Analysis device, analysis method, and computer-readable recording medium

Abstract

The present invention relates to an analysis device, an analysis method and a computer-readable recording medium, which detects the degradation of TCD and suppresses the reduction of sensitivity. The analysis device extracts the baseline part in the chromatographic data obtained from the instrument. The normality of the instrument is determined based on the characteristics of the baseline part of the extracted chromatographic data and the chromatographic data when it is normal.

Description

Technical Field

[0001] The present invention relates to an analysis device, an analysis method, and a computer-readable recording medium. Background Art

[0002] A gas chromatograph known as a TCD (Thermal Conductivity Detector) is known. A gas chromatograph is a gas analyzer that separates a mixed gas of multiple components or a volatile liquid into each single component for detection. It can measure multiple components simultaneously without being interfered by other components. In a thermostat controlled at a constant temperature, a certain amount of a measurement sample is collected and introduced into a chromatographic column using a carrier gas. In the chromatographic column, the measurement sample of multiple components is separated into each single component, and they are eluted from the chromatographic column in sequence. A detector converts the single component eluted from the chromatographic column into an electrical signal to obtain a chromatogram. The concentration of each component is calculated based on the peak area of each obtained chromatogram (chromatographic data).

[0003] A TCD is a detector that extracts an unbalanced voltage generated in a bridge circuit as a concentration signal using the difference between the thermal conductivity of the measurement component and the thermal conductivity of the carrier gas. As a gas chromatograph, various detectors are used according to the measurement component, its concentration, and measurement conditions. Among them, a TCD can be manufactured inexpensively, and on top of that, in principle, it can measure all components except the carrier gas.

[0004] Non-Patent Document 1: "Gas Chromatography Lecture for Beginners, No. 4, What is a TCD Detector?", [online], January 6, 2021, J Science Lab Co., Ltd., [searched on January 26, 2021], Internet <URL: https: / / http: / / j-sl.com / feature / gas / study04>

[0005] However, it is difficult for a TCD to maintain sensitivity. For example, since the sensitivity of a TCD is low, it is not suitable for measuring low-concentration components. Therefore, it is necessary to detect the deterioration of the TCD itself and suppress the decrease in the sensitivity of component measurement. Summary of the Invention

[0006] An object of the present invention is to detect the deterioration of a TCD and suppress the decrease in sensitivity.

[0007] An analysis device according to one aspect is characterized by having: an extraction unit that extracts a baseline part from chromatographic data acquired from an instrument; and a determination unit that determines the normality of the instrument based on the characteristics of the baseline part of the chromatographic data extracted by the extraction unit and the chromatographic data in a normal state.

[0008] An analysis method according to one aspect is characterized in that a computer executes processing for extracting a baseline portion from chromatographic data acquired from an instrument and determining the normality of the instrument based on characteristics of the baseline portion of the extracted chromatographic data and the chromatographic data when the instrument is normal.

[0009] A computer-readable recording medium according to one aspect records an analysis program, wherein the analysis program causes a computer to perform the following processing, namely, extracting a baseline portion from chromatographic data acquired from an instrument, and determining the normality of the instrument based on the characteristics of the baseline portion of the extracted chromatographic data and the chromatographic data under normal conditions.

[0010] Effects of the Invention

[0011] According to one embodiment, it is possible to detect degradation of the TCD while suppressing a decrease in sensitivity. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a functional block diagram showing the functional structure of the analysis device involved in the embodiment.

[0013] Figure 2 This is a diagram for explaining the processing of the analysis device.

[0014] Figure 3A This is a diagram for explaining the processing of the analysis device.

[0015] Figure 3B This is a diagram for explaining the processing of the analysis device.

[0016] Figure 4 This is a diagram for explaining the processing of the analysis device.

[0017] Figure 5 This is a flowchart for explaining the processing flow of the embodiment.

[0018] Figure 6 This is a flowchart for explaining the processing flow of the embodiment.

[0019] Fig. 7A It is a figure explaining an Example.

[0020] Figure 7B It is a figure explaining an Example.

[0021] Figure 7C It is a figure explaining an Example.

[0022] Figure 8 This is a diagram for explaining a hardware configuration example. DETAILED DESCRIPTION

[0023] Below, the embodiments of the analysis method, analysis program, and information processing device disclosed in the present application are described in detail based on the accompanying drawings. In addition, the present invention is not limited by the embodiments. In addition, the same symbols are marked on the same elements and repeated descriptions are appropriately omitted, and the various embodiments can be appropriately combined within the scope of non-contradiction.

[0024] [Analysis device]

[0025] Figure 1 is a functional block diagram showing the functional structure of the analysis device involved in the embodiment. Figure 1 As shown, the analysis device 10 is implemented by a general-purpose computer such as a personal computer or a workstation, and includes an input unit 11, an output unit 12, a communication control unit 13, a storage unit 14, and a control unit 15. In addition, the functional units of the analysis device 10 are not limited to the functional units shown in the figure, and other functional units may be included.

[0026] The input unit 11 is implemented by input devices such as a keyboard and a mouse, and inputs various instruction information such as starting a process to the control unit 15 in response to an input operation by an operator. The output unit 12 is implemented by a display device such as a liquid crystal display, a printing device such as a printer, etc. For example, the output unit 12 displays the result of the analysis process described later.

[0027] The communication control unit 13 is implemented by a NIC (Network Interface Card) or the like, and controls the communication between an external device and the control unit 15 via an electric communication line such as a LAN (Local Area Network) or the Internet. For example, the communication control unit 13 controls the communication between the control unit 15 and a management device that manages chromatographic data output by an instrument to be analyzed.

[0028] The storage unit 14 is implemented by a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 14 stores the processing program for operating the analysis device 10, the data used in the execution of the processing program, etc. in advance, or temporarily stores them each time the processing is performed. In addition, the storage unit 14 stores a model 14a used for the analysis process described later and the results of the analysis process. In addition, the storage unit 14 may be a structure that communicates with the control unit 15 via the communication control unit 13.

[0029] The control unit 15 is implemented by a CPU (Central Processing Unit) and the like, and executes a processing program stored in a memory. Figure 1As shown in the example, the control unit 15 functions as an acquisition unit 15a, an extraction unit 15b, a preprocessing unit 15c, a learning unit 15d, and a determination unit 15e. In addition, the above functional units can be installed in different hardware separately or partially. For example, the learning unit 15d can be installed in a learning device different from the determination unit 15e. In addition, the control unit 15 can have other functional units.

[0030] The acquisition unit 15a acquires chromatogram data from the instrument. For example, the acquisition unit 15a acquires chromatogram data outputted from the instrument to be analyzed via the input unit 11 or from the instrument to be analyzed or a management device that manages chromatogram data outputted from the instrument via the communication control unit 13 .

[0031] Specifically, the acquisition unit 15a acquires the chromatographic data of the TCD to be analyzed when it is normal, and supplies it to the learning unit 15d for processing via the extraction unit 15b and the preprocessing unit 15c described later. For example, the acquisition unit 15a supplies the chromatographic data of the first 100 or more pieces of data after the TCD is replaced with a normal one to the learning unit 15d.

[0032] The acquisition unit 15a acquires the chromatogram data of the TCD operation target and supplies it to the determination unit 15e via the extraction unit 15b and the preprocessing unit 15c described later. At this time, the acquisition unit 15a acquires the same number of chromatogram data as the chromatogram data supplied to the learning unit 15d and supplies it to the determination unit 15e.

[0033] Furthermore, the acquisition unit 15a may store the acquired chromatogram data in the storage unit 14. Alternatively, the acquisition unit 15a may not store the acquired chromatogram data in the storage unit 14 but may immediately transfer the acquired chromatogram data to the subsequent extraction unit 15b.

[0034] The extraction unit 15b extracts the baseline portion in the chromatographic data acquired from the instrument (ie, the baseline portion of the chromatographic data). Figure 2 , Figure 3A as well as Figure 3B This is a diagram for explaining the processing of the analysis device. Figure 2 The chromatographic data of the instrument are shown as an example. In addition to the peaks corresponding to the detected components, Figure 2 As shown by the dotted line in the figure, the chromatographic data of the instrument also includes a baseline portion whose change is extremely small compared with the peak portion.

[0035] When the instrument is normal, the baseline is roughly a straight line, and the noise change is so small that it cannot be visually confirmed. On the other hand, when the instrument has some abnormality, such as Figure 3A As shown, the baseline is tilted, as Figure 3BTherefore, the analysis device of this embodiment detects abnormality caused by instrument deterioration or the like by observing the baseline portion in the analysis process described later.

[0036] Back to Figure 1 As a preprocessing of the analysis process described later, the preprocessing unit 15c extracts the features of the baseline portion. For example, as the features of the baseline portion, the preprocessing unit 15c calculates the average value, maximum value, minimum value, difference between the maximum value and the minimum value, variance value, etc. of the baseline portion.

[0037] The learning unit 15d uses the chromatogram data acquired from the instrument as input data to learn the model 14a that outputs the evaluation value reflecting the normality of the instrument. Specifically, the learning unit 15d uses the characteristics of the baseline part of the chromatogram data when normal as teacher data and generates the model 14a that outputs the evaluation value reflecting the normality of the instrument through learning.

[0038] For example, the learning unit 15 d generates the model 14 a outputting the evaluation value of the instrument by learning using the features of the baseline extracted from the normal chromatogram data by the preprocessing unit 15 c as the teacher data. The learning unit 15 d stores the generated model 14 a in the storage unit 14 .

[0039] Furthermore, the learning unit 15d can learn the model 14a that outputs the evaluation value reflecting the normality of the instrument based on the normal chromatogram data or using the abnormal chromatogram data as the teacher data instead of the normal chromatogram data. Thus, the model 14a that outputs the evaluation value reflecting the normality of the instrument with high accuracy is generated.

[0040] Back to Figure 1 The determination unit 15e determines the normality of the instrument based on the characteristics of the baseline part of the chromatographic data extracted by the extraction unit 15b and the chromatographic data when normal. Specifically, the determination unit 15e uses the chromatographic data obtained from the instrument (i.e., the chromatographic data of the TCD of the analysis object during operation) as input data, and uses the model 14a learned by the learning unit 15d using the chromatographic data when normal as teacher data to determine the normality of the instrument. For example, when the evaluation value output by the model 14a is within the range of a predetermined threshold value, the determination unit 15e determines that the instrument is normal.

[0041] Furthermore, the determination unit 15e can use the chromatogram data acquired from the instrument as input data, and use the model 14a learned based on the chromatogram data of the normal time, or using the chromatogram data of the abnormal time as teacher data instead of the chromatogram data of the normal time, to determine the normality of the instrument. That is, the determination unit 15e can use the model 14a learned by the learning unit 15d based on the chromatogram data of the normal time, or using the chromatogram data of the abnormal time as teacher data instead of the chromatogram data of the normal time. Thus, the normality of the instrument can be determined with higher accuracy.

[0042] Here, as mentioned above, in the case of some abnormality in the instrument, such as Figure 3A As shown, the baseline is tilted, as Figure 3B As shown, the noise increases. Therefore, as a feature of the baseline portion, the determination unit e determines the normality of the instrument based on the slope of the baseline portion and the chromatogram data when normal. Alternatively, as a feature of the baseline portion, the determination unit 15e determines the normality of the instrument based on the size of the noise in the baseline portion and the chromatogram data when normal.

[0043] Figure 4 This is a diagram for explaining the processing of the analysis device. Figure 4 , an example is shown in which the determination unit 15e uses the average value and the difference between the maximum value and the minimum value, i.e., the size of the noise in the baseline portion, as the feature of the baseline portion, and applies the model 14a that outputs the evaluation value of the normality. Here, Figure 4 The horizontal axis of the curve graph is time, and the vertical axis is the evaluation value (unitless) when analyzed by AI.

[0044] In this case, the learning unit 15d generates the model 14a that outputs the evaluation value of the normality by learning using the average value of the baseline portion of the chromatogram data when the instrument is normal and the difference between the maximum value and the minimum value. Furthermore, the determination unit 15e inputs the data of the baseline portion of the chromatogram data when the instrument is operating into the learned model 14a to obtain the evaluation value of the instrument.

[0045] exist Figure 4 In the example shown, the evaluation value indicating normality decreases with the passage of time. When the output evaluation value deviates from the range of the preset threshold value, the determination unit 15e determines that the instrument is abnormal. In this way, the analysis device 10 can detect abnormalities caused by, for example, the deterioration of the instrument over time.

[0046] When the device is judged to be abnormal, the judging unit 15e can issue an alarm to the administrator of the device, etc. For example, when the number or frequency of device abnormalities exceeds a predetermined threshold, the judging unit 15e outputs an alarm or error display to the management device used by the administrator.

[0047] [Analysis method]

[0048] Figure 5 as well as Figure 6 This is a flowchart for explaining an example of the processing flow of the embodiment, that is, an analysis method. The analysis process (analysis method) of this embodiment includes a learning process and a determination process. First, Figure 5 The learning process shown is started when a user including an administrator, an operator, etc. instructs the start of the learning process, for example. The start of the learning process is not limited to the instruction of the user, and may be automatically started by a program, for example.

[0049] First, the acquisition unit 15a acquires chromatogram data when the TCD to be analyzed is normal (step S1). The extraction unit 15b extracts the baseline part in the acquired chromatogram data (step S2). The preprocessing unit 15c extracts the features of the baseline part.

[0050] Then, the learning unit 15d uses the characteristics of the baseline extracted from the normal chromatographic data by the preprocessing unit 15c as teaching data, and generates a model 14a that outputs the evaluation value of the instrument through learning (step S3). The learning unit 15d stores the generated model 14a in the storage unit 14. Thus, a series of learning processes are completed.

[0051] Next, Figure 6 The determination process shown is started when a user including an administrator, an operator, etc. instructs the start of the determination process, for example. The start of the determination process is not limited to the instruction of the user, and may be automatically started by a program, for example.

[0052] First, the acquisition unit 15a acquires the chromatogram data of the TCD during operation to be analyzed (step S11). The extraction unit 15b extracts the baseline portion from the acquired chromatogram data (step S12). The preprocessing unit 15c extracts the features of the baseline portion.

[0053] Then, the determination unit 15 e uses the characteristics of the baseline portion extracted from the acquired chromatogram data by the preprocessing unit 15 c as input data and determines the normality of the device using the learned model 14 a (step S13 ).

[0054] In addition, the determination unit 15e outputs the determination result to the output unit 12 (step S14). For example, when the determination unit 15e determines that the device to be analyzed is abnormal, it outputs an error message, an alarm, etc. to the device management device, etc. Thus, a series of determination processes are terminated.

[0055] [Effect]

[0056] As described above, the analysis device 10 involved in the present embodiment has an extraction unit 15b and a determination unit 15e. The extraction unit 15b extracts the baseline portion of the chromatographic data obtained from the instrument. The determination unit 15e determines the normality of the instrument based on the characteristics of the baseline portion of the chromatographic data extracted by the extraction unit 15b and the chromatographic data when normal. As a result, the analysis device 10 can detect the degradation of the instrument during operation in real time and respond to it. Therefore, it is possible to suppress the situation where the TCD is replaced at an excessively high frequency before a fault occurs, or the situation where a fault occurs and causes a part of the factory's functions to stop and recover. In this way, according to the analysis device 10 involved in the present embodiment, the sensitivity reduction of the TCD in operation can be easily suppressed without adding a new sensor, etc., and the sensitivity can be maintained.

[0057] In addition, as a feature of the baseline portion, the determination unit 15e determines the normality of the instrument based on the slope of the baseline portion and the chromatogram data when normal. Alternatively, as a feature of the baseline portion, the determination unit 15e determines the normality of the instrument based on the size of the noise in the baseline portion and the chromatogram data when normal. Thus, the analysis device 10 can determine the normality of the instrument with high accuracy.

[0058] The determination unit 15e uses the chromatogram data acquired from the device as input data and determines the normality of the device using the model 14a learned using the chromatogram data in normal state as teacher data. Thus, the analyzer 10 can determine the normality of the device with high accuracy.

[0059] The determination unit 15e uses the chromatogram data acquired from the instrument as input data and determines the normality of the instrument using the model 14a learned based on the normal chromatogram data or using the abnormal chromatogram data as teacher data instead of the normal chromatogram data. Thus, the analyzer 10 can determine the normality of the instrument with higher accuracy.

[0060] The learning unit 15d uses the chromatogram data acquired from the instrument as input data to learn the model 14a that outputs the evaluation value reflecting the normality of the instrument. Thus, the analysis device 10 can generate the model 14a that accurately outputs the evaluation value reflecting the normality of the instrument and accurately determine the normality of the instrument.

[0061] [Example]

[0062] Figure 7A to Figure 7C is a diagram for explaining an embodiment. In this embodiment, junk data is intentionally mixed into the TCD to cause an abnormality in the baseline portion. Specifically, Fig. 7AThe device with the piping diagram shown in the figure samples air, thereby using the oxygen contained in the air to oxidize the TCD filaments and cause abnormalities in the baseline. Specifically, the air is sampled at a period of 300 seconds to generate peaks, and data is collected for 24 hours. In this case, the obtained Figure 3A Chromatographic data of abnormal state shown.

[0063] In addition, Fig. 7A In the example shown, the chromatographic column is processed in such a way that the filler of fine particles filled in the chromatographic column installed near the front side of the TCD leaks out. Figure 3B Chromatographic data of abnormal state shown.

[0064] Furthermore, the chromatographic data of the normal state and the chromatographic data of the abnormal state are preprocessed respectively. Specifically, the chromatographic data of the 300-second period is taken as one data and 288 data are obtained within 24 hours. Furthermore, Figure 7B As shown, a baseline portion of a predetermined range of about 1 to 10 seconds (5 seconds in this embodiment) is extracted from each data, and a variance value is calculated. In addition, the noise of the baseline portion when set to about 1 to 10 seconds is extremely small, thereby forming long-term data to prevent the inclusion of elements other than noise.

[0065] The results, such as Figure 7C As shown in FIG. 1 , it can be seen that the evaluation values ​​of the chromatographic data in the abnormal state deviate greatly from those in the normal state many times. For example, Figure 7C Points with evaluation values ​​below -0.5 depict Figure 3B The abnormal state shown, Figure 7C The evaluation value of the multiple points near 0 is drawn Figure 3B As shown in the normal state, it was confirmed that the analyzer 10 can detect an abnormality of the device by the instability of the noise in the baseline portion.

[0066] [Other embodiments]

[0067] Thus, although the embodiment of the present invention has been described so far, the present invention can be implemented in various different modes other than the above-described embodiment.

[0068] For example, the characteristics of the baseline part used in the above embodiment are just examples and can be changed arbitrarily. In addition, various algorithms such as neural network, deep learning, support vector machine, etc. can be used for the model 14a generated by the learning unit 15d.

[0069] [Analysis program]

[0070] The information including the processing sequence, control sequence, specific names, various data, and parameters described above and shown in the drawings can be changed arbitrarily except for cases where special description is given.

[0071] In addition, the structural elements of each device shown in the figure are functional concepts and do not need to be physically configured as shown in the figure. In other words, the specific way of dispersing and integrating each device is not limited to the structure shown in the figure. In other words, all or part of them can be formed into a structure that is functionally or physically dispersed / integrated in any unit according to various loads, usage conditions, etc.

[0072] Furthermore, all or any part of each processing function performed by the device can be implemented by using a CPU and a program executed by the CPU, or as hardware based on wired logic. For example, an analysis program can be executed on a cloud network. The analysis program can cause a computer to perform the following processing, that is, extracting a baseline portion from chromatographic data acquired by the instrument, and determining the normality of the instrument based on the characteristics of the baseline portion of the extracted chromatographic data and the chromatographic data when normal.

[0073] [hardware]

[0074] Next, a hardware configuration example of the analysis device 10 will be described. Figure 8 This is a diagram for explaining an example of a hardware configuration. Figure 8 As shown, the analysis device 10 includes a communication device 10a, a HDD (Hard Disk Drive) 10b, a memory 10c, and a processor 10d. Figure 8 The various parts shown are connected to each other by a bus or the like.

[0075] The communication device 10a is a network interface card, etc., and performs communication with other servers. Figure 1 The program and DB that execute the functions shown are stored.

[0076] The processor 10d reads and executes data from the HDD 10b or the like. Figure 1 The same processing program is developed in the memory 10c for each processing unit as shown in FIG. Figure 1 The process of each function described in the above etc. executes the action. For example, this process executes the same function as each processing unit of the analysis device 10. Specifically, the processor 10d reads a program having the same function as the acquisition unit 15a, the extraction unit 15b, the preprocessing unit 15c, the learning unit 15d, the determination unit 15e, etc. from the HDD 10b, etc. Then, the processor 10d executes the following process, that is, executing the same processing as the acquisition unit 15a, the extraction unit 15b, the preprocessing unit 15c, the learning unit 15d, the determination unit 15e, etc.

[0077] In this way, the analysis device 10 performs an action as an information processing device that reads and executes the analysis program to execute the analysis method. That is, the analysis device 10 can be an information processing device such as a PC installed with the analysis program. In addition, the analysis device 10 can also use a medium reading device to read the above program from the recording medium and execute the read program, thereby realizing the same function as the above embodiment. In addition, the program described in other embodiments is not limited to being executed by the analysis device 10. For example, when other computers or servers execute the program, or when they execute the program in collaboration, the present invention can also be applied in the same way.

[0078] The program can be distributed via a network such as the Internet. In addition, the program can be recorded on a computer-readable recording medium such as a hard disk, a floppy disk (FD), a CD-ROM, an MO (Magneto-Optical disk), or a DVD (Digital Versatile Disc), and can be read from the recording medium by a computer and executed. In addition, as shown in Figure 8 As described in the above and other aspects, a recording medium recording the program is also one aspect of the present disclosure.

Claims

1. An analysis device, characterized in that: The analysis device comprises: an extraction section that extracts a baseline portion from chromatographic data acquired from the instrument; and a determination unit that determines the normality of the instrument based on the characteristics of the baseline portion of the chromatographic data extracted by the extraction unit and the chromatographic data when normal; The instrument is a thermal conductivity detector that uses the difference between the thermal conductivity of the measuring component and the thermal conductivity of the carrier gas to extract the unbalanced voltage generated by the bridge circuit as a concentration signal.

2. The analysis device according to claim 1, characterized in that The determination unit determines the normality of the device based on a slope of the baseline portion and the chromatogram data at the normal time as the feature of the baseline portion.

3. The analysis device according to claim 1, characterized in that The determination unit determines the normality of the device based on the magnitude of noise in the baseline portion and the chromatogram data in the normal state as the feature of the baseline portion.

4. The analysis device according to claim 1, characterized in that The determination unit determines the normality of the device using the chromatogram data acquired from the device as input data and a model learned using the chromatogram data in a normal state as teacher data.

5. The analysis device according to claim 4, characterized in that The determination unit determines the normality of the device using the chromatogram data acquired from the device as input data and the model learned based on the normal chromatogram data or using the abnormal chromatogram data as teacher data instead of the normal chromatogram data.

6. The analysis device according to claim 4, characterized in that The analysis device further includes a learning unit that uses the chromatogram data acquired from the instrument as input data to learn the model that outputs an evaluation value reflecting the normality of the instrument.

7. A parsing method, characterized in that: The computer performs the following processing, namely, Extract the baseline part of the chromatographic data obtained from the instrument. The normality of the instrument is determined based on the characteristics of the baseline portion of the extracted chromatographic data and the chromatographic data when the instrument is normal. The instrument is a thermal conductivity detector that uses the difference between the thermal conductivity of the measuring component and the thermal conductivity of the carrier gas to extract the unbalanced voltage generated by the bridge circuit as a concentration signal.

8. A computer-readable recording medium having an analysis program recorded thereon, characterized in that: The analysis program causes the computer to execute the following processing, namely, Extract the baseline part of the chromatographic data obtained from the instrument. determining the normality of the instrument based on the characteristics of the baseline portion of the extracted chromatographic data and the chromatographic data when the instrument is normal; The instrument is a thermal conductivity detector that uses the difference between the thermal conductivity of the measuring component and the thermal conductivity of the carrier gas to extract the unbalanced voltage generated by the bridge circuit as a concentration signal.

Images