimated. n It functions as a soft sensor that calculates the concentration estimate y^(t) at time t from (t).
[0026] The concentration estimation model f is created offline. On the other hand, the concentration estimate y^(t) is calculated online in real time. Here, "online" means that plant 20 is in operation, and "offline" means that it is unrelated to the operating status of plant 20. Therefore, the concentration estimation model f may be created, for example, when plant 20 is not in operation, or when plant 20 is in operation.
[0027] The monitoring device 70 visualizes (displays) the concentration estimate y^(t) calculated by the concentration estimation device 10 in real time. This allows the operators of the plant 20 to check the raw material concentration (estimated value) in real time during the operation of the plant 20.
[0028] The overall configuration of the concentration estimation system 1 shown in Figure 1 is an example and is not limited to this configuration. For example, the concentration estimation device 10 and the monitoring device 70 may be configured as a single unit. Also, various equipment, devices, etc. not shown may be included.
[0029] <Continuous plant involving precipitation due to evaporation> Figure 2 shows an example of a continuous plant involving precipitation due to evaporation. In the continuous plant shown in Figure 2, the raw material liquid is continuously fed into the evaporator 21 by a valve 22. The raw material liquid 23 fed into the evaporator 21 is stirred by a stirrer 25 and heated by a jacket 24, and its vapor is sent to a condenser 26. In the condenser 26, the vapor is condensed and liquefied and discharged as a distillate. On the other hand, the raw material liquid 23, which has been concentrated by a certain amount of heating and evaporation to reach the desired raw material concentration, is discharged as a concentrated liquid (for example, a supersaturated raw material liquid, a raw material liquid that has become a slurry due to precipitation, etc.). In such a continuous plant, precipitation of the raw material in the raw material liquid 23 may occur due to the heating and evaporation of the raw material liquid 23. The concentrated liquid may also be called a crystallizing liquid, etc.
[0030] In the continuous plant shown in Figure 2, a temperature sensor 401 for measuring the temperature of the raw material liquid 23 and a pressure sensor 402 for measuring the pressure of the steam generated by heating the raw material liquid 23 are installed.
[0031] The continuous plant shown in Figure 2 is a schematic representation of a plant 20 that performs a continuous process involving evaporation and precipitation, and is not limited to this example. For example, there may be multiple pieces of equipment such as evaporators 21, valves 22, jackets 24, agitators 25, and condensers 26, and there may also be various other pieces of equipment such as calandrias and external circulating heat exchangers. In addition, although the example in Figure 2 has one temperature sensor 401 and one pressure sensor 402, there may be multiple temperature sensors 401 or multiple pressure sensors 402. Furthermore, various sensors other than the temperature sensor 401 and pressure sensor 402 may be installed.
[0032] <Example of hardware configuration for concentration estimation device 10> Figure 3 shows an example of the hardware configuration of the concentration estimation device 10 according to this embodiment. As shown in Figure 3, the concentration estimation device 10 according to this embodiment includes an input device 101, a display device 102, an external I / F 103, a communication I / F 104, a RAM (Random Access Memory) 105, a ROM (Read Only Memory) 106, an auxiliary storage device 107, and a processor 108. Each of these hardware components is connected to each other via a bus 109 so as to be able to communicate.
[0033] The input device 101 is, for example, a keyboard, mouse, touch panel, or physical button. The display device 102 is, for example, a display or display panel. The concentration estimation device 10 does not necessarily have to have at least one of the input device 101 and the display device 102.
[0034] External I / F 103 is an interface with external devices such as recording media 103a. Examples of recording media 103a include flexible disks, CDs (Compact Discs), DVDs (Digital Versatile Disks), SD memory cards (Secure Digital memory cards), and USB (Universal Serial Bus) memory cards.
[0035] The communication interface 104 is an interface for the concentration estimation device 10 to connect to the communication network 80. The RAM 105 is a volatile semiconductor memory (storage device) that temporarily holds programs and data. The ROM 106 is a non-volatile semiconductor memory (storage device) that can retain programs and data even when the power is turned off. The auxiliary storage device 107 is a storage device (storage device) such as an HDD (Hard Disk Drive), SSD (Solid State Drive), or flash memory. The processor 108 is a processing unit such as a CPU (Central Processing Unit) or GPU (Graphics Processing Unit).
[0036] The concentration estimation device 10 according to this embodiment can realize various processes described later by having the hardware configuration shown in Figure 3. Note that the hardware configuration shown in Figure 3 is just one example, and the hardware configuration of the concentration estimation device 10 is not limited to this. For example, the concentration estimation device 10 may have multiple auxiliary storage devices 107 and multiple processors 108, it may not have some of the hardware shown, or it may have various hardware other than the hardware shown.
[0037] <Example of Functional Configuration of Concentration Estimation Device 10> Figure 4 shows an example of the functional configuration of the concentration estimation device 10 according to this embodiment. As shown in Figure 4, the concentration estimation device 10 according to this embodiment has an offline processing unit 201 and an online processing unit 202. Each of these parts is realized, for example, by processing that one or more programs installed in the concentration estimation device 10 cause the processor 108 to execute.
[0038] The offline processing unit 201 performs processing to create a concentration estimation model f while offline. Here, the offline processing unit 201 includes a data acquisition unit 211 and a model creation unit 212. The data acquisition unit 211 acquires the actual data set X={(x1(t),x2(t),···,x n The formula (t), y(t)|t∈T'⊆T} is obtained from database 60. The model creation unit 212 creates a concentration estimation model f using the actual data set X obtained by the data acquisition unit 211.
[0039] The online processing unit 202 calculates the concentration estimate y^(t) and performs processing to visualize it on the monitoring device 70 while online. Here, the online processing unit 202 includes a data acquisition unit 221, a concentration estimation unit 222, and a visualization unit 223. The data acquisition unit 221 acquires the measured values (x1(t), x2(t), ..., x) at each time t while online (i.e., the current time t). nThe concentration estimation unit 222 obtains the measured values (x1(t), x2(t), ..., x) obtained by the data acquisition unit 221. n The concentration estimate y^(t) at the current time t is calculated using (t) and the concentration estimation model f. The visualization unit 223 visualizes (displays) the concentration estimate y^(t) (or its moving average) on the display of the monitoring device 70 each time the concentration estimate y^(t) is calculated by the concentration estimation unit 222 (or each time the moving average described later is calculated).
[0040] In the example shown in Figure 4, the same concentration estimation device 10 has both the offline processing unit 201 and the online processing unit 202. However, this is just one example, and the offline processing unit 201 and the online processing unit 202 may be in separate devices. For example, the offline processing unit 201 may be in the first device, and the online processing unit 202 may be in the second device. In this case, for example, the first device may be called the "model creation device," and the second device may be called the "concentration estimation device," etc.
[0041] <Offline processing> The offline processing according to this embodiment will be described below with reference to Figure 5.
[0042] The data acquisition unit 211 of the offline processing unit 201 acquires the actual data set X for model creation from the database 60 (step S101). That is, the data acquisition unit 211 acquires the actual data set X from the database 60 as the training data set.
[0043] The model creation unit 212 of the offline processing unit 201 creates a concentration estimation model f using the set of actual data X acquired in step S101 (step S102). The concentration estimation model f created by the model creation unit 212 is stored in a storage area such as the auxiliary storage device 107.
[0044] Here, the model creation unit 212 generates, for example, x1, x2, ..., x nAfter selecting the variables to be used as explanatory variables from among them, a concentration estimation model f can be created that accurately estimates the concentration measurement value y(t) when concentration is the dependent variable. For example, the model creation method for creating the concentration estimation model f can be partial least squares regression (PLS), support vector regression (SVR), or random forest.
[0045] More specifically, the model creation unit 212 can create a concentration estimation model f by following steps 1 and 2 below.
[0046] Step 1: First, the model creation unit 212 sets x1, x2, ..., x n Select the variables to be used as explanatory variables from the following. Below, the variables selected in step 1 will be x i_1 ,x i_2 ,···,x i_m Let's assume the following: i_1, i_2, ..., i_m ∈ {1, ..., n} and 1 ≤ m ≤ n. For the selection of explanatory variables, any known method can be used; for example, LASSO (Least Absolute Shrinkage and Selection Operator), and if PLS is adopted as the model creation method, VIP (Variable Importance in Projection), etc., can be used.
[0047] Step 2: Then, the model creation unit 212, x i_1 ,x i_2 ,···,x i_m Using x as the explanatory variable and y as the dependent variable, we can construct a model using known modeling methods (partial least squares regression, support vector regression, random forest) to obtain f(x i_1 (t),x i_2 (t),···,x i_mThe parameters of the concentration estimation model f are determined (in other words, the parameters to be learned for the concentration estimation model f) so that (t) accurately estimates y(t). This creates the concentration estimation model f (i.e., the concentration estimation model f expressed as some function with the learned parameters).
[0048] In addition, in step 2 above, for example, multiple concentration estimation models may be created using each of the multiple model creation methods, and the one with the best accuracy among these multiple concentration estimation models may be used as the final concentration estimation model f.
[0049] Furthermore, using partial least squares regression, support vector regression, and random forest as model creation methods is just one example; it is also possible to use any other model creation method capable of creating machine learning models or statistical models.
[0050] <Online Processing> The online processing according to this embodiment will be described below with reference to Figure 6. Here, steps S201 to S203 in Figure 6 are repeatedly executed at each time t while online. Note that the time t while online is updated, for example, t←t+ΔT1 each time the sensing period ΔT1 has elapsed. Steps S201 to S203 at a certain time t will be described below.
[0051] The data acquisition unit 221 of the online processing unit 202 acquires the measured values (x1(t), x2(t), ..., x) at the time t (current time t). n (t)) is obtained from database 60 (step S201). Note that the data acquisition unit 221 is (x1(t),x2(t),···,x n (t)) may be received from the control device 30 instead of being obtained from the database 60. Also, the data acquisition unit 221 may receive (x1(t),x2(t),···,x n Instead of obtaining (t), it is a part of it (x i_1 (t),x i_2 (t),···,x i_mYou may obtain (t).
[0052] The concentration estimation unit 222 of the online processing unit 202 uses the measured values (x1(t), x2(t), ..., x) obtained in step S201 above. n Using (t) and the concentration estimation model f, the concentration estimate y^(t) at time t is calculated (step S202). That is, the concentration estimation unit 222 calculates y^(t) = f(x i_1 (t),x i_2 (t),···,x i_m The concentration estimate y^(t) is calculated using (t).
[0053] The visualization unit 223 of the online processing unit 202 visualizes (displays) the concentration estimate y^(t) calculated in step S202 on the display of the monitoring device 70 (step S203). At this time, the visualization unit 223 may visualize not the concentration estimate y^(t) itself, but rather a moving average that takes into account the average residence time of the raw material liquid in the plant 20, for example. For example, if the average time from when the raw material liquid is introduced into the plant 20 until the concentrated liquid corresponding to that raw material liquid is discharged (this is the average residence time) is μ hours, the visualization unit 223 may calculate a moving average that takes into account the concentration estimate y^(t) for μ hours and visualize those moving averages. Note that the visualization unit 223 may calculate the moving average, but the monitoring device 70 that receives the concentration estimate y^(t) may also calculate and visualize the moving average.
[0054] Steps S201 to S203 described above are repeated at each time t while the system is online, so that a time-series graph of the estimated concentration y^(t) at each time t while the system is online is displayed in real time on the display of the monitoring device 70. This allows the operators of the plant 20 to check the estimated concentration y^(t) at each time t while the system is online, and to know whether the raw material concentration of the raw material liquid being processed in the plant 20 (i.e., for example, the raw material concentration of raw material liquid 23 in the example of Figure 2) is appropriate. Therefore, if the raw material concentration is not appropriate, the operators can take some action to adjust the raw material concentration in the plant 20 (for example, an operation to lower the temperature of the raw material liquid if the raw material concentration is too high, or an operation to raise the temperature of the raw material liquid if the raw material concentration is too low).
[0055] <Concentration estimation results> Figure 7 shows the results of calculating the concentration estimate y^(t) using the concentration estimation device 10 according to this embodiment, with partial least-squares regression as the model creation method for a continuous process involving precipitation due to heating and evaporation of the raw material liquid. Figure 7 shows the measured concentration value y(t), the concentration estimate y^(t), and the moving average value of the concentration estimate y^(t). The moving average value is a moving average value that takes into account the average residence time of the raw material liquid in plant 20.
[0056] As shown in Figure 7, the concentration estimate y^(t) can be estimated to a certain extent with accuracy compared to the measured concentration y(t). In particular, the moving average of the concentration estimate y^(t) can be estimated to be accurate compared to the measured concentration y(t).
[0057] Therefore, it can be seen that accurate concentration estimation results are obtained by applying a moving average that takes into account the average residence time of the raw material liquid in plant 20 to the concentration estimate y^(t). Furthermore, an accurate concentration estimation model f is obtained using a linear modeling method called partial least squares regression, and it can be said that estimation results with good monitoring capabilities are obtained.
[0058] <Variation> The following describes some modified examples of this embodiment.
[0059] • Variation 1 The concentration estimation device 10 according to this embodiment may not only visualize the concentration estimate y^(t) or its moving average value on the monitoring device 70, but may also control the plant 20 according to the concentration estimate y^(t) or its moving average value. That is, the concentration estimation device 10 according to this embodiment may, for example, use the concentration estimate y^(t) or its moving average value and a preset upper concentration threshold th1 and lower concentration threshold th2 to perform an operation to lower the temperature of the raw material liquid if the concentration estimate y^(t) or its moving average value exceeds the upper concentration threshold th1, and perform an operation to raise the temperature of the raw material liquid if the concentration estimate y^(t) or its moving average value falls below the lower concentration threshold th2. In this way, the concentration estimation device 10 according to this embodiment can control the raw material concentration of the raw material liquid in the plant 20 according to the concentration estimate y^(t) or its moving average value.
[0060] Furthermore, operations to raise or lower the temperature of the raw material liquid include, for example, opening or closing a valve that adjusts the amount of heat transfer medium (e.g., high-temperature steam) supplied to the jacket 24, as shown in Figure 2.
[0061] • Variation 2 Even after creating a concentration estimation model f, after a certain period of time, the estimated concentration value y^(t) may deviate from the actual concentration due to some reason (for example, deterioration of equipment in the plant 20 over time). Therefore, if such a situation occurs, the concentration estimation device 10 according to this embodiment may recreate the concentration estimation model f.
[0062] For example, if the absolute value of the difference between the measured concentration value y(t) and the estimated concentration value y^(t) or its moving average value exceeds a preset tolerance, the concentration estimation device 10 according to this embodiment may perform the offline processing shown in Figure 5 again to recreate the concentration estimation model f. This makes it possible to prevent the accuracy of the estimated concentration value y^(t) from falling below a certain level.
[0063] <Summary> As described above, the concentration estimation device 10 according to this embodiment can accurately estimate the raw material concentration from various physical quantities (e.g., temperature, pressure, flow rate, etc.) measured in a continuous process that involves precipitation due to heating and evaporation of the raw material liquid. Furthermore, the concentration estimation device 10 according to this embodiment visualizes the estimated concentration value y^(t) or its moving average value on the monitoring device 70. This makes it possible for plant operators, etc., to know in real time whether the raw material concentration of the raw material liquid being processed in the continuous process is appropriate. For example, if the raw material concentration of the raw material liquid is too high or too low, it becomes possible to perform operations to adjust the raw material concentration to an appropriate value.
[0064] By using the concentration estimation device 10 according to this embodiment, in addition to the above, it is also possible to address the following situations (1) and (2).
[0065] (1) In order to obtain the measured concentration value y(t) using the concentration measuring instrument 50, it is necessary to sample the raw material liquid being processed in a continuous process. Generally, this sampling is done by sampling the raw material liquid from a nozzle called a sampling nozzle. On the other hand, in processes involving precipitation, this sampling nozzle may become blocked by precipitation, making it impossible to sample the raw material liquid. In contrast, by using the concentration estimation device 10 according to this embodiment, it is possible to obtain an accurate concentration estimate value y^(t) or its moving average value even when the sampling nozzle is blocked.
[0066] (2) In a continuous process, rapid process fluctuations (i.e., fluctuations and variations in various physical quantities) generally occur at the start of the process (startup). In contrast, the concentration estimation device 10 according to this embodiment can estimate the concentration in a way that can handle such rapid process fluctuations at startup by creating a concentration estimation model f from the measured concentration value at startup and the measured value at that time. For this reason, by using the concentration estimation device 10 according to this embodiment, it is possible to achieve safe and stable operation, especially at the start-up of the plant 20, when accidents are likely to occur. Not limited to the start-up, the concentration estimation device 10 according to this embodiment can also be applied in the same way at the end of a continuous process (shutdown).
[0067] The present invention is not limited to the embodiments specifically disclosed above, and various modifications, changes, and combinations with known technologies are possible without departing from the scope of the claims. [Explanation of symbols]
[0068] 1. Concentration Estimation System 10 Concentration estimation device 20 plants 21 Evaporator 22 valves 23 Raw material liquid 24 Jackets 25 Agitator 26 Condenser 30 Control device 40 sensor group 50 Concentration measuring instruments 60 databases 70 Monitoring equipment 80 Communication Networks 101 Input Device 102 Display device 103 External I / F 103a Recording medium 104 Communication I / F 105 RAM 106 ROM 107 Auxiliary storage 108 processors 109 Bus 201 Offline Processing Unit 202 Online Processing Unit 211 Data Acquisition Unit 212 Model Creation Department 221 Data Acquisition Unit 222 Concentration estimation section 223 Visualization section
Claims
1. A procedure for obtaining measured values of predetermined physical quantities in a continuous process involving precipitation due to evaporation of the raw material liquid, An estimation procedure for calculating an estimated concentration from the measured value using a model for estimating the concentration of the raw material in the raw material liquid, A visualization procedure for visualizing the moving average of the estimated concentration based on the residence time of the continuous process, A concentration estimation method performed by a computer.
2. The aforementioned model is a statistical model or machine learning model that has parameters to be learned. The computer further executes a model creation procedure to learn the parameters of the model using the measured values from a predetermined period in the past and the actual measured values of the concentration during that period as training data. The estimation procedure described above is: The concentration estimation method according to claim 1, wherein an estimated value of the concentration is calculated from the measured value using the model having the learned parameters.
3. A concentration control method in which a computer performs a control procedure for performing an operation to control at least one of the physical quantities based on an estimated concentration value estimated by the concentration estimation method according to claim 1 or 2 and a preset threshold.
4. The concentration control method according to claim 3, wherein the continuous process is a crystallization process that includes a process of creating a concentrated solution of a target raw material concentration by evaporation of the raw material solution.
5. A continuous crystallization process in which equipment controlled by the concentration control method described in claim 4 produces a concentrated solution of a predetermined concentration.
6. An acquisition unit is configured to acquire measured values of predetermined physical quantities in a continuous process involving precipitation due to evaporation of the raw material liquid, An estimation unit is configured to calculate an estimated value of the concentration from the measured value using a model for estimating the concentration of the raw material in the raw material liquid. A visualization unit configured to visualize the moving average of the estimated concentration based on the residence time of the continuous process, A concentration estimation device having the following features.