TWI855253B - Electromagnetic wave signal analysis device and electromagnetic wave signal analysis program - Google Patents

Abstract

[Topic] By making active use of the fact that electromagnetic waves are absorbed by water vapor, it is possible to easily detect the inherent characteristics of liquid samples. [Solution] It comprises: a spectrum acquisition unit (11) for acquiring a spectrum having characteristic values relative to frequency generated by an electromagnetic wave signal acquired by a spectrometer (20); and a water vapor fitting processing unit (12) for fitting the spectrum of a frequency at which the absorption of electromagnetic waves by water vapor becomes greater to a waveform of a single fitting function or a plurality of fitting functions. and a characteristic analysis unit (14) which uses at least two values determined for the properties of the fitting function used in the fitting to analyze the characteristics of the liquid sample, and processes the spectrum of the frequency at which the absorption of electromagnetic waves caused by water vapor generated from the surface of the liquid sample becomes large to analyze the characteristics of the liquid sample.

Description

Electromagnetic wave signal analysis device and electromagnetic wave signal analysis program

The present invention relates to an electromagnetic wave signal analysis device and an electromagnetic wave signal analysis program, and in particular to a device for analyzing the characteristics of electromagnetic waves after passing through a liquid sample arranged on an optical path of a spectrometer and a program used in the device.

In the prior art, there is provided a spectrometer that uses electromagnetic waves to measure the properties of a substance. In the spectrometer, electromagnetic waves are made to pass through or reflect a sample, and the physical or chemical properties of the sample are measured based on the changes in the electromagnetic waves caused by the interaction between the electromagnetic waves and the sample. The spectrum of the sample observed by this spectrometer has a spectrum structure inherent to the sample. In particular, in spectrometry using megahertz waves, which are one type of electromagnetic waves, intermolecular interactions caused by hydrogen bonding, etc. are observed.

However, since the spectra observed by spectrometers overlap, it is difficult to extract the peak of a specific frequency. Therefore, it is difficult to know "where in the spectrum the sample characteristics are expressed" or "in which waveform the sample characteristics are expressed", and there is a problem that it is extremely difficult to find the characteristics.

In order to solve such a problem, a megahertz wave analysis device is proposed, which allows megahertz waves to pass through or reflect a liquid sample to be measured arranged on an optical path, and analyzes the megahertz wave signal obtained by detecting the megahertz wave acting on the liquid sample in this way by a spectrometer, so as to visualize the characteristics corresponding to the characteristics of the liquid sample easily (for example, refer to Patent Document 1). In the analysis device described in Patent Document 1, a spectrum obtained from the megahertz wave signal is fitted with a synthetic waveform of a complex fitting function, and at least two values specified for the properties of the complex fitting function used in the fitting are used as parameters to generate a graph.

In addition, since electromagnetic waves are also absorbed by water vapor in the atmosphere, the acquired spectrum may contain characteristics of water vapor. Therefore, in the analysis device described in Patent Document 1, before performing the fitting process by the fitting function, the extreme value at the frequency where the absorption of megahertz waves due to water vapor becomes large is first removed. That is, the spectrum is set to a state that is difficult to be affected by the absorption of the spectrum due to water vapor by removing the "extreme value at the frequency where the absorption of megahertz waves due to water vapor other than the liquid sample becomes large" from the absorbance data relative to each frequency in the spectrum, and then fitting is performed. [Prior art literature] [Patent literature]

[Patent Document 1] WO2018/110481 [Patent Document 2] Japanese Patent Publication No. 2010-164511 [Patent Document 3] Japanese Patent Publication No. 2008-46574 [Patent Document 4] Japanese Patent Publication No. 2007-108151 [Patent Document 5] Japanese Patent Publication No. 2000-74827

[The problem that the invention is trying to solve]

As mentioned above, it is well known that electromagnetic waves are absorbed by water vapor. For this reason, when performing spectroscopic measurement of electromagnetic waves, it is generally configured to avoid the influence of spectrum absorption caused by water vapor contained in the atmosphere other than the sample (for example, the above-mentioned patent document 1, and further, refer to patent documents 2 to 5). In contrast, the inventors of this application have found that by actively utilizing the fact that electromagnetic waves are absorbed by water vapor, it is possible to analyze the inherent characteristics of the liquid sample from the spectrum measured by the spectrometer.

That is, the purpose of the present invention is to analyze the electromagnetic wave signal detected by the spectrometer and make active use of the fact that the electromagnetic wave is absorbed by water vapor, thereby making it possible to easily detect the inherent characteristics of liquid samples. [Means for solving the problem]

In order to solve the above-mentioned problems, the electromagnetic wave signal analysis device of the present invention analyzes the electromagnetic wave signal obtained by the spectrometer, and the spectrometer detects the electromagnetic waves transmitted or reflected by the film-like liquid sample generated in the space where the liquid is ejected from the nozzle. Specifically, the electromagnetic wave signal analysis device is configured to fit the spectrum of the water vapor absorption frequency, which is "the frequency at which the absorption of electromagnetic waves due to water vapor becomes larger", among "the spectrum having characteristic values relative to the frequency generated by the electromagnetic wave signal", with a waveform of a fitting function or a composite waveform of multiple fitting functions, and analyze the characteristics of the liquid sample using at least two values determined for the properties of the fitting function used in the fitting. [Effect of the invention]

As described above, if a film-like liquid sample is generated in space by spraying liquid from a nozzle, water vapor is generated from the surface of the liquid sample, and electromagnetic waves also pass through this water vapor. Since this water vapor contains the properties of the liquid sample, the characteristics of the liquid sample can be analyzed by processing the spectrum of the frequency at which the absorption of electromagnetic waves by this water vapor becomes larger. In addition, the spectrum of water vapor is easy to read and easier to analyze because the peak is clear and sharp. According to the present invention constructed as described above, the spectrum caused by water vapor generated from the liquid sample is approximated by a waveform of a single fitting function or a composite waveform of multiple fitting functions in a form that inherits the characteristics of the liquid sample, and the characteristics of the liquid sample are analyzed based on the value of the fitting function used in the approximation. Thus, according to the present invention, by analyzing the electromagnetic wave signal detected by the spectrometer and actively utilizing the fact that the electromagnetic wave is absorbed by water vapor, it is possible to easily detect the inherent characteristics of the liquid sample.

Hereinafter, one embodiment of the present invention will be described with reference to the drawings. FIG1 is a block diagram showing an example of the functional configuration of an electromagnetic wave signal analysis device according to the present embodiment. The electromagnetic wave signal analysis device 10 of the present embodiment is for analyzing "the electromagnetic wave signal obtained by the spectrometer 20 for the liquid sample", and as its functional configuration, it has a spectrum acquisition unit 11, a water vapor fitting processing unit 12, a liquid fitting processing unit 13, and a characteristic analysis unit 14.

The electromagnetic wave that is analyzed by the electromagnetic wave signal analysis device 10 of the present embodiment is, for example, megahertz waves, infrared rays, visible rays, ultraviolet rays, etc. Any electromagnetic wave can be analyzed as long as it contains a frequency band that is absorbed by water vapor. In addition, especially in the case of megahertz waves, since the molecular interactions generated in the liquid sample in response to the megahertz waves are complex processes, the degree of mutual overlap of the spectra observed by the spectrometer is strong, and it is difficult to extract the peak of a specific frequency compared to infrared spectroscopy or photoelectron spectroscopy. The electromagnetic wave signal analysis device 10 of the present embodiment is also capable of analyzing such megahertz wave signals.

Each of the functional blocks 11 to 14 described above can be constructed by any one of hardware, digital signal processor (DSP) or software. For example, when constructed by software, each of the functional blocks 11 to 14 described above is actually constructed by having a CPU, RAM, ROM, etc. of a computer, and is implemented by operating an electromagnetic wave signal analysis program stored in a recording medium such as RAM or ROM, a hard disk or a semiconductor memory.

The spectrum acquisition unit 11 acquires a spectrum showing a characteristic value relative to frequency generated by the electromagnetic wave signal detected by the spectrometer 20. In the present embodiment, as one example of the characteristic value, a spectrum showing absorbance is acquired. The spectrum acquisition unit 11 may be configured to acquire the spectrum detected by the spectrometer 20, or may be configured to acquire the spectrum by generating a spectrum itself based on the electromagnetic wave signal generated by the spectrometer 20.

The spectrometer 20 allows electromagnetic waves to pass through or reflect a liquid sample to be measured arranged on an optical path, and detects the electromagnetic waves that act on the liquid sample in this way. In this embodiment, various known spectrometers can be used as the spectrometer 20. However, the spectrometer 20 used in this embodiment has a specific nozzle, and generates a film-like liquid sample in space (on the path of the electromagnetic wave) by ejecting liquid from the nozzle.

FIG. 2 is a diagram showing an example of the structure of the liquid film generating device used in the present embodiment. As shown in FIG. 2(a), the liquid film generating device is composed of a container 1, a liquid film cartridge 2, a hose pump 3, a forward pipe 4, and a return pipe 5. The container 1 is provided with a liquid recovery tank 1a. The liquid film cartridge 2 is provided with a nozzle 21 (refer to FIG. 2(b)) for ejecting liquid and generating a liquid film (hereinafter referred to as a sample liquid film 100).

The hose pump 3 sucks up the liquid of the measurement object from the recovery tank 1a through the return pipe 5, pressurizes the sucked liquid, and guides it to the liquid film cartridge 2 through the forward pipe 4. The liquid film cartridge 2 generates a sample liquid film 100 with a flat surface in space by ejecting the liquid guided from the recovery tank 1a by the hose pump 3 from the nozzle 21. Holes are provided on the sides of the container 1 and the liquid film cartridge 2 near the height at which the sample liquid film 100 is formed, and electromagnetic waves pass through the sample liquid film 100 through the holes.

The recovery tank 1a is used to recover and store the liquid dripping from the liquid film cartridge 2. The liquid stored in the recovery tank 1a is sucked up again by the hose pump 3, and is pressurized and ejected from the nozzle 21 of the liquid film cartridge 2. In this way, the liquid in the recovery tank 1a is circulated, and the sample liquid film 100 is generated through the nozzle 21 during the circulation process.

FIG. 2( b ) is a diagram for explaining the sample liquid film 100 produced by the nozzle 21. Here, the three-dimensional coordinate axes defining the space are indicated by x-y-z. The center axis 21a of the nozzle 21 is set to face the y-axis direction. At the front end of the nozzle 21, a slit-shaped opening 21b is provided which is orthogonal to the center axis 21a, and the slit is set to be parallel to the x-axis.

As shown in FIG. 2( b ), the liquid ejected from the opening 21 b provided at the front end of the nozzle 21 sequentially forms a plurality of mutually orthogonal liquid film surfaces 101 to 103. The first liquid film surface 101 is formed by the surface tension of the liquid between two strip-shaped fluid columns 111 flowing in the z-y plane by the liquid ejected from the opening 21 b of the nozzle 21. That is, the two strip-shaped fluid columns 111 draw a smooth arc while colliding with each other at the fluid column gathering point 121, and the first liquid film surface 101 is formed by the surface tension of the liquid in the space from the opening 21 b of the nozzle 21 to the fluid column gathering point 121. Therefore, the first liquid film surface 101 is a surface perpendicular to the x-axis and parallel to the zy plane.

The two strip-shaped fluid columns 111 that collided with each other at the fluid column collection point 121 change their angle by 90 degrees and become two strip-shaped fluid columns 112 that flow in the x-y plane, and collide with each other at the next fluid column collection point 122 while drawing a smooth arc. Thus, in the space from the first fluid column collection point 121 to the second fluid column collection point 122, the second liquid film surface 102 is formed by the surface tension of the liquid. Therefore, the second liquid film surface 102 is perpendicular to the first liquid film surface 101, and becomes a surface perpendicular to the x-axis and parallel to the x-y plane.

The third liquid film surface 103 is also the same as the first liquid film surface 101 and the second liquid film surface 102, and is formed by the surface tension of the liquid in the space from the second fluid column gathering point 122 to the third fluid column gathering point 123. The third liquid film surface 103 is perpendicular to the second liquid film surface 102, and is a surface perpendicular to the x-axis and parallel to the z-y plane. At the spectroscopic device 20, the electromagnetic wave is made to pass through or reflect the first liquid film surface 101 among the liquid film surfaces 101 to 103 formed in this way. This first liquid film surface 101 is equivalent to the "liquid sample" in the scope of the patent application.

In addition, the structure of the liquid film generating device and the nozzle 21 shown here is only one example and is not limited thereto. As long as it is a device and a nozzle that can generate a liquid film in space by ejecting a liquid as shown in FIG. 2(b), no matter what structure it has, it can be used in this embodiment.

FIG3 is a diagram showing the appearance of water vapor around the liquid sample generated by the nozzle 21. If a sample liquid film 100 (liquid sample) is generated in space by spraying liquid from the nozzle 21, water vapor 110 is generated from the surface of the sample liquid film 100. The electromagnetic wave advancing on the optical path of the spectrometer 20 also passes through the water vapor 110 and passes through or reflects the sample liquid film 100.

FIG4 is a diagram showing an example of a spectrum obtained by the spectrum acquisition unit 11. In FIG4, the vertical axis represents absorbance, the horizontal axis represents frequency, and a spectrum in the range of 0 to 2 THz is shown. This spectrum has different waveforms in each of the liquid samples with different properties, but it is difficult to know "where in the waveform the characteristics of the liquid sample are expressed" or "in which waveform the characteristics of the liquid sample are expressed". The electromagnetic wave signal analysis device 10 of this embodiment is a device that analyzes such a spectrum to easily indicate the characteristics corresponding to the characteristics of the liquid sample.

Returning to FIG. 1 , the configuration of the electromagnetic wave signal analysis device 10 will be described. The water vapor fitting processing unit 12 performs a process of fitting a frequency spectrum where the absorption of electromagnetic waves due to water vapor increases in the spectrum acquired by the spectrum acquisition unit 11 with a waveform of a single fitting function or a synthesized waveform of multiple fitting functions. Hereinafter, the frequency where the absorption of electromagnetic waves due to water vapor increases is referred to as a "water vapor absorption frequency", and the spectrum of the water vapor absorption frequency generated by the water vapor fitting processing unit 12 is referred to as a "water vapor spectrum".

The frequency at which the absorption of electromagnetic waves by water vapor increases can be determined by using data provided by NICT (National Research and Development Corporation, Information and Communications Technology), for example. NICT publishes data on the radio wave attenuation rate of air (including water vapor) for electromagnetic wave communication. Alternatively, the HITRANonline database published on the website can be used. By using this data, the frequency at which the absorption of electromagnetic waves by water vapor increases can be determined. In FIG4, several frequencies at which the peaks of absorbance increase extremely are observed, but these are the absorption frequencies of water vapor.

The water vapor fitting processing section 12 performs fitting using a complex normal distribution function (Gaussian function) having at least one different center frequency, amplitude, and width as one example of a fitting function. That is, the water vapor fitting processing section 12 fits the spectrum of the complex water vapor absorption frequency in the spectrum obtained by the spectrum acquisition section 11 to a waveform of a complex normal distribution function having at least one different center frequency, amplitude, and width (e.g., 1/e width).

The water vapor fitting processing section 12, for each of the plurality of water vapor absorption frequencies, is basically based on the premise that "one spectrum (peak waveform) can be approximated by the waveform of one normal distribution function", and a normal distribution function that minimizes the residual between "the value of the absorbance at each frequency" and "the value of the waveform of the normal distribution function corresponding to it at each frequency" is calculated for each of the plurality of water vapor absorption frequencies by optimization calculation using the center frequency, amplitude and width as variables.

The water vapor spectrum generated by the optimization calculation of the water vapor fitting processing unit 12 can be expressed by the following (Formula 1). In this (Formula 1), i is the identification code of the complex water vapor absorption frequency (i=1, 2, ..., m), λ _i is the water vapor absorption frequency, I _i is the absorbance at the water vapor absorption frequency, λ _0i is the center frequency of the normal distribution function, a _0i is the amplitude of the normal distribution function, and a _1i is the width of the normal distribution function.

FIG5 is a diagram showing an example of a water vapor spectrum generated by the water vapor fitting processing section 12. As shown in FIG5, the water vapor spectrum generated by the water vapor fitting processing section 12 is a spectrum in which the spectrum around the multiple water vapor absorption frequencies where the peak value of the absorbance in the entire spectrum shown in FIG4 increases extremely is approximated by the waveform of the multiple normal distribution function. In addition, the water vapor spectrum shown in FIG5 is a waveform showing only the peak component of the absorbance at the multiple water vapor absorption frequencies.

In addition, the absorbance of the water vapor spectrum shown in FIG5 shows a difference value of the intensity relative to the reference value at each water vapor absorption frequency. The so-called reference value is a value of the absorbance at the fundamental part (base) of the waveform that becomes a peak. In this embodiment, in the spectrum generated by the liquid fitting processing unit 13 described below, each absorbance at a plurality of water vapor absorption frequencies is used as a reference value at the plurality of water vapor absorption frequencies.

Here, although the example of "fitting a waveform of a normal distribution function to a spectrum of a water vapor absorption frequency (a peak waveform)" is described, the invention is not limited to this. For example, "fitting a composite waveform of multiple normal distribution functions to a spectrum of a water vapor absorption frequency" may be configured. In the case where the accuracy of the fitting is improved by "approximating by a composite waveform of multiple normal distribution functions" rather than "approximating by a waveform of a single normal distribution function", the fitting may be performed using a composite waveform.

The liquid fitting processing unit 13 performs a process of "fitting the spectrum other than the peak component of the water vapor absorption frequency in the spectrum acquired by the spectrum acquisition unit 11 with a synthetic waveform of a complex fitting function." Hereinafter, the spectrum generated by the liquid fitting processing unit 13 is referred to as a "liquid spectrum."

Here, the phrase "except for the peak component of the water vapor absorption frequency" does not mean deleting the "absorbance data of the water vapor absorption frequency", but means "separating only the peak component of the water vapor absorption frequency from the spectrum such as FIG. 4 obtained by the spectrum acquisition unit 11". The phrase "separating only the peak component of the water vapor absorption frequency" is equivalent to performing fitting with respect to the water vapor absorption frequency as the absorbance as the above-mentioned reference value when performing fitting by the liquid fitting processing unit 13. This point is different from Patent Document 1, which states that "fitting is performed after removing the absorbance data at the water vapor absorption frequency". The separated peak components of the water vapor absorption frequency are fitted by the water vapor fitting processing unit 12 as described above.

In this embodiment, the liquid fitting processing unit 13 performs fitting using a polynomial function including complex terms as a complex fitting function. That is, the liquid fitting processing unit 13 fits the spectrum obtained by the spectrum acquisition unit 11 after separating the peak component with respect to the water vapor absorption frequency, with a composite waveform of waveforms specified by terms of each order in an n-order polynomial (n>1). Here, terms of each order in the n-order polynomial are equivalent to a "complex fitting function".

The liquid fitting processing unit 13 calculates each coefficient of the complex term in the polynomial function by optimization calculation based on the premise that "the liquid spectrum can be approximated by an n-order polynomial function" in a manner that minimizes the residual between "the value of the absorbance at each frequency" and "the value of the corresponding synthetic waveform at each frequency (the value of the polynomial function)". The liquid spectrum generated by the optimization calculation of the liquid fitting processing unit 13 can be expressed by the following (Formula 2). In this (Formula 2), x is a variable of a 1-variable polynomial (a polynomial with only one indeterminate variable), and here, it represents frequency. j is the degree of the n-th degree polynomial (j=1, 2, ..., n), and _bj represents the coefficient of each term.

Fig. 6 is a diagram showing an example of a liquid spectrum generated by the liquid fitting processing unit 13. As shown in Fig. 6, the liquid spectrum generated by the liquid fitting processing unit 13 is a spectrum that does not include a peak component of absorbance at the water vapor absorption frequency in the entire spectrum shown in Fig. 4, and is approximated by a synthetic waveform of a polynomial function.

If the water vapor spectrum shown in FIG. 5 is synthesized with the liquid spectrum shown in FIG. 6, a spectrum that approximates the entire spectrum shown in FIG. 4 (hereinafter referred to as a fully fitted spectrum) is obtained. FIG. 7 is a diagram showing one example of this fully fitted spectrum. In addition, in this embodiment, the spectrum obtained by the spectrum acquisition unit 11 as shown in FIG. 4 is fitted by synthesizing the water vapor spectrum and the liquid spectrum as shown in the following formula. Fully fitted spectrum = water vapor spectrum + liquid spectrum

In addition, for the sake of convenience, although the water vapor simulation processing section 12 and the liquid simulation processing section 13 are illustrated as different functional blocks, it is possible to calculate the water vapor spectrum and the liquid spectrum at the same time and obtain the entire simulation spectrum.

The characteristic analysis unit 14 analyzes the characteristics of the liquid sample using at least two values including a value for the property of the complex fitting function used in the fitting of the water vapor fitting processing unit 12 and a value for the property of the complex fitting function used in the fitting of the liquid fitting processing unit 13. The value for the property of the complex fitting function used in the fitting of the water vapor fitting processing unit 12 is the center frequency, amplitude and width of the complex normal distribution function (the values of λ _0i , a _0i , a _1i in the above (Formula 1)). The values for determining the properties of the complex fitting function used in the fitting of the liquid fitting processing unit 13 are the coefficients of the polynomial function (the values of _bj in the above (Formula 2)). The characteristic analysis unit 14 analyzes the characteristics of the liquid sample using at least two of these values λ _0i , a _0i , a _1i , and _bj .

Here, the characteristic analysis unit 14 may be configured to analyze the characteristics of the liquid sample using at least two of the center frequency λ _0i , amplitude a _{0i ,} and width a _1i of the complex normal distribution function used in the fitting of the water vapor fitting processing unit 12. Furthermore, the characteristic analysis unit 14 may be configured to analyze the characteristics of the liquid sample using at least two of the coefficients b _j of the polynomial function used in the fitting of the liquid fitting processing unit 13. Furthermore, the characteristic analysis section 14 may also be configured to analyze the characteristics of the liquid sample using at least one of the center frequency λ _0i , amplitude a _0i , and width a _1i of the complex normal distribution function used in the fitting of the water vapor fitting processing section 12, and at least one of the coefficients b _j of the polynomial function used in the fitting of the liquid fitting processing section 13.

The content of the analysis performed by the characteristic analysis unit 14 is, for example, a specific statistical processing or function processing performed using the above-mentioned at least two values. The at least two values used in this statistical processing or function processing are values set for the properties of the complex fitting function used in the generation of the water vapor spectrum and the properties of the complex fitting function used in the generation of the liquid spectrum, and they are all values that reflect the characteristics of the liquid sample. Therefore, the value of the result after the specific statistical processing or function processing is performed using these values becomes a unique value that reflects the characteristics of the liquid sample.

Therefore, it is possible to use the statistical value or function value calculated in this way to identify the characteristics of the liquid sample or to classify liquid samples having the same or similar characteristics. In addition, it is also possible to perform analysis such as "using the statistical value or function value calculated based on "spectrums measured multiple times by the spectrometer 20 for the same liquid sample with time intervals" to detect changes in the characteristics of the liquid sample."

Furthermore, the characteristic analysis unit 14 is also capable of performing general analysis such as "using a combination of "at least two of the above values calculated based on a liquid sample with known characteristics" and "data representing the characteristics of the liquid sample" as teacher data, and generating a completed learning model (prediction model) by machine learning using a plurality of teacher data, and then inputting the above at least two values calculated based on a liquid sample with unknown characteristics into the completed learning model, thereby outputting data representing the characteristics of the liquid sample from the prediction model." For example, the prediction model that has been subjected to machine learning can be used to analyze characteristics such as "the concentration of a specific solvent or solvent in a solution", "the type of solvent or solvent contained in a solution", "whether there is foreign matter mixed into the reference liquid", etc. In addition, the characteristics of the liquid sample shown here as the analysis object are just one example and are not limited to this.

Furthermore, the characteristic analysis unit 14 may be configured to generate a specific graph using at least two of the above values as parameters. For example, the characteristic analysis unit 14 may be configured to generate a graph showing the relationship between at least two of the center frequency λ _0i , the amplitude a _0i , and the width a _1i of the complex normal distribution function used in the generation of the water vapor spectrum. Furthermore, the characteristic analysis unit 14 may be configured to calculate the area A i of a specific region (a waveform region having an amplitude greater than or equal to 1/e of the width) of the normal distribution waveform from the amplitude a _0j and the width a _1j for each of the complex normal distribution functions, and generate a graph showing the relationship between the center frequency λ _0i and _{the area A i .}

Furthermore, the characteristic analysis unit 14 may also be configured to "take one of the multiple normal distribution functions as a reference, and for each of the multiple center frequencies λ _0y , calculate the ratio R xy (equivalent to a value representing the balance between the peak values of the amplitude a 0i or the width a _1i that is characteristic of the waveform of the peak component at each water vapor absorption frequency) between "the amplitude a _0x or the width a _1x (x is one of 1 to i) of the normal distribution function used as the reference" and "the amplitude a _0y or the width a 1y (y is one of ₁ to i, x≠ _y ) of other normal distribution functions", and generate a _graph showing the relationship between the center frequency λ _0y and the ratio R _xy .

Hereinafter, an example of generating a graph using at least two of the center frequency λ _0i , the amplitude a _0i , and the width a _1i of the complex normal distribution function used in the fitting in the water vapor fitting processing unit 12 will be described.

Fig. 8 is a diagram showing an example of a graph generated by the characteristic analysis unit 14. The graph shown in Fig. 8 is a radar diagram showing five central frequencies λ _0i (representing values roughly approximating the central frequency of the complex water vapor absorption frequency) of the normal distribution function used in the generation of the water vapor spectrum as axes, and amplitude a _0i , width a _1i or area _Ai as values of each axis.

The radar diagram of FIG8 shows an example of a radar diagram generated when "a megahertz wave is used as an example of an electromagnetic wave, and five central frequencies λ _0i of 0.75 THz, 0.99 THz, 1.10 THz, 1.16 THz, and 1.41 THz are used as axes." The five central frequencies λ _0i used here correspond to "a water vapor absorption frequency with a large absorption of a megahertz wave, that is, a water vapor absorption frequency with a large peak value."

In addition, the frequency value (center frequency λ _0i , amplitude a _0i , width a _1i ) associated with a plurality of water vapor absorption frequencies to be used in the graphing can be arbitrarily defined. For example, it is possible to generate a graph using the value associated with the water vapor absorption frequency whose absorbance in the water vapor spectrum generated as shown in FIG. 5 is equal to or greater than a specific value. In addition, it is possible to generate a graph using all the values associated with a plurality of water vapor absorption frequencies. In addition, it is possible to generate a graph using the value associated with the water vapor absorption frequency arbitrarily selected by the user from a plurality of water vapor absorption frequencies.

The radar diagram shown in FIG8(a) is generated based on a megahertz wave signal related to a certain liquid sample. The radar diagram shown in FIG8(b) is generated based on a megahertz wave signal related to other liquid samples. In this way, the graph generated by the characteristic analysis unit 14 reflects the difference in the characteristics of the liquid samples, and the difference in the characteristics becomes a difference in the shape of the graph and appears clearly. By this, the characteristics corresponding to the characteristics of the liquid sample can be easily visualized in the form of a graph. For example, the characteristics of the liquid sample that can only be grasped by human senses in the prior art can also be objectively visualized as the shape of the radar diagram.

Here, although an example of generating two radar images based on two different liquid samples is shown, the present invention is not limited to this. For example, it is also possible to calculate the ratio Ri between the amplitude _a0i or width a1i of the normal distribution function calculated based on the spectrum of one of the liquid samples and the amplitude _a0i or width _a1i of the normal distribution function calculated based on the spectrum of the other liquid sample for each of _the multiple center frequencies _λ0i , and generate a graph showing the relationship between the center frequency _λ0i and _the ratio _Ri .

In addition, it is also possible to construct a system that overlays and visualizes multiple radar images generated by multiple electromagnetic wave signals. FIG. 9 is a diagram showing one example. The example shown in FIG. 9 is a system that overlays and visualizes multiple radar images obtained by analyzing "spectra measured multiple times by the spectrometer 20 for the same liquid sample with time intervals." In this example, "how the characteristics of the liquid sample change over time" is visualized.

The example of FIG. 9 above is a representation of the state change of one liquid sample with the passage of time by using multiple radar images, but it is also possible to construct a configuration in which multiple radar images generated based on multiple liquid samples are overlapped and visualized. In this case, if the multiple liquid samples have the same characteristics, the multiple radar images generated are overlapped as approximately the same shape. On the other hand, if the multiple liquid samples have different characteristics, the radar images generated are different shapes.

According to this, by generating a radar map for a plurality of liquid samples with unknown characteristics, it is possible to easily determine whether the liquid samples have the same characteristics or different characteristics. Furthermore, by generating a radar map based on a liquid sample with a known characteristic and each of a plurality of liquid samples with unknown characteristics, it is also possible to easily identify a liquid sample with the same characteristics as the known characteristics.

Here, an example of superimposing and visualizing a plurality of radar images obtained based on "a plurality of frequency spectra measured at a plurality of time points for a single liquid sample" is shown, but the present invention is not limited to this. For example, it can also be constructed as "for each of the multiple center frequencies λ _0i , the ratio R it between "the amplitude a _0i or the width a _1i of the normal distribution function calculated based on the spectrum of the liquid sample at a certain time point" and "the amplitude a _0i or the width a _1i of the normal distribution function calculated based on the spectrum of the liquid sample at other time points" _is calculated separately, and a graph showing the relationship between the center frequency λ _0i and the ratio R _it is generated."

In addition, the form of the generated graph is not limited to a radar graph. For example, it can be configured to generate a line graph, a bar graph, a distribution graph, etc. In addition, it can be configured to generate the amplitude, width, or area size of each center frequency by a pie chart. In addition, it can be configured to generate a "bubble chart with the vertical axis set as the amplitude, the horizontal axis set as the width, and the center frequency expressed by the size of the circle." For example, it can be configured to generate a graph (radar graph, line graph, bar graph, distribution graph, etc.) that displays the relationship between amplitude and width without introducing the center frequency into the elements of the graph.

Furthermore, it can also be configured to generate a branch graph as shown in FIG10. FIG10 uses the central frequency λ _0i of the complex number as the axis (branch) of the complex number, and expresses one of the amplitude a _0i , width a _1i , and area A _i of the normal distribution function by the length of each axis, and expresses the other of the amplitude a _0i , width a _1i , and area A _i of the normal distribution function by the size of the circle drawn at the front end of the axis.

The above are several examples of generating graphs using at least two of the center frequency λ _0i , amplitude a _0i , and width a _1i of the complex normal distribution function used in generating the water vapor spectrum.

Furthermore, the characteristic analysis unit 14 may be configured to generate a graph showing the relationship between at least two of the coefficients _bj of the polynomial function used in generating the liquid spectrum. The graph generated using at least two of the coefficients b _i of the polynomial function may be a radar graph, a line graph, a bar graph, a distribution graph, a pie graph, a branch graph, or the like.

Furthermore, the characteristic analysis unit 14 may be configured to generate a graph using at least two values as parameters, namely, a value set for the property of the complex fitting function used in the fitting of the water vapor fitting processing unit 12 and a value set for the property of the complex fitting function used in the fitting of the liquid fitting processing unit 13. For example, the characteristic analysis unit 14 may be configured to generate a graph showing the relationship between "at least one of the center frequency λ _0i , amplitude a _0i , and width a _1i of the complex normal distribution function used in the fitting of the water vapor fitting processing unit 12" and "at least one of the coefficients b _j of the polynomial function used in the fitting of the liquid fitting processing unit 13".

Here, it is also possible to construct a graph that directly shows the relationship between "at least one of the center frequency λ _0i , the amplitude a _0i , and the width a _1i " and "at least one of the coefficients b _j of the polynomial function".

Alternatively, it may be configured such that after the types of parameters are made consistent by "replacing each coefficient _bj of the polynomial function with the center frequency _λ0j , amplitude _a0j , and width _a1j ", a graph is generated showing the relationship between "at least one of the center frequency _λ0i , amplitude _a0i , and width _a1i obtained from the water vapor spectrum" and "at least one of the center frequency _λ0j , amplitude _a0j , and width _a1j obtained from the liquid spectrum". In this case, the liquid fitting processing unit 13 performs the first fitting using the polynomial function shown in formula (2) as a complex fitting function, and performs the second fitting using a synthetic waveform of a complex normal distribution function on the spectrum obtained by the first fitting (see Figure 6).

FIG11 shows an example of a "branch diagram generated by making the parameters obtained from the water vapor spectrum and the parameters obtained from the liquid spectrum consistent with each other and including both the elements of the water vapor spectrum and the elements of the liquid spectrum." The branch diagram shown in FIG11 is generated by making one of the five axes (branches) an element of the water vapor spectrum and the remaining four axes as elements of the liquid spectrum. That is, a central frequency λ _0i obtained from the water vapor spectrum is used as an axis, and one of the amplitude a _0i , width a _1i , and area A _i of the normal distribution function is expressed by the length of the axis, and the other one of the amplitude a _0i , width a _1i , and area A _i of the normal distribution function is expressed by the size of the circle drawn at the front end of the axis. Furthermore, the four central frequencies obtained from the liquid spectrum are used as four axes, and one of the amplitude a _0j , width a _1j , and area A _j of the normal distribution function is expressed by the length of the four axes, and the other one of the amplitude a _0j , width a _1j , and area A _j of the normal distribution function is expressed by the size of the circle drawn at the front end of the four axes.

In addition, in order to obtain the values of the center frequency λ _0j , amplitude a _0j and width a _1j of the complex normal distribution function in relation to the liquid spectrum, an example is described in which the first fitting is performed using the polynomial function shown in formula (2) and then the second fitting is performed using the synthetic waveform of the complex normal distribution function for the spectrum obtained by the first fitting. However, the present invention is not limited to this. That is, it is also possible to configure the fitting using the synthetic waveform of the complex normal distribution function directly for the spectrum other than the peak component of the water vapor absorption frequency in the spectrum obtained by the spectrum acquisition unit 11.

As described in detail above, in the present embodiment, the spectrum of the water vapor absorption frequency in the spectrum of the electromagnetic wave signal obtained by the spectrometer 20 is fitted with the waveform of a fitting function or the synthetic waveform of a plurality of fitting functions, and the characteristics of the liquid sample are analyzed using a value determined for the property of the fitting function used in the fitting.

The spectrum of water vapor absorption frequency reflects the "properties of water vapor generated around the liquid sample". Since water vapor contains the properties of the liquid sample, the properties of the liquid sample can be analyzed by processing the spectrum of water vapor absorption frequency. In addition, the spectrum of water vapor is easy to read because the peak is clear and sharp. According to the present embodiment constructed as described above, the spectrum caused by water vapor generated from the liquid sample is approximated by a waveform of a single fitting function or a composite waveform of multiple fitting functions in a form that inherits the characteristics of the liquid sample, and the characteristics of the liquid sample are analyzed based on the value of the fitting function used in the approximation. Thus, according to the present embodiment, by analyzing the electromagnetic wave signal detected by the spectrometer 20 and actively utilizing the fact that electromagnetic waves are absorbed by water vapor, it is possible to easily detect the inherent characteristics of the liquid sample.

Furthermore, in the present embodiment, the spectrum of the electromagnetic wave signal obtained by the spectrometer 20, except for the peak component of the water vapor absorption frequency, is fitted with the synthetic waveform of the complex fitting function, and the characteristics of the liquid sample are analyzed using a value determined for the property of the complex fitting function used in the fitting. Here, the entire spectrum is fitted in such a way that the relationship "entire fitting spectrum = water vapor spectrum + liquid spectrum" holds.

By doing so, we can analyze the characteristics of the liquid sample, which are reflected in the spectrum of the water vapor absorption frequency due to the electromagnetic wave passing through the water vapor around the liquid sample, and the characteristics of the liquid sample, which are reflected in the spectrum of frequencies other than the water vapor absorption frequency due to the electromagnetic wave passing through or reflecting the liquid sample, and can easily detect the inherent characteristics of the liquid sample.

Furthermore, in this embodiment, when fitting the liquid spectrum, the absorbance data of the water vapor absorption frequency is not deleted, but only the peak component of the water vapor absorption frequency is separated and fitted. Therefore, the liquid spectrum can be obtained without "loss of spectral data at the water vapor absorption frequency". By this, the characteristics of the liquid sample reflected in the electromagnetic wave by passing through or reflecting the liquid sample can be analyzed with higher accuracy than the method described in Patent Document 1.

In addition, in the above-mentioned embodiment, although the example of "making the water vapor fitting processing unit 12 fit using a complex normal distribution function as in (Formula 1), and making the characteristic analysis unit 14 analyze the characteristics of the liquid sample using the center frequency, amplitude, and width of the complex normal distribution function" is explained, the present invention is not limited to this. For example, it can also be configured as "making the water vapor fitting processing unit 12 fit using a polynomial function as in (Formula 2), and making the characteristic analysis unit 14 analyze the characteristics of the liquid sample using each coefficient of the polynomial function".

Furthermore, in the above-mentioned embodiment, although the example of "making the liquid fitting processing unit 13 fit using a polynomial function as in (Formula 2), and making the characteristic analysis unit 14 analyze the characteristics of the liquid sample using at least two of the coefficients of the polynomial function" is explained, the present invention is not limited to this. For example, it can also be configured as "making the liquid fitting processing unit 13 fit using a composite waveform of a complex normal distribution function, and making the characteristic analysis unit 14 analyze the characteristics of the liquid sample using at least two of the center frequency, amplitude, and width of the complex normal distribution function". Furthermore, when the spectrum of the liquid frequency can be approximated to a certain extent by the waveform of a normal distribution function, it can also be constituted as "using the waveform of the normal distribution function to perform fitting, and allowing the characteristic analysis unit 14 to use any one of the center frequency, amplitude and width of the normal distribution function to analyze the characteristics of the liquid sample."

Furthermore, in the above-mentioned embodiment, although several examples of "using at least two values including a predetermined value for the property of a plurality of fitting functions used in the fitting of the water vapor fitting processing section 12 and a predetermined value for the property of a plurality of fitting functions used in the fitting of the liquid fitting processing section 13 to analyze the characteristics of the liquid sample" are described, the present invention is not limited to the above-mentioned examples.

Furthermore, in the above-mentioned embodiment, although the electromagnetic wave signal analysis device 10 is described as having both the water vapor fitting processing unit 12 and the liquid fitting processing unit 13, the present invention is not limited thereto. For example, the electromagnetic wave signal analysis device 10 may be configured to have only the water vapor fitting processing unit 12. In this case, the spectrum acquisition unit 11 may be configured to "acquire the spectrum of the water vapor absorption frequency" instead of "acquire the spectrum of the entire frequency".

In the above-mentioned embodiment, as one example of the function used in the fitting, a normal distribution function (Gaussian function) is used, but it can also be realized by using a Lorenz function, a Voigt function, etc. In addition, a probability distribution function such as a Poisson distribution function (probability mass function, cumulative distribution function) or a Chi-square distribution function (probability density function, cumulative distribution function) which is not centrally symmetric but has an asymmetric shape may be used, and other functions whose waveform shape becomes a mountain shape may be used. When a probability distribution function is used, values that represent the properties of the probability distribution (e.g., the central value or the most frequent value of the amplitude, the frequency at which the amplitude value is obtained, the frequency bandwidth within which the amplitude is above or below a specific value, etc.) are used as parameters for fitting. When a mountain-shaped function is used, the maximum amplitude that is the vertex, the frequency at which the maximum amplitude value is obtained, the frequency bandwidth within which the amplitude is above or below a specific value, etc.) are used as parameters for fitting.

Furthermore, in the above embodiment, although the example of "using absorbance as the characteristic value of the electromagnetic wave signal and obtaining a spectrum expressing the absorbance relative to the frequency" is described, other characteristic values such as transmittance may also be used.

In addition, the above-mentioned embodiments are only examples of the embodiments of the present invention, and the technical scope of the present invention should not be limited based on these embodiments. That is, the present invention can be implemented in various forms without departing from its gist or main features.

10: Electromagnetic wave signal analysis device 11: Spectrum acquisition unit 12: Water vapor simulation processing unit 13: Liquid simulation processing unit 14: Characteristics analysis unit 20: Spectrometer

[Figure 1] is a block diagram showing an example of the functional configuration of the electromagnetic wave signal analysis device according to the present embodiment. [Figure 2] is a diagram showing an example of the configuration of the liquid film generating device used in the present embodiment. [Figure 3] is a diagram showing the appearance of water vapor generated around the liquid sample generated by the nozzle. [Figure 4] is a diagram showing an example of the spectrum obtained by the spectrum acquisition unit of the present embodiment. [Figure 5] is a diagram showing an example of the water vapor spectrum generated by the water vapor simulation processing unit of the present embodiment. [Fig. 6] is a diagram showing one example of a liquid spectrum generated by the liquid fitting processing unit of the present embodiment. [Fig. 7] is a diagram showing one example of a spectrum of the entire body fitted by the present embodiment. [Fig. 8] is a diagram showing one example of a graph generated by the characteristic analysis unit of the present embodiment. [Fig. 9] is a diagram showing one example of a graph generated by the characteristic analysis unit of the present embodiment. [Fig. 10] is a diagram showing one example of a graph generated by the characteristic analysis unit of the present embodiment. [Fig. 11] is a diagram showing one example of a graph generated by the characteristic analysis unit of the present embodiment.

10: Electromagnetic wave signal analysis device

11: Spectrum acquisition unit

12: Water vapor preparation treatment department

13: Liquid fitting processing department

14: Characteristic Analysis Department

20: Spectrometer

Claims (12)

An electromagnetic wave signal analysis device analyzes an electromagnetic wave signal obtained by a spectrometer, wherein the spectrometer detects electromagnetic waves transmitted or reflected by a film-like liquid sample generated in a space where liquid is ejected from a nozzle. The electromagnetic wave signal analysis device is characterized by comprising: A spectrum acquisition unit that acquires a spectrum having characteristic values relative to frequency generated by the electromagnetic wave signal; and The water vapor fitting processing section fits the spectrum of the water vapor absorption frequency in the spectrum acquired by the spectrum acquisition section with a waveform of a fitting function or a composite waveform of multiple fitting functions, wherein the water vapor absorption frequency is a frequency at which the absorption of electromagnetic waves by water vapor becomes larger; and The characteristic analysis section analyzes the characteristics of the liquid sample using at least two values determined for the properties of the fitting function used in the fitting of the water vapor fitting processing section. The electromagnetic wave signal analysis device as described in claim 1, further comprising: A liquid fitting processing unit, which fits the spectrum other than the peak component of the water vapor absorption frequency in the spectrum acquired by the spectrum acquisition unit with a waveform of a fitting function or a composite waveform of multiple fitting functions, The characteristic analysis unit, which uses at least two values including a value determined for the property of the fitting function used in the fitting of the water vapor fitting processing unit and a value determined for the property of the fitting function used in the fitting of the liquid fitting processing unit, to analyze the characteristics of the liquid sample. The electromagnetic wave signal analysis device as described in claim 1 or 2, wherein, the water vapor fitting processing section uses a single or multiple normal distribution function, Lorentz function, Voigt function, probability distribution function or other functions whose waveform shape becomes mountain-shaped as the fitting function to perform the fitting, the characteristic analysis section uses at least one of the center frequency, amplitude and width of the fitting function used in the fitting of the water vapor fitting processing section to analyze the characteristics of the liquid sample. The electromagnetic wave signal analysis device as described in claim 1 or 2, wherein, the water vapor fitting processing section uses a polynomial function including complex terms as the fitting function to perform the fitting, and the characteristic analysis section uses at least one of the coefficients of the complex terms in the polynomial function used in the fitting of the water vapor fitting processing section to analyze the characteristics of the liquid sample. The electromagnetic wave signal analysis device as described in claim 2, wherein, the liquid fitting processing unit uses a polynomial function including complex terms as the fitting function to perform the fitting, and the characteristic analysis unit uses at least one of the coefficients of the complex terms in the polynomial function used in the fitting of the liquid fitting processing unit to analyze the characteristics of the liquid sample. The electromagnetic wave signal analysis device as described in claim 2, wherein, the liquid fitting processing unit uses a 1 or plural normal distribution function, Lorentz function, Voigt function, probability distribution function or other functions whose waveform shape becomes mountain-shaped as the fitting function to perform the fitting, the characteristic analysis unit uses at least one of the center frequency, amplitude and width of the fitting function used in the fitting of the liquid fitting processing unit to analyze the characteristics of the liquid sample. The electromagnetic wave signal analysis device as described in claim 2, wherein, the water vapor fitting processing unit uses 1 or a complex normal distribution function, Lorentz function, Voigt function, probability distribution function or other functions whose waveform shape becomes mountain-shaped as the fitting function to perform the fitting, the liquid fitting processing unit uses a polynomial function including complex terms as the fitting function to perform the first fitting, and for the spectrum obtained by the first fitting, uses 1 or a complex normal distribution function, Lorentz function, Voigt function as the fitting function. function), probability distribution function or other functions whose waveform shape becomes mountain-shaped to perform the second fitting, The above-mentioned characteristic analysis section uses at least one of the center frequency, amplitude and width of the above-mentioned fitting function used in the fitting of the above-mentioned water vapor fitting processing section, and at least one of the center frequency, amplitude and width of the above-mentioned fitting function used in the above-mentioned second fitting of the above-mentioned liquid fitting processing section to analyze the characteristics of the above-mentioned liquid sample. An electromagnetic wave signal analysis device analyzes an electromagnetic wave signal obtained by a spectrometer, wherein the spectrometer detects electromagnetic waves transmitted or reflected by a film-like liquid sample generated in a space where liquid is sprayed from a nozzle. The electromagnetic wave signal analysis device is characterized by comprising: A spectrum acquisition unit that acquires a spectrum having characteristic values relative to a water vapor absorption frequency generated based on the electromagnetic wave signal, wherein the water vapor absorption frequency is a frequency at which the absorption of electromagnetic waves by water vapor becomes greater; and The water vapor fitting processing section fits the spectrum acquired by the spectrum acquisition section with a waveform of a fitting function or a composite waveform of multiple fitting functions; and The characteristic analysis section analyzes the characteristics of the liquid sample using at least two values determined for the properties of the fitting function used in the fitting of the water vapor fitting processing section. The electromagnetic wave signal analysis device as described in claim 1, wherein the characteristic analysis section generates a graph by using at least two values of the properties of the fitting function used in the fitting of the water vapor fitting processing section as parameters. The electromagnetic wave signal analysis device as described in claim 2, wherein the characteristic analysis section generates a graph using at least two values as parameters, including a value for the property of the fitting function used in the fitting of the water vapor fitting processing section and a value for the property of the fitting function used in the fitting of the liquid fitting processing section. An electromagnetic wave signal analysis program analyzes an electromagnetic wave signal obtained by a spectroscopic device, wherein the spectroscopic device detects electromagnetic waves transmitted or reflected by a film-like liquid sample generated in a space where liquid is ejected from a nozzle. The electromagnetic wave signal analysis program is characterized in that it is used to make a computer function as follows: Spectrum acquisition means is to acquire a spectrum having characteristic values relative to frequency generated by the above-mentioned electromagnetic wave signal; and The water vapor fitting processing means is to fit the spectrum of the water vapor absorption frequency in the above-mentioned spectrum obtained by the above-mentioned spectrum acquisition means with a waveform of a fitting function or a composite waveform of multiple fitting functions, wherein the water vapor absorption frequency is a frequency at which the absorption of electromagnetic waves due to water vapor becomes larger; and The characteristic analysis means is to analyze the characteristics of the above-mentioned liquid sample using at least two values determined for the properties of the above-mentioned fitting function used in the fitting of the above-mentioned water vapor fitting processing means. The electromagnetic wave signal analysis program as described in claim 11, wherein the computer is further made to function as the following means: The liquid fitting processing means is to fit the spectrum other than the peak component of the water vapor absorption frequency in the spectrum obtained by the spectrum acquisition means with a waveform of a fitting function or a composite waveform of multiple fitting functions, The characteristic analysis means is to analyze the characteristics of the liquid sample using at least two values including a value for the property of the fitting function used in the fitting of the water vapor fitting processing means and a value for the property of the fitting function used in the fitting of the liquid fitting processing means.

Images