Measurement device, measurement method, and measurement program

Abstract

A gas chromatograph (10) separates constituents of a sample by gas chromatography using components each provided with a heating device and controlled to a predetermined temperature by the heating device, and outputs measurement data indicating a concentration of the target constituent having been separated.

Description

Measuring device, measuring method, and measuring program

[0001] This disclosure relates to a measuring device, a measuring method, and a measuring program.

[0002] A gas chromatograph is an analytical instrument that takes a fixed amount of a multi-component mixed sample, sends it to a column (separation tube) using a carrier gas that has been reduced to a constant pressure by a pressure reducing valve, and then sequentially measures the concentration of each component separated in the column using a detector.

[0003] International Publication No. 2017 / 149794

[0004] However, the above techniques make it difficult to improve the measurement accuracy of gas chromatography. For example, in situations where the separation of components is insufficient, the detected peaks may overlap, making quantitative or qualitative analysis impossible.

[0005] This disclosure has been made in view of the above and aims to improve the measurement accuracy of gas chromatography.

[0006] A measuring device according to one embodiment of the present disclosure comprises a component and a processor connected to each component, each component having a heating device installed, and the processor performs the following: using each component controlled to a predetermined temperature by the heating device, separates each component contained in a sample by gas chromatography and outputs measurement data indicating the concentration of the separated target component.

[0007] A measuring device according to one embodiment of the present disclosure comprises a component and a processor connected to each component, each component having a drive device installed thereon, and the processor performs the following: using each component controlled by the drive device to a predetermined pressure, separates each component contained in a sample by gas chromatography and outputs measurement data indicating the concentration of the separated target component.

[0008] A measurement method according to one embodiment of the present disclosure involves a computer separating each component contained in a sample by gas chromatography using components each equipped with a heating device, the heating devices each controlled to a predetermined temperature, and outputting measurement data indicating the concentration of the separated target component.

[0009] A measurement method according to one embodiment of the present disclosure involves a computer using components each equipped with a drive device, which controls each drive device to a predetermined pressure, to separate each component contained in a sample by gas chromatography, and outputting measurement data indicating the concentration of the separated target component.

[0010] A measurement program according to one embodiment of the present disclosure causes a computer to perform the following actions: separate each component contained in a sample by gas chromatography using each component equipped with a heating device, each of which is controlled to a predetermined temperature; and output measurement data indicating the concentration of the separated target component.

[0011] A measurement program according to one embodiment of the present disclosure causes a computer to perform the following actions: separate each component contained in a sample by gas chromatography using components each equipped with a drive device, each of which is controlled to a predetermined pressure by the drive device, and output measurement data indicating the concentration of the separated target component.

[0012] According to this disclosure, there is an effect that can improve the measurement accuracy of gas chromatography.

[0013] This figure shows an example of the configuration and processing of a gas measurement system according to the embodiment. This figure shows an example of the configuration of the oven of a gas chromatograph according to the embodiment. This block diagram shows an example of the configuration of the gas chromatograph of a gas measurement system according to the embodiment. This block diagram shows an example of the configuration of the oven of a gas chromatograph of a gas measurement system according to the embodiment. This figure shows an example of the measurement data storage unit of a gas chromatograph according to the embodiment. This figure shows an example of the machine learning model storage unit of a gas chromatograph according to the embodiment. This figure shows a specific example of the peak separation calculation process of a gas chromatograph according to the embodiment. This flowchart shows an example of the overall processing flow of a gas measurement system according to the embodiment. This flowchart shows an example of the model design processing flow of a gas measurement system according to the embodiment. This flowchart shows an example of the optimal condition identification processing flow of a gas measurement system according to the embodiment. This flowchart shows an example of the measurement control processing flow of a gas measurement system according to the embodiment. This figure shows an example of the hardware configuration according to the embodiment.

[0014] A measuring apparatus, a measuring method, and a measuring program according to one embodiment of this disclosure will be described in detail below with reference to the drawings. However, this disclosure is not limited to the embodiments described below.

[0015] The configuration and processing of the gas measurement system 100 according to the embodiment, the configuration and processing of each device of the gas measurement system 100, specific examples of each process of the gas measurement system 100, the flow of each process of the gas measurement system 100, and the effects of the embodiment will be described below.

[0016] [1. Configuration and Processing of Gas Measurement System 100] The configuration and processing of the gas measurement system 100 according to the embodiment will be described using Figures 1 and 2. Figure 1 is a diagram showing an example of the configuration and processing of the gas measurement system 100 according to the embodiment. Below, an example of the overall configuration of the gas measurement system 100, an example of the overall processing of the gas measurement system 100, and the effects of the gas measurement system 100 will be described.

[0017] (1-1. Example of the overall configuration of the gas measurement system 100) An example of the overall configuration of the gas measurement system 100 will be described. The gas measurement system 100 consists of a gas chromatograph 10. Here, the gas chromatograph 10 is connected to various devices (not shown) via a predetermined communication network (not shown) by wired or wireless means. Various communication networks such as the internet or dedicated lines can be used as the predetermined communication network.

[0018] (1-1-1. Gas Chromatograph 10) The gas chromatograph 10 is a measuring device controlled by operator W that uses the principle of gas chromatography to measure the concentration of each component in the sample SP. For example, the gas chromatograph 10 performs process gas chromatography in a plant to measure gas components as process data. The gas chromatograph 10 may be an air-bus type or an airless type. Note that the gas measurement system 100 shown in Figure 1 may include multiple gas chromatographs 10.

[0019] (1-2. Example of the entire gas measurement system 100) An example of the entire gas measurement system 100 process will be described. Note that some of the processes in steps S1 to S3 below may be omitted. For example, the process in step S3 below may be omitted.

[0020] (1-2-1. Model Training Process) The gas chromatograph 10 performs a model training process (step S1). For example, the gas chromatograph 10 generates training data for each target component TC, including the temperature T and peak separation R of each component contained in the oven 16 (not shown), and trains the machine learning model LM. Alternatively, the gas chromatograph 10 may generate training data for each target component TC, including the pressure P and peak separation R of each component contained in the oven 16, and train the machine learning model LM. Alternatively, the gas chromatograph 10 may generate training data for each target component TC, including the flow rate F and peak separation R of each component contained in the oven 16, and train the machine learning model LM.

[0021] Here, with reference to FIG. 2, each component included in the oven 16 of the gas chromatograph 10 will be described. FIG. 2 is a diagram showing a configuration example of the oven 16 of the gas chromatograph 10 according to the embodiment. As shown in FIG. 2, the oven 16, which is a thermostatic chamber, is composed of an oven heater 16a, a pressure reducing valve 16b, a sample valve 16c, a column 16d, a detector 16e, a flow rate adjusting valve 16f, a pressure reducing valve heater 16g, a column heater 16h, and a flow rate adjusting valve heater 16i. Note that the oven heater 16a, the pressure reducing valve 16b, and the flow rate adjusting valve 16f may be installed outside the oven 16. Also, a plurality of pressure reducing valves 16b, columns 16d, and flow rate adjusting valves 16f may be installed.

[0022] A pressure reducing valve heater 16g, which is a heating device for controlling the temperature T1 of the pressure reducing valve 16b, is installed in the pressure reducing valve 16b. Also, a column heater 16h, which is a heating device for controlling the temperature T2 of the column 16d, is installed in the column 16d. Further, a flow rate adjusting valve heater 16i, which is a heating device for controlling the temperature T3 of the flow rate adjusting valve 16f, is installed in the flow rate adjusting valve 16f. Note that the installation location of the heating device may be in contact with each component as long as the temperature of each component can be controlled, or may be in a separated location.

[0023] A pressure reducing valve temperature sensor TS1 (not shown), which is a sensor device for measuring the temperature T1 of the pressure reducing valve 16b, is installed in the pressure reducing valve 16b. Also, a column temperature sensor TS2 (not shown), which is a sensor device for measuring the temperature T2 of the column 16d, is installed in the column 16d. Further, a flow rate adjusting valve temperature sensor TS3 (not shown), which is a sensor device for measuring the temperature T3 of the flow rate adjusting valve 16f, is installed in the flow rate adjusting valve 16f.

[0024] An electric drive type pressure reducing valve drive device 16j (not shown) for controlling the pressure P may be installed in the pressure reducing valve 16b. Also, an electric drive type flow rate adjusting valve drive device 16k (not shown) for controlling the flow rate F may be installed in the flow rate adjusting valve 16f. Further, a pressure reducing valve pressure sensor PS (not shown), which is a sensor device for measuring the pressure P, may be installed in the pressure reducing valve 16b. Also, a flow rate sensor FS (not shown), which is a sensor device for measuring the flow rate F, may be installed in the flow rate adjusting valve 16f.

[0025] Next, the outline of the gas chromatography of the gas chromatograph 10 will be described with reference to FIG. 2. First, the oven heater 16a heats the entire oven 16 and maintains it at a constant temperature. Second, the pressure reducing valve 16b reduces the pressure of the gas G, which is a carrier gas (e.g., hydrogen gas, nitrogen gas, argon gas, helium gas), and sends it to the column 16d. Third, the sample valve 16c collects the sample SP to be measured. Fourth, the flow rate adjusting valve 16f adjusts the flow rate of the gas G sent to the column 16d or controls the discharge rate of the gas G discharged without passing through the column 16d. Fifth, the column 16d changes the moving speed of each component of the sample SP by the interaction with the stationary phase and separates them in time. Sixth, the detector 16e detects each component sent.

[0026] (1-2-2. Condition setting process) The gas chromatograph 10 executes a condition setting process (step S2). For example, the gas chromatograph 10 sets the pressure reducing valve heater 16g so that the pressure reducing valve 16b is at the temperature T1, sets the column heater 16h so that the column 16d is at the temperature T2, and sets the flow rate adjusting valve heater 16i so that the flow rate adjusting valve 16f is at the temperature T3 as the temperature conditions output by the trained machine learning model LM. Also, the gas chromatograph 10 may set the pressure reducing valve drive device 16j so that the pressure reducing valve 16b is at the pressure P1 as the pressure conditions output by the trained machine learning model LM. Further, the gas chromatograph 10 may set the flow rate adjusting valve drive device 16k so that the flow rate adjusting valve 16f is at the flow rate F3 as the flow rate conditions output by the trained machine learning model LM.

[0027] (1-2-3. Measurement Result Output Processing) The gas chromatograph 10 performs measurement result output processing (step S3). For example, the gas chromatograph 10 performs gas chromatography measurement under set temperature conditions and outputs measurement data MD showing the concentration of the target component TC. Alternatively, the gas chromatograph 10 may perform gas chromatography measurement under set pressure conditions and output measurement data MD showing the concentration of the target component TC. Alternatively, the gas chromatograph 10 may perform gas chromatography measurement under set flow rate conditions and output measurement data MD showing the concentration of the target component TC.

[0028] (1-3. Effects of the Gas Measurement System 100) Below, we will explain the overview, principle, and problems of the gas measurement system 100-P related to the reference technology, and then explain the effects of the gas measurement system 100 according to the embodiment.

[0029] (1-3-1. Overview of Gas Measurement System 100-P) In the gas measurement system 100-P relating to the reference technology, a certain amount of multi-component mixed sample is sent to a column 16d with a carrier gas that has been reduced to a constant pressure by a pressure reducing valve 16b, and the concentration of each component separated in the column 16d is measured sequentially via a detector 16e.

[0030] The gas measurement system 100-P sometimes utilizes a process gas chromatograph for rapid analysis during industrial processes. In this case, the gas measurement system 100-P can be directly integrated into the production line, performing continuous real-time analysis of gas components and providing rapid feedback, thus contributing to the optimization of manufacturing processes and quality control. Furthermore, the gas measurement system 100-P is automated and used for repeated measurement of specific components. For example, the gas measurement system 100-P is widely used for component analysis in a wide range of applications, including the chemical industry, petroleum refining, and environmental monitoring.

[0031] (1-3-2. Principle of Gas Measurement System 100-P) Here, the principle of the gas measurement system 100-P related to the reference technology will be explained. Below, the principle of gas chromatography will be explained, including the separation of components, the retention time of components, and the relationship between the temperature of the oven 16 and the retention time.

[0032] (1-3-2-1. Separation of Components) In the gas measurement system 100-P, each component contained in the sample SP is separated by a column 16d positioned within the flow path through which each component passes. Each component is separated according to the principle of chromatography based on the partition equilibrium between the gas phase and the liquid phase (or solid phase) in the column 16d. The movement of each component in the column 16d is determined by the interaction of release and dissolution. In this case, components that are more easily released or less easily dissolved pass through the column 16d faster. On the other hand, components that are less easily released or more easily dissolved pass through the column 16d more slowly.

[0033] (1-3-2-2. Retention Time of Components) In gas chromatography, the time from when each component is introduced into the flow path until it reaches the detector 16e is called the retention time (or retention time). Here, the retention time is the sum of the time the component is in the gas phase and the time it is in the liquid phase in the column 16d, and the time it is in the liquid phase in the column 16d is affected by the solubility of the measured component. Therefore, in the gas measurement system 100-P, the retention time depends on conditions such as the type, length, and state of the column 16d and the measured component, as well as the flow rate and pressure of the carrier gas and the temperature of the column 16d. Furthermore, in the gas measurement system 100-P, by promoting the separation of each component, the difference in retention times widens, and the overlap of peaks detected by the detector 16e is reduced, making it possible to measure concentrations more accurately.

[0034] (1-3-2-3. Relationship between oven temperature and holding time) In the gas measurement system 100-P, a constant temperature bath and a heating bath may be used as the gas chromatograph 10-P. In the oven 16, which is a constant temperature bath, the temperature T of the gas inside is controlled so that each component such as the column 16d, pressure reducing valve 16b, and restrictor maintains a constant temperature. On the other hand, in the heating bath, the temperature is controlled to a programmed range of temperature rise in order to sequentially separate compounds of various measurement components with different boiling points and complete the analysis quickly. The higher the average temperature T, the shorter the holding time, but this has the effect of worsening the peak separation R and shortening the lifespan of each component.

[0035] (1-3-3. Problems with the Gas Measurement System 100-P) The gas measurement system 100-P, which is the reference technology, has the following problems. Firstly, in the gas measurement system 100-P, when component separation is insufficient, the detected peaks overlap. In this case, since each peak in the gas measurement system 100-P has its own unique temporal peak width, quantitative or qualitative analysis is difficult when different components reach the detector 16e at the same time. Also, in the gas measurement system 100-P, since the concentration of a component depends on the area of ​​the peak, it becomes difficult to accurately calculate the area of ​​the overlapping part of the peaks, and it also becomes difficult to identify the components of adjacent peaks. Secondly, in the gas measurement system 100-P, if the temperature of the entire oven 16 is lowered or the column length is increased in order to improve peak separation, the retention time of all components increases, which increases the analysis time and creates a disadvantageous situation for measurement.

[0036] (1-3-4. Overview of the Gas Measurement System 100) In the gas measurement system 100 according to this embodiment, the following processes are performed. First, the gas chromatograph 10 generates training data for each target component TC, including the conditions (temperature T, pressure P, flow rate F) and peak separation degree R of each component contained in the oven 16, and trains the machine learning model LM. Second, the gas chromatograph 10 processes the optimized optimal conditions (optimal temperature T) output by the trained machine learning model LM. ＯＰ , optimal pressure P ＯＰ, optimal flow rate F ＯＰ ) Each component is set. Thirdly, the gas chromatograph 10 performs gas chromatography measurement under the set optimal conditions and outputs measurement data MD showing the concentration of the target component TC.

[0037] (1-3-5. Effects of the Gas Measurement System 100) The gas measurement system 100 according to this embodiment has the following effects. Firstly, the gas measurement system 100 can improve the separation of adjacent peaks on the chromatogram by installing additional heaters and drive devices on specific components in the oven 16, controlling the temperature T and pressure P for each component to adjust the local retention time of the components, and then using an algorithm to optimize the peak separation degree R. Secondly, the gas measurement system 100 can shorten the measurement time while maintaining measurement accuracy by using the above-mentioned algorithm to optimize the peak separation degree R.

[0038] As described above, the gas measurement system 100 can improve the measurement accuracy of gas chromatography.

[0039] [2. Configuration and Processing of Each Device in the Gas Measurement System 100] The configuration and processing of each device in the gas measurement system 100 shown in Figure 1 will be explained using Figures 3 to 6. Below, an example of the overall configuration of the gas measurement system 100 according to the embodiment, as well as an example of the configuration and processing of the gas chromatograph 10, will be described.

[0040] (2-1. Example of the overall configuration of the gas measurement system 100) An example of the overall configuration of the gas measurement system 100 shown in Figure 1 will be explained using Figure 3. Figure 3 is a block diagram showing an example of the configuration of the gas chromatograph 10 of the gas measurement system 100 according to the embodiment. As shown in Figure 3, the gas measurement system 100 is composed of a gas chromatograph 10. The gas chromatograph 10 is also connected to various devices (not shown) via a communication network realized by the internet or a dedicated line.

[0041] (2-2. Configuration Examples and Processing Examples of Gas Chromatograph 10) Configuration examples and processing examples of the gas chromatograph 10 will be described using Figures 3 and 4. Figure 4 is a block diagram showing a configuration example of the oven 16M of the gas chromatograph 10 of the gas measurement system 100 according to the embodiment. Note that oven 16M is a modified version of oven 16, and its details will be described later. As shown in Figure 3, the gas chromatograph 10 has an input unit 11, an output unit 12, a communication unit 13, a storage unit 14, a control unit 15, and an oven 16. Also, as shown in Figure 4, the gas chromatograph 10 may have an oven 16M. Note that in Figure 4, the configurations of the input unit 11, output unit 12, communication unit 13, storage unit 14, and control unit 15 of the gas chromatograph 10 are the same as in Figure 3, so they are omitted.

[0042] (2-2-1. Input Unit 11) The input unit 11 is responsible for inputting various types of information to the gas chromatograph 10. For example, the input unit 11 can be implemented using a mouse, keyboard, touch panel, etc., and accepts input of various types of information to the gas chromatograph 10.

[0043] (2-2-2. Output Unit 12) The output unit 12 is responsible for outputting various types of information from the gas chromatograph 10. For example, the output unit 12 is implemented using a display, speaker, etc., and displays various types of information stored in the gas chromatograph 10.

[0044] (2-2-3. Communication Unit 13) The communication unit 13 is responsible for data communication with other devices. For example, the communication unit 13 performs data communication with each communication device via a router or the like. The communication unit 13 can also perform data communication with terminals (not shown).

[0045] (2-2-4. Storage Unit 14) The storage unit 14 stores various information that the control unit 15 refers to when it operates, and various information acquired when the control unit 15 operates. The storage unit 14 is composed of a measurement data storage unit 14a and a machine learning model storage unit 14b. Here, the storage unit 14 can be implemented as, for example, a semiconductor memory element such as RAM (Random Access Memory) or flash memory, or a storage device such as a hard disk or optical disc. In the example in Figure 3, the storage unit 14 is installed inside the gas chromatograph 10, but it may be installed outside the gas chromatograph 10, or multiple storage units may be installed.

[0046] (2-2-4-1. Measurement Data Storage Unit 14a) The measurement data storage unit 14a stores the measurement data MD. For example, the measurement data storage unit 14a stores the measurement data MD output by the measurement unit 15c of the control unit 15, which will be described later. The measurement data storage unit 14a also stores the peak separation degree R calculated by the calculation unit 15d of the control unit 15, which will be described later.

[0047] Here, an example of the data stored in the measurement data storage unit 14a will be explained using Figure 5. Figure 5 is a diagram showing an example of the measurement data storage unit 14a of the gas chromatograph 10 according to the embodiment. In the example in Figure 5, the measurement data storage unit 14a has items such as "gas chromatograph", "measurement data", and "peak separation".

[0048] "Gas chromatograph" indicates identification information for identifying the gas chromatograph 10, which is the measuring device. For example, "Gas chromatograph" is the identification number or identification symbol of the gas chromatograph 10. "Measurement data" indicates the detection time and detection concentration of each component contained in the sample SP. For example, "Measurement data" expresses the detection time in hours, minutes, and seconds, and the detection concentration in %, or ppm. Alternatively, "Measurement data" expresses the detection concentration in terms of detection intensity, such as the detector signal. Alternatively, "Measurement data" may be a chromatogram showing the relationship between the detection time and the detection intensity, such as the detector signal. "Peak separation" indicates the degree to which the target component TC is separated from adjacent peaks. For example, "Peak separation" is the peak separation R between the measurement data MD and the adjacent peaks before and after it.

[0049] In other words, Figure 5 shows an example in which the measurement data storage unit 14a stores the following data for the gas chromatograph 10 identified as "GC001": {measurement data: "MD101", peak resolution: "PR101"}, {measurement data: "MD102", peak resolution: "PR102"}, {measurement data: "MD103", peak resolution: "PR103"}, ...

[0050] (2-2-4-2. Machine Learning Model Storage Unit 14b) The machine learning model storage unit 14b stores machine learning models LM. For example, the machine learning model storage unit 14b stores machine learning models LM used by the setting unit 15b of the control unit 15, which will be described later. The machine learning model storage unit 14b also stores machine learning models LM that are trained by the training unit 15a of the control unit 15, which will be described later.

[0051] Here, an example of the information stored in the machine learning model storage unit 14b will be explained using Figure 6. Figure 6 is a diagram showing an example of the machine learning model storage unit 14b of the gas chromatograph 10 according to the embodiment. In the example in Figure 6, the machine learning model storage unit 14b has an item such as "machine learning model".

[0052] A "machine learning model" is model data of a trained machine learning model LM, and includes, for example, execution data for running the algorithm of the machine learning model LM, model parameters which are setting values, hyperparameters, etc.

[0053] Figure 6 shows an example in which trained machine learning models LM, namely "LM001," "LM002," "LM003," etc., are stored in the machine learning model storage unit 14b. Furthermore, "LM001," "LM002," "LM003," etc., can be used as different machine learning models LM depending on the type and amount of input data. In addition, "LM001," "LM002," "LM003," etc., can be used when redesigning the machine learning model LM.

[0054] (2-2-5. Oven 16) Here, we will describe the oven 16 in the gas measurement system 100 that enables temperature control of each component. As shown in Figure 3, the oven 16 is composed of an oven heater 16a, a pressure reducing valve 16b, a sample valve 16c, a column 16d, a detector 16e, a flow control valve 16f, a pressure reducing valve heater 16g, a column heater 16h, a flow control valve heater 16i, a pressure reducing valve temperature sensor TS1, a column temperature sensor TS2, and a flow control valve temperature sensor TS3. The oven 16 may further include some or all of the pressure reducing valve drive device 16j, a flow control valve drive device 16k, a pressure reducing valve pressure sensor PS, and a flow sensor FS that constitute the oven 16M described later.

[0055] (2-2-5-1. Oven heater 16a) The oven heater 16a is a heating device installed in the oven 16 that heats the entire oven 16 and maintains it at a constant temperature.

[0056] (2-2-5-2. Pressure Reducing Valve 16b) The pressure reducing valve 16b reduces the pressure of gas G, which is a carrier gas (e.g., hydrogen gas, nitrogen gas, argon gas, helium gas), and sends it to column 16d.

[0057] (2-2-5-3. Sample valve 16c) The sample valve 16c collects the sample SP to be measured. The sample valve 16c also releases the collected sample SP.

[0058] (2-2-5-4. Column 16d) Column 16d changes the migration rate of each component of the sample SP through interaction with the stationary phase, thereby separating them over time.

[0059] (2-2-5-5. Detector 16e) Detector 16e detects each emitted component. Detector 16e can be implemented using, but is not limited to, a flame ionization detector (FID), thermal conductivity detector (TCD), barrier discharge ionization detector (BID), electron capture detector (ECD), thermal ionization detector (FTD), flame photometric detector (FPD), or chemiluminescence sulfur detector (SCD).

[0060] (2-2-5-6. Flow control valve 16f) The flow control valve 16f adjusts the flow rate of gas G delivered to column 16d.

[0061] (2-2-5-7. Pressure Reducing Valve Heater 16g) The pressure reducing valve heater 16g is a heating device installed on the pressure reducing valve 16b, which heats the pressure reducing valve 16b according to the temperature conditions set by the setting unit 15b of the control unit 15 (described later), and maintains a constant temperature. The pressure reducing valve heater 16g also heats the pressure reducing valve 16b according to the temperature conditions set by the setting unit 15b and maintains a high temperature for a predetermined time. The pressure reducing valve heater 16g also stops heating the pressure reducing valve 16b according to the temperature conditions set by the setting unit 15b and maintains a low temperature for a predetermined time.

[0062] (2-2-5-8. Column heater 16h) The column heater 16h is a heating device installed on the column 16d, which heats the column 16d according to the temperature conditions set by the setting unit 15b of the control unit 15 (described later), and maintains it at a constant temperature. The column heater 16h also heats the column 16d according to the temperature conditions set by the setting unit 15b and maintains it at a high temperature for a predetermined time. The column heater 16h also stops heating the column 16d according to the temperature conditions set by the setting unit 15b and maintains it at a low temperature for a predetermined time.

[0063] (2-2-5-9. Flow Control Valve Heater 16i) The flow control valve heater 16i is a heating device installed on the flow control valve 16f, which heats the flow control valve 16f according to the temperature conditions set by the setting unit 15b of the control unit 15 (described later), and maintains it at a constant temperature. The flow control valve heater 16i also heats the flow control valve 16f according to the temperature conditions set by the setting unit 15b and maintains it at a high temperature for a predetermined time. The flow control valve heater 16i also stops heating the flow control valve 16f according to the temperature conditions set by the setting unit 15b and maintains it at a low temperature for a predetermined time.

[0064] (2-2-5-10. Pressure Reducing Valve Temperature Sensor TS1) The pressure reducing valve temperature sensor TS1 is a sensor device installed on the pressure reducing valve 16b, and measures the temperature T1 of the pressure reducing valve 16b.

[0065] (2-2-5-11. Column temperature sensor TS2) The column temperature sensor TS2 is a sensor device installed on column 16d and measures the temperature T2 of column 16d.

[0066] (2-2-5-12. Flow Control Valve Temperature Sensor TS3) The flow control valve temperature sensor TS3 is a sensor device installed on the flow control valve 16f, and measures the temperature T3 of the flow control valve 16f.

[0067] (2-2-6. Oven 16M) Here, we will describe the oven 16M in the gas measurement system 100, which enables pressure control or flow rate control of each component. As shown in Figure 4, the oven 16M is composed of an oven heater 16a, a pressure reducing valve 16b, a sample valve 16c, a column 16d, a detector 16e, a flow rate control valve 16f, a pressure reducing valve drive device 16j, a flow rate control valve drive device 16k, a pressure reducing valve pressure sensor PS, and a flow rate sensor FS. The oven 16M may further include some or all of the components of the oven 16 described above: a pressure reducing valve heater 16g, a column heater 16h, a flow rate control valve heater 16i, a pressure reducing valve temperature sensor TS1, a column temperature sensor TS2, and a flow rate control valve temperature sensor TS3. The configuration examples and processing examples of the oven heater 16a, pressure reducing valve 16b, sample valve 16c, column 16d, detector 16e, and flow rate control valve 16f are the same as those of the oven 16, so we will omit their explanation.

[0068] (2-2-6-1. Pressure Reducing Valve Drive Device 16j) The pressure reducing valve drive device 16j is a drive device installed on the pressure reducing valve 16b, which electrically opens and closes the pressure reducing valve 16b according to the pressure conditions set by the setting unit 15b of the control unit 15 (described later), and maintains a constant pressure. The pressure reducing valve drive device 16j also opens the pressure reducing valve 16b according to the pressure conditions set by the setting unit 15b and maintains a high pressure for a predetermined time. The pressure reducing valve drive device 16j also closes the pressure reducing valve 16b according to the pressure conditions set by the setting unit 15b and maintains a low pressure for a predetermined time.

[0069] (2-2-6-2. Flow Control Valve Drive Device 16k) The flow control valve drive device 16k is a drive device installed on the flow control valve 16f, which electrically opens and closes the flow control valve 16f according to the flow rate conditions set by the setting unit 15b of the control unit 15 (described later), and maintains a constant flow rate. The flow control valve drive device 16k also opens the flow control valve 16f according to the flow rate conditions set by the setting unit 15b, maintaining a high flow rate for a predetermined time. The flow control valve drive device 16k also closes the flow control valve 16f according to the flow rate conditions set by the setting unit 15b, maintaining a low flow rate for a predetermined time.

[0070] (2-2-6-3. Pressure Reducing Valve Pressure Sensor PS) The pressure reducing valve pressure sensor PS is a sensor device installed on the pressure reducing valve 16b, and measures the pressure P1 of the pressure reducing valve 16b.

[0071] (2-2-6-4. Flow Sensor FS) The flow sensor FS is a sensor device installed on the flow control valve 16f, and measures the flow rate F3 of the flow control valve 16f.

[0072] (2-2-7. Control Unit 15) The control unit 15 is responsible for the overall control of the gas chromatograph 10. The control unit 15 consists of a training unit 15a, a setting unit 15b, a measurement unit 15c, and a calculation unit 15d. Here, the control unit 15 can be realized by electronic circuits such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit), or integrated circuits such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

[0073] (2-2-7-1. Training Unit 15a) The training unit 15a trains the machine learning model LM. The training unit 15a may also refer to various information stored in the memory unit 14. The machine learning model training control processes (temperature condition training control process, pressure condition training control process, flow rate condition training control process) will be described below.

[0074] (Machine Learning Model Training Control Processing) The training unit 15a executes machine learning model training control processing. For example, the training unit 15a trains the machine learning model LM for each target component TC by inputting training data consisting of each controlled component's conditions and the peak separation degree R of the target component TC calculated from the measurement data MD under the corresponding conditions. The training unit 15a also trains the machine learning model LM to output each component's conditions that optimize the input peak separation degree R of the target component TC. At this time, the training unit 15a trains the machine learning model LM to output each component's conditions in response to the input of multiple target components TC. Furthermore, the training unit 15a trains the machine learning model LM to output temporal changes (e.g., changes in each condition before and after peak separation of the target component TC) as each component's conditions.

[0075] (Temperature Condition Training Control Processing) The training unit 15a trains the machine learning model LM for each target component TC by inputting training data consisting of the temperature conditions of each component controlled by the heating device and the peak separation degree R of the target component TC calculated from the measurement data MD under the corresponding temperature conditions. At this time, the training unit 15a trains the machine learning model LM to output the temperature conditions of each component in response to the input of multiple target components TC. In addition, the training unit 15a trains the machine learning model LM to output temporal changes (e.g., temperature conditions changed before and after peak separation of the target component TC) as the temperature conditions of each component.

[0076] A specific example of temperature condition training control processing as a machine learning model training control processing will be described. First, the training unit 15a controls the target component TC stored in the measurement data storage unit 14a for the gas chromatograph 10 identified as "GC001". Ａ The temperature conditions for each component corresponding to the measurement data MD "MD111" are referred to as {pressure reducing valve temperature: "T1-111", column temperature: "T2-111", flow control valve temperature: "T3-111"}. Secondly, the training unit 15a refers to the target component TC stored in the measurement data storage unit 14a for the gas chromatograph 10 identified as "GC001". ＡThe measurement data MD, "MD111", is referred to as "PR111" as the peak separation degree R. Thirdly, the training unit 15a refers to the target component TC stored in the measurement data storage unit 14a for the gas chromatograph 10 identified as "GC001". Ａ The training data generated is {pressure reducing valve temperature: "T1-111", column temperature: "T2-111", flow control valve temperature: "T3-111", peak separation degree: "PR111"}. Fourthly, the training unit 15a trains the machine learning model LM, "LM001", stored in the machine learning model storage unit 14b, by inputting the generated training data.

[0077] (Pressure Condition Training Control Processing) The training unit 15a trains the machine learning model LM for each target component TC by inputting training data consisting of the pressure conditions of each component controlled by the drive unit and the peak separation degree R of the target component TC calculated from the measurement data MD under the corresponding pressure conditions. At this time, the training unit 15a trains the machine learning model LM to output the pressure conditions of each component in response to the input of multiple target components TC. The training unit 15a also trains the machine learning model LM to output temporal changes (e.g., changes in pressure conditions before and after peak separation of the target component TC) as pressure conditions for each component.

[0078] A specific example of pressure condition training control processing as a machine learning model training control processing will be described. First, the training unit 15a controls the target component TC stored in the measurement data storage unit 14a for the gas chromatograph 10 identified as "GC001". Ａ The pressure conditions for each component corresponding to the measurement data MD "MD121" are referred to as {pressure reducing valve pressure: "P1-121"}. Secondly, the training unit 15a refers to the target component TC stored in the measurement data storage unit 14a for the gas chromatograph 10 identified as "GC001". ＡRefer to "PR121" as the peak resolution R corresponding to "MD121", which is the measurement data MD. Thirdly, for the gas chromatograph 10 identified by "GC001", the training unit 15a generates {pressure reducing valve pressure: "P1-121", peak resolution: "PR121"} as the training data of the target component stored in the measurement data storage unit 14a. Fourthly, the training unit 15a trains the machine learning model LM, which is "LM002", stored in the machine learning model storage unit 14b by inputting the generated training data.

[0079] (Flow rate condition training control process) The training unit 15a trains the machine learning model LM for each target component TC by inputting the training data of the flow rate conditions of each component controlled by the drive device and the peak resolution R of the target component calculated from the measurement data MD under the corresponding flow rate conditions. At this time, the training unit 15a trains the machine learning model LM so as to output the flow rate conditions of each component in response to the input of a plurality of target components TC. In addition, the training unit 15a trains the machine learning model LM so as to output a temporal change (e.g., changing the flow rate conditions before and after the peak separation of the target component TC) as the flow rate conditions of each component.

[0080] A specific example of the flow rate condition training control process as the machine learning model training control process will be described. First, for the gas chromatograph 10 identified by "GC001", the training unit 15a refers to {flow rate regulating valve flow rate: "F3-131"} as the flow rate conditions of each component corresponding to "MD131", which is the measurement data MD of the target component TC stored in the measurement data storage unit 14a. Second, for the gas chromatograph 10 identified by "GC001", the training unit 15a refers to "PR131" as the peak resolution R corresponding to "MD131", which is the measurement data MD of the target component TC stored in the measurement data storage unit 14a. Third, for the gas chromatograph 10 identified by "GC001", the training unit 15a refers to the target component TC Ａ stored in the measurement data storage unit 14a. Ａ For the measurement data MD of "MD131", "PR131" is referred to as the peak resolution R corresponding thereto. Thirdly, for the gas chromatograph 10 identified by "GC001", the training unit 15a refers to the target component TC ＡAs training data, {flow control valve flow rate: "F3-131", peak separation degree: "PR131"} is generated. Fourthly, the training unit 15a trains the machine learning model LM "LM003" stored in the machine learning model storage unit 14b by inputting the generated training data.

[0081] Furthermore, the training unit 15a can also train a machine learning model LM for each target component TC by inputting training data consisting of combinations of temperature, pressure, and flow rate conditions for each component controlled by the heating device or drive device, and the peak separation degree R of the target component TC calculated from the measurement data MD under the corresponding conditions. In this case, the training unit 15a trains the machine learning model LM to output combinations of temperature, pressure, and flow rate conditions for each component in response to the input of multiple target components TC. The training unit 15a also trains the machine learning model LM to output temporal changes (e.g., changes in temperature conditions, pressure conditions, and flow rate conditions before and after peak separation of the target component TC) as combinations of temperature, pressure, and flow rate conditions for each component.

[0082] (2-2-7-2. Setting Unit 15b) The setting unit 15b acquires various information. The setting unit 15b may also refer to the various information stored in the storage unit 14. The following describes the condition setting process (temperature condition setting process, pressure condition setting process, flow rate condition setting process).

[0083] (Condition Setting Process) The setting unit 15b executes the condition setting process. For example, the setting unit 15b sets predetermined conditions for each component using a trained machine learning model LM that outputs conditions for each component that optimize the peak indicated by the target component TC. At this time, the setting unit 15b sets predetermined conditions for each component so that the peak separation R between the peak indicated by the target component TC and the peak indicated by the adjacent component is greater than or equal to a predetermined value.

[0084] (Temperature condition setting process) The setting unit 15b sets a predetermined temperature T for each component using a trained machine learning model LM that outputs the temperature T for each component that optimizes the peak indicated by the target component TC. At this time, the setting unit 15b sets the predetermined temperature T for each component so that the peak separation R between the peak indicated by the target component TC and the peak indicated by the adjacent component is equal to or greater than a predetermined value.

[0085] As a condition setting process, a specific example of the temperature condition setting process will be explained. First, the setting unit 15b sets the target component TC output by the trained machine learning model LM, "LM001", for the gas chromatograph 10 identified as "GC001". Ａ The optimal temperature conditions for each component are obtained as follows: {Pressure reducing valve temperature: "T1-111-OP", Column temperature: "T2-111-OP", Flow control valve temperature: "T3-111-OP"}. Secondly, the setting unit 15b causes the pressure reducing valve heater 16g to perform temperature control for the pressure reducing valve 16b of the gas chromatograph 10 identified as "GC001", and maintains the temperature T1 of the pressure reducing valve 16b detected by the pressure reducing valve temperature sensor TS1 at "T1-111-OP". Thirdly, the setting unit 15b causes the column heater 16h to perform temperature control for the column 16d of the gas chromatograph 10 identified as "GC001", and maintains the temperature T2 of the column 16d detected by the column temperature sensor TS2 at "T2-111-OP". Fourth, the setting unit 15b causes the flow control valve heater 16i to perform temperature control on the flow control valve 16f of the gas chromatograph 10 identified as "GC001", and maintains the temperature T3 of the flow control valve 16f detected by the flow control valve temperature sensor TS3 at "T3-111-OP".

[0086] (Pressure condition setting process) The setting unit 15b sets a predetermined pressure P for each component using a trained machine learning model LM that outputs the pressure P for each component that optimizes the peak indicated by the target component TC. At this time, the setting unit 15b sets the predetermined pressure P for each component so that the peak separation R between the peak indicated by the target component TC and the peak indicated by the adjacent component is equal to or greater than a predetermined value.

[0087] As a condition setting process, a specific example of the pressure condition setting process will be described. First, the setting unit 15b sets the target component TC output by the trained machine learning model LM, "LM002", for the gas chromatograph 10 identified as "GC001". Ａ The optimal pressure condition for each component, {pressure reducing valve pressure: "P1-121-OP"}, is obtained. Secondly, the setting unit 15b causes the pressure reducing valve drive device 16j to perform pressure control for the pressure reducing valve 16b of the gas chromatograph 10 identified as "GC001", and maintains the pressure P1 of the pressure reducing valve 16b detected by the pressure reducing valve pressure sensor PS at "P1-121-OP".

[0088] (Flow rate condition setting process) The setting unit 15b sets a predetermined flow rate F for each component using a trained machine learning model LM that outputs the flow rate F for each component that optimizes the peak indicated by the target component TC. At this time, the setting unit 15b sets the predetermined flow rate F for each component so that the peak separation R between the peak indicated by the target component TC and the peak indicated by the adjacent component is equal to or greater than a predetermined value.

[0089] As a condition setting process, a specific example of the flow rate condition setting process will be explained. First, the setting unit 15b sets the target component TC output by the trained machine learning model LM, "LM003", for the gas chromatograph 10 identified as "GC001". Ａ The optimal flow rate condition for each component, {flow control valve flow rate: "F3-131-OP"}, is obtained. Secondly, the setting unit 15b causes the flow control valve drive device 16k to perform flow control for the flow control valve 16f of the gas chromatograph 10 identified as "GC001", and maintains the flow rate F3 of the flow control valve 16f detected by the flow sensor FS at "F3-131-OP".

[0090] Furthermore, the setting unit 15b can also set predetermined temperatures T, pressure P, and flow rates F for each component using a trained machine learning model LM that outputs combinations of temperature T, pressure P, and flow rate F for each component that optimize the peak indicated by the target component TC.

[0091] (2-2-7-3. Measurement Unit 15c) The measurement unit 15c outputs the measurement result. The measurement unit 15c may store the outputted measurement result in the storage unit 14. The measurement unit 15c may also refer to various information stored in the storage unit 14. The measurement result output control process will be described below.

[0092] (Measurement Result Output Control Processing) The measurement unit 15c performs measurement result output control processing. For example, the measurement unit 15c separates each component contained in the sample SP by gas chromatography using each component, each equipped with a heating device and controlled to a predetermined temperature T, and outputs measurement data MD showing the concentration of the separated target component TC. The measurement unit 15c also outputs measurement data MD showing the concentrations of multiple separated target components TC. The measurement unit 15c also measures process data in the plant. In this case, each component includes a pressure reducing valve 16b on which a pressure reducing valve heater 16g is installed. Each component also includes a column 16d on which a column heater 16h is installed. Each component also includes a flow control valve 16f on which a flow control valve heater 16i is installed.

[0093] Furthermore, the measurement unit 15c uses components, each equipped with a drive device and controlled to a predetermined pressure P by the drive device, to separate the components contained in the sample SP by gas chromatography and output measurement data MD indicating the concentration of the separated target component TC. The measurement unit 15c also measures process data in the plant. In this case, each component includes a pressure reducing valve 16b on which a pressure reducing valve drive device 16j is installed.

[0094] Furthermore, the measurement unit 15c uses components, each equipped with a drive device and controlled by the drive device to a predetermined flow rate F, to separate the components contained in the sample SP by gas chromatography and output measurement data MD indicating the concentration of the separated target component TC. The measurement unit 15c also measures process data in the plant. In this case, each component includes a flow control valve 16f on which a flow control valve drive device 16k is installed.

[0095] A specific example of the measurement result output control process will be described. First, the measurement unit 15c performs gas chromatography on the gas chromatograph 10 identified as "GC001" using each component set to optimal conditions by the setting unit 15b, and displays a chromatogram showing the detection results of each component output by the detector 16e. Second, the measurement unit 15c converts the detection results of each component output by the detector 16e on the gas chromatograph 10 identified as "GC001" into concentrations. Third, the measurement unit 15c stores "MD101", "MD102", "MD103", ... as measurement data MDs including the chromatogram and the concentrations of each component on the gas chromatograph 10 identified as "GC001" in the measurement data storage unit 14a.

[0096] (2-2-7-4. Calculation Unit 15d) The calculation unit 15d calculates various information. The calculation unit 15d may store the calculated information in the storage unit 14. The calculation unit 15d may also refer to the information stored in the storage unit 14. The peak separation degree calculation control process will be described below.

[0097] (Peak Separation Calculation Control Processing) The calculation unit 15d executes peak separation calculation control processing. For example, the calculation unit 15d calculates the peak separation R, which is a measure of how far the target component TC is separated from adjacent peaks on the chromatogram. Details of the peak separation R will be described later.

[0098] A specific example of the peak separation calculation control process will be described. First, the calculation unit 15d refers to "MD101", "MD102", "MD103", ... for the gas chromatograph 10 identified as "GC001" as measurement data MD stored in the measurement data storage unit 14a. Second, the calculation unit 15d calculates "PR101", "PR102", "PR103", ... as the peak separation R for the gas chromatograph 10 identified as "GC001". Third, the calculation unit 15d stores the calculated peak separation R in the measurement data storage unit 14a.

[0099] [3. Specific Examples of Each Process of the Gas Measurement System 100] Specific examples of each process of the gas measurement system 100 according to the embodiment will be described below. Specific examples of the peak separation optimization process, the peak separation calculation process, and the data display process will be described below.

[0100] (3-1. Specific Examples of Peak Separation Optimization Process) Specific examples of peak separation optimization process are explained below. The algorithm for performing peak separation optimization is described below.

[0101] (3-1-1. Initial Setup and Data Collection) The gas chromatograph 10 sets initial conditions and collects various data. First, the gas chromatograph 10 sets the initial conditions (initial temperature, initial pressure, initial flow rate) appropriately. Second, the gas chromatograph 10 measures the gas chromatogram. Third, the gas chromatograph 10 calculates feature quantities such as peak position, peak height, and peak width under the initial conditions and calculates the peak separation R between adjacent peaks. Fourth, the gas chromatograph 10 uses machine learning to train the machine learning model LM on the initial conditions and the peak separation R.

[0102] (3-1-2. Adjustment of Conditions) The gas chromatograph 10 adjusts the conditions. Firstly, the gas chromatograph 10 gradually changes the controlled conditions (temperature, pressure, flow rate) and records the change in peak position for each condition. Secondly, the gas chromatograph 10 sets a certain range for the change in conditions, fine-tunes it, and collects multiple datasets.

[0103] (3-1-3. Modeling of Peak Position) The gas chromatograph 10 models the peak position. First, the gas chromatograph 10 calculates the peak separation R. Second, after calculating the peak separation L times, the gas chromatograph 10 models the change in peak position in response to changes in conditions, i.e., designs an evaluation function. The gas chromatograph 10 predicts how the peak position will change in response to conditions by using the machine learning model LM designed as described above.

[0104] (3-1-4. Optimization Loop) The gas chromatograph 10 executes an optimization loop. First, the gas chromatograph 10 sets the optimal conditions (optimal temperature, optimal pressure, optimal flow rate) based on the designed machine learning model LM. Second, the gas chromatograph 10 measures the gas chromatogram under the set optimal conditions. Third, the gas chromatograph 10 calculates feature quantities such as peak position, peak height, and peak width under the optimal conditions and evaluates the peak separation R between adjacent peaks. Fourth, the gas chromatograph 10 sets the target peak separation R, which is the target value of the evaluation standard peak separation R. ｇｏａｌ It is determined whether or not the target peak separation R is reached. At this time, the gas chromatograph 10 determines the target peak separation R according to the object being measured. ｇｏａｌ This can be set arbitrarily.

[0105] (3-1-5. Output of final conditions) The gas chromatograph 10 outputs the final conditions. The gas chromatograph 10 outputs the peak resolution R to the target peak resolution R. ｇｏａｌ If the target is reached, the optimal conditions (optimal temperature, optimal pressure, optimal flow rate) and peak separation R are output as the final conditions.

[0106] (3-1-6. Model Retraining) The gas chromatograph 10 retrains the machine learning model LM. The gas chromatograph 10 sets the peak resolution R to the target peak resolution R. ｇｏａｌ If the target peak separation R is not reached, the machine learning model LM is retrained and its parameters are optimized. At this time, the gas chromatograph 10 changes the conditions to maximize the peak separation R, dynamically adjusts the conditions (temperature, pressure, flow rate) of each component in the oven 16 based on the changed conditions, calculates the peak separation R under each adjusted condition, and uses the results as feedback to retrain the machine learning model LM. Then, the gas chromatograph 10 checks if the peak separation R after retraining is equal to the target peak separation R. ｇｏａｌ If the condition is reached, the final condition will be output.

[0107] (3-1-7. Model Redesign) The gas chromatograph 10 redesigns the machine learning model LM. The gas chromatograph 10 redesigns the machine learning model LM after the set number of model retrainings, which is M times. At this time, the gas chromatograph 10 repeatedly adjusts the conditions to maximize the peak separation R. Then, the gas chromatograph 10 determines that the peak separation R after redesign is equal to the target peak separation R. ｇｏａｌ If the condition is reached, the final condition will be output.

[0108] (3-1-8. Alarm Output) The gas chromatograph 10 outputs an alarm. After the set number of model redesigns N times, the gas chromatograph 10 sets the target peak separation R. ｇｏａｌ If a machine learning model (LM) that satisfies the conditions cannot be created, the algorithm will stop and an error message will be output as an alarm.

[0109] (3-1-9. Setting the Final Conditions) The gas chromatograph 10 sets the final conditions. The gas chromatograph 10 sets the optimal conditions (optimal temperature, optimal pressure, optimal flow rate), which are the final conditions for optimizing the peak separation degree R, as the operating conditions.

[0110] (3-2. Specific Examples of Peak Resolution Calculation Process) Specific examples of the peak resolution calculation process will be explained using Figure 7 and mathematical formulas. Figure 7 is a diagram showing a specific example of the peak resolution calculation process of the gas chromatograph 10 according to the embodiment. Below, the outline of the gas chromatogram will be explained, followed by the definition of peak resolution R.

[0111] (3-2-1. Outline of Gas Chromatogram) An outline of the gas chromatogram will be explained using Figure 7. As shown in Figure 7, component C appears as two adjacent peaks on the gas chromatogram. Ａ Peak P Ａ and component C Ｂ Peak P Ｂ This is shown. In the gas chromatogram in Figure 7, the vertical axis represents the detector signal. Also, the horizontal axis of the gas chromatogram in Figure 7 represents the elapsed time. Here, component C Ａ The holding time is t ＲＡ , component C ＡThe peak width is W Ａ It is shown as follows. Also, component C Ｂ The holding time is t ＲＢ , component C Ｂ The peak width is W Ｂ This is shown.

[0112] (3-2-2. Definition of Peak Resolve R) The definition of peak resolve R between two adjacent peaks on a gas chromatogram will be explained using a mathematical formula. Peak resolve R is a measure of how well the target component is separated from adjacent peaks, and is defined as shown in equation (1) below.

[0113]

[0114] As shown in equation (1) above, the peak separation R is equal to the component C Ａ Retention time t ＲＡ and component C Ｂ Retention time t ＲＢ The difference between this and component C Ａ Peak width W Ａ and component C Ｂ Peak width W Ｂ It is defined as the value obtained by dividing by the average value of the two factors. Furthermore, the peak separation R is defined as the peak width at half maximum W. １／２ Using this, it is defined as shown in equation (2) below.

[0115]

[0116] Furthermore, assuming the peak shape follows a normal distribution, W = 4σ, W １／２ This is approximately 2.35σ. Furthermore, a peak separation R of 1.5 or higher is considered complete separation.

[0117] (3-3. Specific Examples of Data Display Processing) Specific examples of data display processing will be explained below. The following describes the data that is displayed on the gas chromatograph 10 and provided to the operator W when executing the peak separation optimization processing algorithm described in (3-1. Specific Examples of Peak Separation Optimization Processing) above.

[0118] (3-3-1. Initial Setup and Data Acquisition) After the initial setup is complete, during data acquisition, and upon completion of data acquisition, the gas chromatograph 10 displays data such as initial conditions (initial temperature, initial pressure, initial flow rate), progress during chromatogram measurement, measured peak positions, peak heights, peak widths and other characteristic quantities, and the calculated peak resolution R.

[0119] (3-3-2. Adjustment of Conditions) When adjusting the conditions, the gas chromatograph 10 displays data such as the current conditions (temperature, pressure, flow rate), the progress of the chromatogram measurement, and the change in peak position under each condition, and updates this data in real time whenever the conditions change.

[0120] (3-3-3. Modeling of Peak Position) The gas chromatograph 10 displays data such as predictions of changes in peak position in relation to conditions (temperature, pressure, flow rate) during and after the modeling of peak position is complete.

[0121] (3-3-4. Optimization Loop) The gas chromatograph 10 displays data such as the optimal conditions (optimal temperature, optimal pressure, optimal flow rate), measured peak position, and calculated peak separation R at the start of the optimization loop and at the completion of each optimization loop.

[0122] (3-3-5. Final Condition Output and Model Retraining) The gas chromatograph 10 displays data such as the current conditions (temperature, pressure, flow rate), the calculated peak separation R, and the results of condition changes due to feedback during the feedback loop and upon completion of each feedback loop.

[0123] (3-3-6. Model Redesign) When redesigning the machine learning model LM, the gas chromatograph 10 displays data such as the current conditions (temperature, pressure, flow rate), the progress of peak resolution optimization, and the change in peak resolution R, and updates this data in real time each time the conditions are adjusted.

[0124] (3-3-7. Alarm Output) When the creation of the machine learning model LM fails, the gas chromatograph 10 displays error messages and other data.

[0125] (3-3-8. Setting the final conditions) When outputting the final conditions, the gas chromatograph 10 displays as data the optimal conditions (optimal temperature, optimal pressure, optimal flow rate), the optimized peak separation R, an overview of the entire process, and whether the settings at the time of outputting the final conditions can be fixed.

[0126] (3-4. Examples of applying peak separation optimization processing) Examples of applying peak separation optimization processing will be explained below. Examples 1 to 4 of applying peak separation optimization processing will be explained below.

[0127] (3-4-1. Application Example 1) As an application example 1 of the peak separation optimization process, an example from the initial specification review will be described. In application example 1 of the peak separation optimization process, the initial conditions for the gas chromatograph 10 are set such that the peak separation R is maximized when each component in the oven 16 is under constant conditions. In this case, in application example 1 of the peak separation optimization process, the initial conditions are kept constant throughout the analysis interval.

[0128] (3-4-2. Application Example 2) As an application example 2 of the peak separation optimization process, an example at the manufacturing stage will be described. In application example 2 of the peak separation optimization process, initial conditions are input to the gas chromatograph 10 that will actually be sold, and the optimization loop is executed by the algorithm to determine the final conditions.

[0129] (3-4-3. Application Example 3) As an application example 3 of the peak separation optimization process, we will describe an example of initial operation at the site. In application example 3 of the peak separation optimization process, the optimization loop is executed again at the time of installation to respond to fluctuations due to the environment and other factors at the site, such as the plant. At this time, in application example 3 of the peak separation optimization process, the initial conditions are set to the optimal conditions output as the final conditions at the manufacturing stage. In application example 3 of the peak separation optimization process, the conditions output in the above optimization loop are fixed to the condition fluctuation settings of the equipment, so that subsequent chromatogram measurements are performed under conditions in which the peak separation R is optimized.

[0130] (3-4-4. Application Example 4) As an application example 4 of the peak separation optimization process, we will explain an example of its use in field operations. In application example 4 of the peak separation optimization process, an alarm is output when the peak separation R falls below an acceptable value, in response to changes in the environment or equipment status. In addition, in application example 4 of the peak separation optimization process, the machine learning model LM is calibrated by executing an optimization loop at any time after the alarm is output. At this time, in application example 4 of the peak separation optimization process, it is possible to arbitrarily choose whether to redesign the evaluation function (model redesign) or to perform only parameter optimization (model retraining).

[0131] [4. Flow of each process in the gas measurement system 100] The flow of the gas measurement system 100 according to the embodiment will be explained using Figures 8 to 11. Below, the overall flow of the gas measurement system 100 will be explained, and then the model design process, the optimal condition identification process, and the measurement control process will be explained as individual processes.

[0132] (4-1. Overall Processing of the Gas Measurement System 100) The overall processing flow of the gas measurement system 100 according to the embodiment will be explained using Figure 8. Figure 8 is a flowchart showing an example of the overall processing flow of the gas measurement system 100 according to the embodiment. Note that the processes in steps S101 to S109 below can be executed in a different order. Also, some of the processes in steps S101 to S109 below may be omitted.

[0133] (4-1-1. Model Design Process) First, the gas chromatograph 10 performs a model design process (step S101). For example, the gas chromatograph 10 trains a machine learning model LM that outputs the conditions of each component of the gas chromatograph 10 by performing the processes described in steps S201 to S207.

[0134] (4-1-2. Optimal Condition Identification Process) Secondly, the gas chromatograph 10 performs an optimal condition identification process (step S102). For example, the gas chromatograph 10 identifies the optimal conditions for each component of the gas chromatograph 10 by performing the processes described in steps S301 to S303.

[0135] (4-1-3. Peak Resolving Ability Determination Process) Thirdly, the gas chromatograph 10 performs a peak resolving ability determination process (step S103). At this time, if the peak resolving ability R of the target component TC is equal to or greater than the target value (step S103: Yes), the gas chromatograph 10 proceeds to the process in step S104. On the other hand, if the peak resolving ability R of the target component TC is not equal to or greater than the target value (step S103: No), the gas chromatograph 10 proceeds to the process in step S105.

[0136] (4-1-4. Final Condition Output Processing) Fourth, the gas chromatograph 10 performs final condition output processing (step S104). For example, the gas chromatograph 10 displays the final conditions for each component of the gas chromatograph 10.

[0137] (4-1-5. Measurement Control Processing) Fifth, the gas chromatograph 10 performs measurement control processing (step S109) and completes the overall process. For example, the gas chromatograph 10 sets optimal conditions for each component of the gas chromatograph 10 by performing the processes described in steps S401 to S403, performs gas chromatography, and outputs measurement data MD including a chromatogram and the concentrations of each component.

[0138] (4-1-6. Retraining Determination Process) Sixth, the gas chromatograph 10 performs a retraining determination process (step S105). At this time, if the machine learning model LM has been retrained more than a predetermined number of times (step S105: Yes), the gas chromatograph 10 proceeds to the process in step S107. On the other hand, if the machine learning model LM has not been retrained more than a predetermined number of times (step S105: No), the gas chromatograph 10 proceeds to the process in step S106.

[0139] (4-1-7. Redesign Judgment Process) Seventh, the gas chromatograph 10 performs a redesign judgment process (step S107). At this time, if the machine learning model LM has been redesigned more than a predetermined number of times (step S107: Yes), the gas chromatograph 10 proceeds to the process in step S108. On the other hand, if the machine learning model LM has not been redesigned more than a predetermined number of times (step S107: No), the process returns to the process in step S101.

[0140] (4-1-8. Parameter Optimization Process) Eighth, the gas chromatograph 10 performs a parameter optimization process (step S106) and returns to the process in step S102. For example, the gas chromatograph 10 retrains the machine learning model LM by optimizing the parameters.

[0141] (4-1-9. Alarm Output Processing) Ninth, the gas chromatograph 10 performs alarm output processing (step S108) and terminates the entire process. For example, the gas chromatograph 10 displays an error message indicating that the machine learning model LM could not be created.

[0142] (4-2. Model Design Process) The flow of the model design process for the gas measurement system 100 according to the embodiment will be explained using Figure 9. Figure 9 is a flowchart showing an example of the flow of the model design process for the gas measurement system 100 according to the embodiment. Note that the processes in steps S201 to S207 below can be executed in a different order. Also, some of the processes in steps S201 to S207 below may be omitted.

[0143] (4-2-1. Initial Condition Setting Process) First, the gas chromatograph 10 performs an initial condition setting process (step S201). For example, the gas chromatograph 10 sets initial conditions (initial temperature, initial pressure, initial flow rate) for each component of the gas chromatograph 10.

[0144] (4-2-2. Chromatogram Measurement Process) Secondly, the gas chromatograph 10 performs a chromatogram measurement process (step S202). For example, the gas chromatograph 10 performs gas chromatography using each component under the set initial conditions and outputs a chromatogram and measurement data MD showing the concentrations of each component.

[0145] (4-2-3. Peak Resolve Calculation Process) Thirdly, the gas chromatograph 10 performs a peak resolve calculation process (step S203). For example, the gas chromatograph 10 calculates the peak resolve R of the target component TC based on the chromatogram included in the measurement data MD.

[0146] (4-2-4. Model Training Process) Fourth, the gas chromatograph 10 performs a model training process (step S204). For example, the gas chromatograph 10 learns the relationship between each condition and the peak resolution R for each target component TC by inputting training data of the initial conditions of each component and the corresponding peak resolution R into the machine learning model LM.

[0147] (4-2-5. Training Determination Process) Fifth, the gas chromatograph 10 performs a training determination process (step S205). At this time, if the training of the machine learning model LM has been performed at a predetermined number of times or more (step S205: Yes), the gas chromatograph 10 proceeds to the process in step S207. On the other hand, if the training of the machine learning model LM has not been performed at a predetermined number of times or more (step S205: No), the gas chromatograph 10 proceeds to the process in step S206.

[0148] (4-2-6. Condition Change Processing) Sixth, the gas chromatograph 10 performs condition change processing (step S206) and returns to the process in step S201. For example, the gas measurement system 100 makes a small change to the initial conditions and sets each of the changed conditions.

[0149] (4-2-7. Evaluation Function Design Process) Seventh, the gas chromatograph 10 executes the evaluation function design process (step S207) and completes the model design process. For example, the gas chromatograph 10 completes the design of a machine learning model LM as the evaluation function, which machine learning the relationship between each condition and the peak separation degree R for each target component TC.

[0150] (4-3. Optimal Condition Identification Process) The flow of the optimal condition identification process for the gas measurement system 100 according to the embodiment will be explained using Figure 10. Figure 10 is a flowchart showing an example of the flow of the optimal condition identification process for the gas measurement system 100 according to the embodiment. Note that the processes in steps S301 to S303 below can be executed in a different order. Also, some of the processes in steps S301 to S303 below may be omitted.

[0151] (4-3-1. Optimal Condition Prediction Processing) First, the gas chromatograph 10 performs optimal condition prediction processing (step S301). For example, the gas chromatograph 10 obtains the optimal conditions (optimal temperature, optimal pressure, optimal flow rate) for the target component TC from the machine learning model LM and sets them in each component of the gas chromatograph 10.

[0152] (4-3-2. Chromatogram Measurement Process) Secondly, the gas chromatograph 10 performs a chromatogram measurement process (step S302). For example, the gas chromatograph 10 performs gas chromatography using each component under the set optimal conditions and outputs measurement data MD including a chromatogram and the concentrations of each component.

[0153] (4-3-3. Peak Resolving Ability Calculation Process) Thirdly, the gas chromatograph 10 performs the peak resolving ability calculation process (step S303) and terminates the optimal condition identification process. For example, the gas chromatograph 10 calculates the peak resolving ability R of the target component TC based on the chromatogram included in the measurement data MD.

[0154] (4-4. Measurement Control Processing) The flow of the measurement control processing of the gas measurement system 100 according to the embodiment will be explained using Figure 11. Figure 11 is a flowchart showing an example of the flow of the measurement control processing of the gas measurement system 100 according to the embodiment. Note that the processes in steps S401 to S403 below can be executed in a different order. Also, some of the processes in steps S401 to S403 below may be omitted.

[0155] (4-4-1. Final Condition Setting Process) First, the gas chromatograph 10 performs the final condition setting process (step S401). For example, the gas chromatograph 10 sets the final conditions for each component of the gas chromatograph 10 such that the peak separation R of the target component TC is equal to or greater than the target value.

[0156] (4-4-2. Measurement Result Output Processing) Secondly, the gas chromatograph 10 performs measurement result output processing (step S402). For example, the gas chromatograph 10 performs gas chromatography using each component of the set final conditions and outputs measurement data MD including a chromatogram and the concentrations of each component.

[0157] (4-4-3. Measurement Result Storage Process) Thirdly, the gas chromatograph 10 performs the measurement result storage process (step S403) and terminates the measurement control process. For example, the gas chromatograph 10 stores the output chromatogram and the measurement data MD, which includes the concentrations of each component, in the measurement data storage unit 14a.

[0158] [5. Effects of the Embodiment] The effects of the embodiment will be described below. Effects 1 to 11 corresponding to the processing according to the embodiment will be described below.

[0159] (5-1. Effect 1) Firstly, in this embodiment, the gas chromatograph 10 is equipped with heating devices, and each component is controlled to a predetermined temperature by these heating devices to separate each component contained in the sample SP by gas chromatography and output measurement data MD indicating the concentration of the separated target component TC. Therefore, in this embodiment, each component of the gas chromatograph 10 is temperature controllable, which improves the measurement accuracy of gas chromatography.

[0160] (5-2. Effect 2) Secondly, in this embodiment, the gas chromatograph 10 is equipped with a drive device, and each component is controlled to a predetermined pressure by the drive device to separate the components contained in the sample SP by gas chromatography and output measurement data MD indicating the concentration of the separated target component TC. Therefore, in this embodiment, each component of the gas chromatograph 10 is pressure-controllable, which improves the measurement accuracy of gas chromatography.

[0161] (5-3. Effect 3) Thirdly, in the embodiment, the gas chromatograph 10 sets a predetermined temperature T for each component using a trained machine learning model LM that outputs the temperature T for each component that optimizes the peak shown by the target component TC. For this reason, in the embodiment, the optimal temperature T for each component of the gas chromatograph 10 ＯＰ By using the machine learning model LM that outputs the result, the measurement accuracy of gas chromatography can be improved.

[0162] (5-4. Effect 4) Fourth, in the embodiment, the gas chromatograph 10 sets a predetermined pressure P for each component using a trained machine learning model LM that outputs the pressure P for each component that optimizes the peak shown by the target component TC. For this reason, in the embodiment, the optimal pressure P for each component of the gas chromatograph 10 ＯＰ By using the machine learning model LM that outputs the result, the measurement accuracy of gas chromatography can be improved.

[0163] (5-5. Effect 5) Fifth, in the embodiment, the gas chromatograph 10 sets a predetermined temperature T for each component so that the peak separation R between the peak shown by the target component TC and the peak shown by the adjacent component is equal to or greater than a predetermined value. For this reason, in the embodiment, the optimal temperature T for each component of the gas chromatograph 10 ＯＰ By setting this parameter, the separation of the target component TC peak can be optimized, thereby improving the measurement accuracy of gas chromatography.

[0164] (5-6. Effect 6) Sixth, in the embodiment, the gas chromatograph 10 sets a predetermined pressure P for each component so that the peak separation R between the peak shown by the target component TC and the peak shown by the adjacent component is equal to or greater than a predetermined value. For this reason, in the embodiment, the optimal pressure P for each component of the gas chromatograph 10 ＯＰ By setting this parameter, the separation of the target component TC peak can be optimized, thereby improving the measurement accuracy of gas chromatography.

[0165] (5-7. Effect 7) Seventh, in the embodiment, the gas chromatograph 10 includes a column 16d on which a column heater 16h is installed as a component. Therefore, in the embodiment, the temperature T2 of the column 16d of the gas chromatograph 10 can be optimized, thereby improving the measurement accuracy of gas chromatography.

[0166] (5-8. Effect 8) Eighth, in the embodiment, the gas chromatograph 10 includes a pressure reducing valve 16b on which a pressure reducing valve heater 16g is installed as a component. Therefore, in the embodiment, the temperature T1 of the pressure reducing valve 16b of the gas chromatograph 10 can be optimized, thereby improving the measurement accuracy of gas chromatography.

[0167] (5-9. Effect 9) Ninthly, in the embodiment, the gas chromatograph 10 includes a flow control valve 16f on which a flow control valve heater 16i is installed as a component. Therefore, in the embodiment, the temperature T3 of the flow control valve 16f of the gas chromatograph 10 can be optimized, thereby improving the measurement accuracy of gas chromatography.

[0168] (5-10. Effect 10) In the tenth embodiment, the gas chromatograph 10 includes a pressure reducing valve 16b on which a pressure reducing valve drive device 16j is installed as a component. Therefore, in the embodiment, the pressure P1 of the pressure reducing valve 16b of the gas chromatograph 10 can be optimized, thereby improving the measurement accuracy of gas chromatography.

[0169] (5-11. Effect 11) Eleventhly, in this embodiment, the gas chromatograph 10 measures process data in the plant. Therefore, in this embodiment, the measurement accuracy of process gas chromatography can be improved.

[0170] [6. System] Unless otherwise specified, the processing procedures, control procedures, specific names, and various data and parameters shown in the above documents and drawings may be changed at will.

[0171] Furthermore, the components of each illustrated device are functionally conceptual and do not necessarily need to be physically configured as shown. In other words, the specific forms of distribution and integration of each device are not limited to those shown. That is, all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions.

[0172] Furthermore, each processing function performed by each device can be implemented, in whole or in part, by a CPU and a program executed for analysis by that CPU, or by hardware using wired logic.

[0173] [7. Hardware] An example of the hardware configuration of the gas chromatograph 10 will be described using Figure 12. Note that other devices can have a similar hardware configuration. Figure 12 is a diagram showing an example of the hardware configuration according to the embodiment. As shown in Figure 12, the gas chromatograph 10 has a communication device 10a, an HDD (Hard Disk Drive) 10b, memory 10c, and a processor 10d. Furthermore, each part shown in Figure 12 is interconnected by a bus or the like.

[0174] The communication device 10a is implemented using a network interface card or the like, and communicates with other servers. The HDD 10b stores programs and databases that operate the functions shown in Figure 3.

[0175] The processor 10d operates a process that performs the functions described in Figure 3 by reading a program that performs the same processing as each processing unit shown in Figure 3 from the HDD 10b or the like and loading it into memory 10c. For example, this process performs the same functions as each processing unit of the gas chromatograph 10. Specifically, the processor 10d reads a program that has the same functions as the training unit 15a, setting unit 15b, measurement unit 15c, calculation unit 15d, etc. from the HDD 10b or the like. Then, the processor 10d executes a process that performs the same processing as the training unit 15a, setting unit 15b, measurement unit 15c, calculation unit 15d, etc.

[0176] As described above, the gas chromatograph 10 operates as a device that performs various processing methods by reading and executing the program according to the embodiment. Furthermore, the gas chromatograph 10 can also achieve the same functionality as the embodiment by reading the program from the recording medium using a media reader and executing the read program. Note that the program according to the embodiment is not limited to being executed by the gas chromatograph 10. For example, this disclosure can be similarly applied when another computer or server executes the program, or when they cooperate to execute the program.

[0177] The program according to this embodiment can be distributed via a network such as the Internet. Furthermore, this program can be recorded on a computer-readable recording medium such as a hard disk, flexible disk (FD), CD-ROM, MO (Magneto-Optical disk), or DVD (Digital Versatile Disc), and executed by reading it from the recording medium by a computer.

[0178] [8. Other] Some examples of combinations of technical features to be disclosed are listed below.

[0179] (1) A measuring device comprising a component and a processor connected to each component, wherein each component is equipped with a heating device, and the processor performs the following: using each component controlled by the heating device to a predetermined temperature, separates each component contained in a sample by gas chromatography, and outputs measurement data indicating the concentration of the separated target component.

[0180] (2) A measuring device comprising a component and a processor connected to each component, wherein each component is equipped with a drive device, and the processor performs the following: using each component controlled by the drive device to a predetermined pressure, separates each component contained in a sample by gas chromatography and outputs measurement data indicating the concentration of the separated target component.

[0181] (3) The measuring apparatus according to (1), wherein the processor performs the following: setting the predetermined temperature of each component using a trained machine learning model that outputs the temperature of each component that optimizes the peak indicated by the target component.

[0182] (4) The measuring device according to (2), wherein the processor sets the predetermined pressure for each of the components using a trained machine learning model that outputs the pressure for each of the components that optimizes the peak indicated by the target component.

[0183] (5) The measuring device according to (1) or (3), wherein the processor sets the predetermined temperature of each component such that the degree of separation between the peak indicated by the target component and the peak indicated by the adjacent component is equal to or greater than a predetermined value.

[0184] (6) The measuring device according to (2) or (4), wherein the processor sets the predetermined pressure of each component such that the degree of separation between the peak indicated by the target component and the peak indicated by the adjacent component is equal to or greater than a predetermined value.

[0185] (7) The measuring apparatus according to (1), (3), or (5), wherein each of the components includes a column on which the heating device is installed.

[0186] (8) The measuring device according to (1), (3), (5), or (7), wherein each of the components includes a pressure reducing valve on which the heating device is installed.

[0187] (9) The measuring device according to (1), (3), (5), (7), or (8), wherein each of the components includes a flow control valve on which the heating device is installed.

[0188] (10) The measuring device according to (2), (4), or (6), wherein each of the components includes a pressure reducing valve on which the drive device is installed.

[0189] (11) The processor is a measuring device according to any one of (1) to (10) for measuring process data in a plant.

[0190] (12) A measurement method in which a computer performs the following actions: separates each component contained in a sample by gas chromatography using each component equipped with a heating device, each of which is controlled to a predetermined temperature; and outputs measurement data indicating the concentration of the separated target component.

[0191] (13) A measurement method in which a computer performs the following: separates each component contained in a sample by gas chromatography using each component, each of which is equipped with a drive device and controlled to a predetermined pressure by the drive devices, and outputs measurement data indicating the concentration of the separated target component.

[0192] (14) A measurement program that causes a computer to perform the following actions: separate each component contained in a sample by gas chromatography using each component equipped with a heating device, each of which is controlled to a predetermined temperature by the heating device, and output measurement data indicating the concentration of the separated target component.

[0193] (15) A measurement program that causes a computer to perform the following actions: separate each component contained in a sample by gas chromatography using components each equipped with a drive device and controlled to a predetermined pressure by the drive device, and output measurement data indicating the concentration of the separated target component.

[0194] (16) A measuring device comprising a component and a processor connected to each component, wherein each component is equipped with a heating device and a drive device, and the processor performs the following: separate each component contained in a sample by gas chromatography using each component, which is controlled by the heating device to a predetermined temperature and by the drive device to a predetermined pressure, and output measurement data indicating the concentration of the separated target component.

[0195] 10 Gas chromatograph 10a Communication device 10b HDD 10c Memory 10d Processor 11 Input unit 12 Output unit 13 Communication unit 14 Storage unit 14a Measurement data storage unit 14b Machine learning model storage unit 15 Control unit 15a Training unit 15b Setting unit 15c Measurement unit 15d Calculation unit 16, 16M Oven 16a Oven heater 16b Pressure reducing valve 16c Sample valve 16d Column 16e Detector 16f Flow control valve 16g Pressure reducing valve heater 16h Column heater 16i Flow control valve heater 16j Pressure reducing valve drive unit 16k Flow control valve drive unit 100 Gas measurement system TS1 Pressure reducing valve temperature sensor TS2 Column temperature sensor TS3 Flow control valve temperature sensor PS Pressure reducing valve pressure sensor FS Flow sensor W Operator

Claims

1. A measuring device comprising: a component and a processor connected to each component, wherein each component is equipped with a heating device, and the processor performs the following: separating each component contained in a sample by gas chromatography using each component, which has been controlled to a predetermined temperature by the heating device, and outputting measurement data indicating the concentration of the separated target component.

2. A measuring device comprising: a component and a processor connected to each of the components, wherein each component is equipped with a drive device, and the processor performs the following: separating each component contained in a sample by gas chromatography using each component controlled to a predetermined pressure by the drive device, and outputting measurement data indicating the concentration of the separated target component.

3. The measuring device according to claim 1, wherein the processor performs the following: setting the predetermined temperature of each component using a trained machine learning model that outputs the temperature of each component that optimizes the peak indicated by the target component.

4. The measuring device according to claim 2, wherein the processor performs the following: setting the predetermined pressure for each of the components using a trained machine learning model that outputs the pressure for each of the components that optimizes the peak indicated by the target component.

5. The measuring device according to claim 1 or 3, wherein the processor sets the predetermined temperature of each component such that the degree of separation between the peak represented by the target component and the peak represented by an adjacent component is equal to or greater than a predetermined value.

6. The measuring device according to claim 2 or 4, wherein the processor sets the predetermined pressure of each component such that the degree of separation between the peak represented by the target component and the peak represented by the adjacent component is greater than or equal to a predetermined value.

7. The measuring apparatus according to claim 1, wherein each of the components includes a column on which the heating device is installed.

8. The measuring device according to claim 1, wherein each of the components includes a pressure reducing valve on which the heating device is installed.

9. The measuring device according to claim 1, wherein each of the components includes a flow control valve on which the heating device is installed.

10. The measuring device according to claim 2, wherein each of the components includes a pressure reducing valve on which the drive device is installed.

11. The measuring device according to claim 1 or 2, wherein the processor measures process data in a plant.

12. A measurement method comprising: a computer performing the following actions: separating each component contained in a sample by gas chromatography using each component equipped with a heating device, each of which is controlled to a predetermined temperature; and outputting measurement data indicating the concentration of the separated target component.

13. A measurement method comprising: a computer performing the following actions: separating each component contained in a sample by gas chromatography using components each equipped with a drive device, each of which is controlled to a predetermined pressure; and outputting measurement data indicating the concentration of the separated target component.

14. A measurement program that causes a computer to perform the following actions: separate each component contained in a sample by gas chromatography using components each equipped with a heating device and controlled to a predetermined temperature by the heating device, and output measurement data indicating the concentration of the separated target component.

15. A measurement program that causes a computer to perform the following actions: separate each component contained in a sample by gas chromatography using components each driven by a drive device controlled to a predetermined pressure, and output measurement data indicating the concentration of the separated target component.

Images