CN111007033A - Trace acetylene gas concentration detection method based on spectrum and power spectrum feature fusion - Google Patents
Trace acetylene gas concentration detection method based on spectrum and power spectrum feature fusion Download PDFInfo
- Publication number
- CN111007033A CN111007033A CN201911248926.1A CN201911248926A CN111007033A CN 111007033 A CN111007033 A CN 111007033A CN 201911248926 A CN201911248926 A CN 201911248926A CN 111007033 A CN111007033 A CN 111007033A
- Authority
- CN
- China
- Prior art keywords
- acetylene gas
- concentration
- trace
- trace acetylene
- fusion
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- HSFWRNGVRCDJHI-UHFFFAOYSA-N alpha-acetylene Natural products C#C HSFWRNGVRCDJHI-UHFFFAOYSA-N 0.000 title claims abstract description 216
- 125000002534 ethynyl group Chemical group [H]C#C* 0.000 title claims abstract description 216
- 230000004927 fusion Effects 0.000 title claims abstract description 77
- 238000001514 detection method Methods 0.000 title claims abstract description 66
- 238000001228 spectrum Methods 0.000 title claims abstract description 35
- 239000011159 matrix material Substances 0.000 claims abstract description 57
- 238000002329 infrared spectrum Methods 0.000 claims abstract description 45
- 238000000034 method Methods 0.000 claims abstract description 14
- 230000003595 spectral effect Effects 0.000 claims description 45
- 238000000862 absorption spectrum Methods 0.000 claims description 19
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000012847 principal component analysis method Methods 0.000 claims description 3
- 239000007789 gas Substances 0.000 description 149
- 238000010521 absorption reaction Methods 0.000 description 9
- 238000011088 calibration curve Methods 0.000 description 4
- 230000035945 sensitivity Effects 0.000 description 4
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000004566 IR spectroscopy Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- QFXZANXYUCUTQH-UHFFFAOYSA-N ethynol Chemical group OC#C QFXZANXYUCUTQH-UHFFFAOYSA-N 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 150000002739 metals Chemical class 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229910052757 nitrogen Inorganic materials 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000010926 purge Methods 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3504—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing gases, e.g. multi-gas analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/359—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light using near infrared light
Abstract
A trace acetylene gas concentration detection method based on spectrum and power spectrum feature fusion belongs to the field of trace gas detection and comprises the following steps: trace acetylene gas samples of different concentrations were prepared. Raw infrared spectra were collected for each gas sample. And acquiring a characteristic wavelength information matrix according to the preprocessed infrared spectrum. And acquiring a characteristic frequency information matrix according to the power spectrum of the original infrared spectrum. And fusing the two information matrixes to obtain a fusion characteristic information matrix of each gas sample. And repeatedly acquiring the original infrared spectrum of each gas sample and executing related subsequent steps to enable the number of the fusion characteristic information matrixes of each gas sample to reach a preset value. And dividing the fusion characteristic information matrix of each gas sample into a training sample and a correction sample. And obtaining a final detection model based on the training sample and the correction sample. And detecting the acetylene gas sample to be detected based on the final detection model. The method can realize the rapid and accurate detection of the concentration of the trace acetylene gas.
Description
Technical Field
The invention relates to a trace acetylene gas concentration detection method, and belongs to the field of trace gas detection.
Background
Acetylene is a colorless, flammable gas that produces an oxyacetylene flame when burned that can be used to cut or weld metals. In addition, acetylene is an important organic raw material and is widely used in industrial production. However, since acetylene has flammability, when the content of acetylene in the air reaches 2.3% to 72.3%, explosion occurs upon exposure to open flame. Therefore, in order to ensure the production safety, the concentration of the trace acetylene gas in the industrial field needs to be accurately detected.
The existing acetylene gas concentration detection method mainly comprises an infrared absorption spectrometry, wherein the method selects absorption characteristics based on the near infrared spectrum of acetylene gas molecules, and determines the concentration of acetylene gas by utilizing the relation between the concentration of acetylene gas and absorption intensity.
However, since the absorption spectrum width of acetylene gas in the near-infrared band is only nanometer level, when the concentration of acetylene gas is low, the absorption peak signal of acetylene gas is very weak, and the acetylene gas is easily submerged in the background noise of the spectrum, so that it is difficult to detect the concentration of trace acetylene gas quickly and accurately directly according to the absorption spectrum.
Disclosure of Invention
The invention provides a trace acetylene gas concentration detection method based on the fusion of spectral and power spectral characteristics, which aims to solve the problem that the concentration of trace acetylene gas cannot be quickly and accurately detected by the existing acetylene gas concentration detection method based on the infrared absorption spectrum principle.
The trace acetylene gas concentration detection method based on the fusion of the spectral and power spectral characteristics comprises the following steps:
s1, preparing a preset number of trace acetylene gas samples with different concentrations;
s2, collecting the original infrared spectrum of each trace acetylene gas sample;
s3, preprocessing the original infrared spectrum of each trace acetylene gas sample;
s4, extracting a plurality of characteristic wavelength information related to acetylene concentration in the infrared spectrum of each trace acetylene gas sample after pretreatment to obtain a characteristic wavelength information matrix formed by the characteristic wavelength information;
s5, obtaining a power spectrum of an original infrared spectrum of each trace acetylene gas sample;
s6, extracting a plurality of characteristic frequency information related to acetylene concentration in a power spectrum of an original infrared spectrum of each trace acetylene gas sample to obtain a characteristic frequency information matrix formed by the characteristic frequency information;
s7, fusing the characteristic wavelength information matrix and the characteristic frequency information matrix of each trace acetylene gas sample to obtain a fused characteristic information matrix;
s8, repeatedly executing the steps S2 to S7 until the number of the fusion characteristic information matrixes of each trace acetylene gas sample reaches a preset value;
s9, dividing a plurality of fusion characteristic information matrixes of each trace acetylene gas sample into a training sample and a correction sample;
s10, establishing a trace acetylene gas concentration initial detection model by adopting each fusion characteristic information matrix as a training sample and the known concentration information of the trace acetylene gas sample corresponding to the fusion characteristic information matrix;
s11, correcting the initial detection model of the concentration of the trace acetylene gas by adopting each fusion characteristic information matrix as a correction sample and the known concentration information of the trace acetylene gas sample corresponding to the fusion characteristic information matrix to obtain a final detection model of the concentration of the trace acetylene gas;
and S12, collecting the acetylene gas sample to be detected, obtaining a fusion characteristic information matrix of the acetylene gas sample, and inputting the fusion characteristic information matrix into a final detection model of the concentration of the trace acetylene gas to obtain the acetylene concentration information of the acetylene gas sample.
Preferably, the step S1 of preparing the predetermined parts of the trace acetylene gas samples with different concentrations specifically includes:
preparing 9 parts of trace acetylene gas samples, wherein the acetylene concentrations of the 9 parts of trace acetylene gas samples are 100-900 ppm respectively, and the concentration gradient is 100 ppm.
Preferably, the raw infrared spectrum of each trace acetylene gas sample collected in step S2 is specifically:
under the environment condition of standard atmospheric pressure and room temperature of 26 ℃, an infrared spectrometer is used for collecting the absorption spectrum of each trace acetylene gas sample within the wavelength range of 1529.5-1532.5 nm.
Preferably, the preprocessing of the raw infrared spectrum of each trace acetylene gas sample in step S3 is specifically:
and performing wavelet denoising, multivariate scattering correction and standard normal variable correction on the original infrared spectrum of each trace acetylene gas sample.
Preferably, in step S4, a plurality of characteristic wavelength information related to the acetylene concentration in the pre-processed infrared spectrum of each trace acetylene gas sample is extracted by using a principal component analysis method.
Preferably, step S5 obtains the corresponding power spectrum by fourier transforming the raw infrared spectrum of each trace acetylene gas sample.
Preferably, the characteristic frequency information related to the acetylene concentration in the power spectrum of the original infrared spectrum of each trace acetylene gas sample extracted in step S6 is 18Hz and 23 Hz.
Preferably, the number of the fusion characteristic information matrix per trace acetylene gas sample is 8 at the end of the execution of step S8;
in step S9, 5 fusion characteristic information matrices are randomly selected from 8 fusion characteristic information matrices of each trace acetylene gas sample as training samples, and the remaining 3 fusion characteristic information matrices are used as correction samples.
Preferably, step S10 establishes an initial detection model of the trace acetylene gas concentration based on a partial least squares algorithm.
Preferably, the correction decision coefficient of the final detection model of the trace acetylene gas concentration is more than 0.95.
The method for detecting the concentration of the trace acetylene gas based on the characteristic fusion of the spectrum and the power spectrum obtains a characteristic wavelength information matrix according to the preprocessed infrared spectrum of each trace acetylene gas sample, obtains a characteristic frequency information matrix according to the power spectrum of the original infrared spectrum of each trace acetylene gas sample, and obtains a fusion characteristic information matrix of each trace acetylene gas sample by fusing the characteristic frequency information matrix and the characteristic frequency information matrix which correspond to each other. And repeatedly acquiring the original infrared spectrum of each trace acetylene gas sample and executing related subsequent steps to enable the number of the fusion characteristic information matrixes of each trace acetylene gas sample to reach a preset value. Then, dividing a plurality of fusion characteristic information matrixes of each trace acetylene gas sample into a training sample and a correction sample, establishing a trace acetylene gas concentration initial detection model by adopting each fusion characteristic information matrix as the training sample and the known acetylene concentration information of the trace acetylene gas sample corresponding to the fusion characteristic information matrix, and correcting the trace acetylene gas concentration initial detection model by adopting each fusion characteristic information matrix as the correction sample and the known acetylene concentration information of the trace acetylene gas sample corresponding to the fusion characteristic information matrix to obtain a trace acetylene gas concentration final detection model.
In the method for detecting the concentration of the trace acetylene gas based on the fusion of the spectral characteristics and the power spectral characteristics, the fusion characteristic information matrix of the trace acetylene gas sample is a basis for establishing and correcting an initial detection model of the concentration of the trace acetylene gas so as to obtain a final detection model of the concentration of the trace acetylene gas. And the fusion characteristic information matrix of the trace acetylene gas sample is obtained by fusing the characteristic wavelength information matrix and the characteristic frequency information matrix of the trace acetylene gas sample. On one hand, because a plurality of pieces of characteristic wavelength information forming the characteristic wavelength information matrix are all from the infrared spectrum after the pretreatment of the trace acetylene gas sample, the final detection model for the trace acetylene gas concentration has higher sensitivity, and can realize the rapid detection of the trace acetylene gas concentration. On the other hand, a plurality of characteristic frequency information forming the characteristic frequency information matrix are all from a power spectrum of an original infrared spectrum of the trace acetylene gas sample, the power spectrum is obtained by performing Fourier transform on the corresponding original infrared spectrum, and the amplitudes of the plurality of characteristic frequency information and the concentration of the acetylene gas sample are in an obvious linear relation. The distribution range of the energy of the infrared spectrum absorption signal and the distribution range of the energy of the noise on the frequency domain are different, so that the characteristic frequency information is not interfered by the system noise, the detection result of the final detection model of the concentration of the trace acetylene gas has higher accuracy, and the concentration of the trace acetylene gas can be accurately detected.
Therefore, the final detection model of the concentration of the trace acetylene gas, provided by the method for detecting the concentration of the trace acetylene gas based on the characteristic fusion of the spectrum and the power spectrum, can give consideration to both the sensitivity and the accuracy of detection due to the comprehensive consideration of the characteristic information of the infrared absorption spectrum and the characteristic information of the power spectrum of the infrared absorption spectrum, and can realize the rapid and accurate detection of the trace acetylene gas. Furthermore, the detection object of the trace gas concentration detection method provided by the invention is not limited to acetylene gas, and is also suitable for the concentration detection of other gases with periodic absorption spectrum characteristics.
Drawings
The trace acetylene gas concentration detection method based on the fusion of spectral and power spectral features according to the present invention will be described in more detail below based on examples and with reference to the accompanying drawings, in which:
FIG. 1 is a flow chart of an implementation of a trace acetylene gas concentration detection method based on the fusion of spectral and power spectral features according to an embodiment;
FIG. 2 is a schematic block diagram of a trace acetylene gas concentration detection method based on the fusion of spectral and power spectral features according to an embodiment;
FIG. 3 is a graph showing the standard absorption spectrum of acetylene gas in the wavelength range of 1529.5-1532.5 nm;
FIG. 4 is a power spectrum of a standard absorption spectrum of acetylene gas in the wavelength range of 1529.5-1532.5 nm;
FIG. 5 is an absorption spectrum of a trace acetylene gas sample with the concentration of 100-900 ppm mentioned in the example in the wavelength range of 1529.5-1532.5 nm;
FIG. 6 is a power spectrum of an absorption spectrum of a trace acetylene gas sample with the concentration of 100-900 ppm in the wavelength range of 1529.5-1532.5 nm mentioned in the example.
Detailed Description
The trace acetylene gas concentration detection method based on the fusion of the spectral characteristics and the power spectral characteristics will be further described with reference to the accompanying drawings.
Example (b): the present embodiment is described in detail below with reference to fig. 1 to 6.
Fig. 1 is a flow chart of an implementation of the trace acetylene gas concentration detection method based on the fusion of spectral and power spectral features according to this embodiment. Fig. 2 is a schematic block diagram of a trace acetylene gas concentration detection method based on the fusion of spectral and power spectral features according to the embodiment.
Referring to fig. 1 and fig. 2, the trace acetylene gas concentration detection method based on the fusion of spectral and power spectral features according to the embodiment includes the following steps:
s1, preparing a preset number of trace acetylene gas samples with different concentrations;
s2, collecting the original infrared spectrum of each trace acetylene gas sample;
s3, preprocessing the original infrared spectrum of each trace acetylene gas sample;
s4, extracting a plurality of characteristic wavelength information related to acetylene concentration in the infrared spectrum of each trace acetylene gas sample after pretreatment to obtain a characteristic wavelength information matrix formed by the characteristic wavelength information;
s5, obtaining a power spectrum of an original infrared spectrum of each trace acetylene gas sample;
s6, extracting a plurality of characteristic frequency information related to acetylene concentration in a power spectrum of an original infrared spectrum of each trace acetylene gas sample to obtain a characteristic frequency information matrix formed by the characteristic frequency information;
s7, fusing the characteristic wavelength information matrix and the characteristic frequency information matrix of each trace acetylene gas sample to obtain a fused characteristic information matrix;
s8, repeatedly executing the steps S2 to S7 until the number of the fusion characteristic information matrixes of each trace acetylene gas sample reaches a preset value;
s9, dividing a plurality of fusion characteristic information matrixes of each trace acetylene gas sample into a training sample and a correction sample;
s10, establishing a trace acetylene gas concentration initial detection model by adopting each fusion characteristic information matrix as a training sample and the known concentration information of the trace acetylene gas sample corresponding to the fusion characteristic information matrix;
s11, correcting the initial detection model of the concentration of the trace acetylene gas by adopting each fusion characteristic information matrix as a correction sample and the known concentration information of the trace acetylene gas sample corresponding to the fusion characteristic information matrix to obtain a final detection model of the concentration of the trace acetylene gas;
and S12, collecting the acetylene gas sample to be detected, obtaining a fusion characteristic information matrix of the acetylene gas sample, and inputting the fusion characteristic information matrix into a final detection model of the concentration of the trace acetylene gas to obtain the acetylene concentration information of the acetylene gas sample.
In this embodiment, the steps of preparing the predetermined portions of the trace acetylene gas samples with different concentrations in step S1 are specifically as follows:
preparing 9 parts of trace acetylene gas samples, wherein the acetylene concentrations of the 9 parts of trace acetylene gas samples are 100-900 ppm respectively, and the concentration gradient is 100 ppm.
Before preparing a trace acetylene gas sample, firstly, purging a gas chamber by using nitrogen to remove the interference of residual gas in the gas chamber, and then respectively injecting 100-900 ppm acetylene gas into the gas chamber. In the preparation process, an acetylene gas analyzer is adopted to calibrate the acetylene concentration of each trace acetylene gas sample so as to ensure the accuracy of the concentration of each trace acetylene gas sample.
In this embodiment, the original infrared spectrum obtained by collecting each trace acetylene gas sample in step S2 specifically includes:
under the environment condition of standard atmospheric pressure and room temperature of 26 ℃, an infrared spectrometer is used for collecting the absorption spectrum of each trace acetylene gas sample within the wavelength range of 1529.5-1532.5 nm.
In this embodiment, the preprocessing of the raw infrared spectrum of each trace acetylene gas sample in step S3 specifically includes:
and performing wavelet denoising, multivariate scattering correction and standard normal variable correction on the original infrared spectrum of each trace acetylene gas sample.
In the present embodiment, step S4 adopts a principal component analysis method to extract a plurality of characteristic wavelength information related to the acetylene concentration in the preprocessed infrared spectrum of each trace acetylene gas sample.
In the present embodiment, step S5 obtains a corresponding power spectrum by performing fourier transform on the original infrared spectrum of each trace acetylene gas sample.
In the present embodiment, the characteristic frequency information related to the acetylene concentration in the power spectrum of the original infrared spectrum of each trace acetylene gas sample extracted in step S6 is 18Hz and 23 Hz.
In this embodiment, the characteristic wavelength information matrix is a, the characteristic frequency information matrix is B, and the corresponding fusion characteristic information matrix M is:
M=[A;B]
in the present embodiment, the number of the fusion characteristic information matrices per trace acetylene gas sample is 8 at the time of completion of step S8;
in step S9, 5 fusion characteristic information matrices are randomly selected from 8 fusion characteristic information matrices of each trace acetylene gas sample as training samples, and the remaining 3 fusion characteristic information matrices are used as correction samples.
In the present embodiment, step S10 establishes an initial detection model of the trace acetylene gas concentration based on a partial least squares algorithm:
the concentration characteristic values of a certain trace acetylene gas sample are X1 and X2 … Xn, a fusion characteristic information matrix M is composed of X1 and X2 … Xn, model fusion is carried out by using a weighted average method, and the weight is βiAnd measuring the contribution rate of different fusion characteristic information matrixes.
In this embodiment, the correction decision coefficient of the final detection model of the trace acetylene gas concentration is greater than 0.95.
The infrared absorption characteristics of acetylene gas in the time domain and the frequency domain are explained in detail below with reference to fig. 3 to 6:
FIG. 3 is a standard absorption spectrum of acetylene gas in the wavelength range of 1529.5-1532.5 nm.
FIG. 4 is a power spectrum of a standard absorption spectrum of acetylene gas in a wavelength range of 1529.5-1532.5 nm, which is obtained by Fourier transform of FIG. 3.
FIG. 5 is an absorption spectrum of a trace acetylene gas sample with a concentration of 100-900 ppm in a wavelength range of 1529.5-1532.5 nm. In fig. 5, the matte burrs in the absorption spectrum together with the fluctuating spectral background constitute spectral background noise.
FIG. 6 is a power spectrum of an absorption spectrum of a trace acetylene gas sample with a concentration of 100-900 ppm in a wavelength range of 1529.5-1532.5 nm. As can be seen from FIG. 6, the spectral characteristics of the power spectrum of the absorption spectrum of the trace acetylene gas sample with the concentration of 100-900 ppm in the wavelength range of 1529.5-1532.5 nm are the same, i.e. the peaks appear at the same frequency components, and the amplitude of the characteristic absorption component has a linear relationship with the acetylene concentration. As shown in FIG. 6, the absorption characteristic components of acetylene gas are mainly and intensively distributed in the middle-low frequency range of 0-40 Hz, wherein the absorption characteristic components are relatively obvious at 12Hz, 15Hz, 18Hz, 23Hz, 29Hz and 34 Hz. And the linearity of the amplitude of the characteristic absorption component and the acetylene concentration at 2Hz, 4Hz and 6Hz of the low frequency band is low, mainly because these three frequency components are affected by 1/f noise. Since the linearity of the concentration calibration curves of the 18Hz and 23Hz frequency components is higher than that of the concentration calibration curves of other frequency components, and the sensitivity of the concentration calibration curves of the 18Hz and 23Hz frequency components is higher than that of the concentration calibration curves of other frequency components, the embodiment extracts 18Hz and 23Hz as characteristic frequency information.
In this embodiment, the final detection model of the concentration of the trace acetylene gas proposed in this embodiment is also compared with a detection model of the concentration of the trace acetylene gas constructed based on single spectral data and a detection model of the concentration of the trace acetylene gas constructed based on single power spectral data, and the comparison results are shown in table 1:
TABLE 1 comparison of three trace acetylene gas concentration detection models
As can be seen from table 1, both the detection accuracy and the minimum detection limit of the final detection model for the concentration of trace acetylene gas provided by this embodiment are significantly better than those of the detection model for the concentration of trace acetylene gas constructed based on single spectral data and those of the detection model for the concentration of trace acetylene gas constructed based on single power spectral data. The result shows that the final detection model for the concentration of the trace acetylene gas, which is provided by the embodiment, can effectively remove noise interference, well retains the high-sensitivity characteristic vector related to the concentration of acetylene, and has high detection sensitivity and accuracy.
Although the invention herein has been described with reference to particular embodiments, it is to be understood that these embodiments are merely illustrative of the principles and applications of the present invention. It is therefore to be understood that numerous modifications may be made to the illustrative embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that features described in different dependent claims and herein may be combined in ways different from those described in the original claims. It is also to be understood that features described in connection with individual embodiments may be used in other described embodiments.
Claims (10)
1. The trace acetylene gas concentration detection method based on the fusion of the spectral and power spectral features is characterized by comprising the following steps of:
s1, preparing a preset number of trace acetylene gas samples with different concentrations;
s2, collecting the original infrared spectrum of each trace acetylene gas sample;
s3, preprocessing the original infrared spectrum of each trace acetylene gas sample;
s4, extracting a plurality of characteristic wavelength information related to acetylene concentration in the infrared spectrum of each trace acetylene gas sample after pretreatment to obtain a characteristic wavelength information matrix formed by the characteristic wavelength information;
s5, obtaining a power spectrum of an original infrared spectrum of each trace acetylene gas sample;
s6, extracting a plurality of characteristic frequency information related to acetylene concentration in a power spectrum of an original infrared spectrum of each trace acetylene gas sample to obtain a characteristic frequency information matrix formed by the characteristic frequency information;
s7, fusing the characteristic wavelength information matrix and the characteristic frequency information matrix of each trace acetylene gas sample to obtain a fused characteristic information matrix;
s8, repeatedly executing the steps S2 to S7 until the number of the fusion characteristic information matrixes of each trace acetylene gas sample reaches a preset value;
s9, dividing a plurality of fusion characteristic information matrixes of each trace acetylene gas sample into a training sample and a correction sample;
s10, establishing a trace acetylene gas concentration initial detection model by adopting each fusion characteristic information matrix as a training sample and the known concentration information of the trace acetylene gas sample corresponding to the fusion characteristic information matrix;
s11, correcting the initial detection model of the concentration of the trace acetylene gas by adopting each fusion characteristic information matrix as a correction sample and the known concentration information of the trace acetylene gas sample corresponding to the fusion characteristic information matrix to obtain a final detection model of the concentration of the trace acetylene gas;
and S12, collecting the acetylene gas sample to be detected, obtaining a fusion characteristic information matrix of the acetylene gas sample, and inputting the fusion characteristic information matrix into a final detection model of the concentration of the trace acetylene gas to obtain the acetylene concentration information of the acetylene gas sample.
2. The method for detecting the concentration of the trace acetylene gas based on the fusion of the spectral features and the power spectral features as claimed in claim 1, wherein the step S1 of preparing the predetermined parts of trace acetylene gas samples with different concentrations specifically comprises:
preparing 9 parts of trace acetylene gas samples, wherein the acetylene concentrations of the 9 parts of trace acetylene gas samples are 100-900 ppm respectively, and the concentration gradient is 100 ppm.
3. The method for detecting the concentration of the trace acetylene gas based on the fusion of the spectral features and the power spectral features as claimed in claim 2, wherein the step S2 is to collect the original infrared spectrum of each trace acetylene gas sample specifically as follows:
under the environment condition of standard atmospheric pressure and room temperature of 26 ℃, an infrared spectrometer is used for collecting the absorption spectrum of each trace acetylene gas sample within the wavelength range of 1529.5-1532.5 nm.
4. The method for detecting the concentration of the trace acetylene gas based on the fusion of the spectral features and the power spectral features as claimed in claim 3, wherein the step S3 of preprocessing the original infrared spectrum of each trace acetylene gas sample specifically comprises:
and performing wavelet denoising, multivariate scattering correction and standard normal variable correction on the original infrared spectrum of each trace acetylene gas sample.
5. The method for detecting the concentration of trace acetylene gas based on the fusion of spectral and power spectral features according to claim 4, wherein in step S4, a principal component analysis method is used to extract a plurality of characteristic wavelength information related to acetylene concentration in the preprocessed infrared spectrum of each trace acetylene gas sample.
6. The method for detecting the concentration of the trace acetylene gas based on the fusion of the spectral features and the power spectral features as claimed in claim 5, wherein the step S5 is to perform Fourier transform on the original infrared spectrum of each trace acetylene gas sample to obtain the corresponding power spectrum.
7. The method for detecting the concentration of trace acetylene gas based on the characteristic fusion of the spectrum and the power spectrum as claimed in claim 6, wherein the characteristic frequency information related to the acetylene concentration in the power spectrum of the original infrared spectrum of each trace acetylene gas sample extracted in the step S6 is 18Hz and 23 Hz.
8. The trace acetylene gas concentration detection method based on the fusion of spectral and power spectral features according to claim 7, wherein the number of the fusion feature information matrixes per trace acetylene gas sample is 8 when step S8 is executed;
in step S9, 5 fusion characteristic information matrices are randomly selected from 8 fusion characteristic information matrices of each trace acetylene gas sample as training samples, and the remaining 3 fusion characteristic information matrices are used as correction samples.
9. The method for detecting the concentration of the trace acetylene gas based on the fusion of the spectral features and the power spectral features as claimed in claim 8, wherein the step S10 is to establish an initial detection model of the concentration of the trace acetylene gas based on a partial least squares algorithm.
10. The method for detecting the concentration of trace acetylene gas based on the fusion of spectral features and power spectral features of claim 9, wherein the correction decision coefficient of a final detection model of the concentration of trace acetylene gas is greater than 0.95.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911248926.1A CN111007033B (en) | 2019-12-09 | 2019-12-09 | Trace acetylene gas concentration detection method based on spectrum and power spectrum feature fusion |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911248926.1A CN111007033B (en) | 2019-12-09 | 2019-12-09 | Trace acetylene gas concentration detection method based on spectrum and power spectrum feature fusion |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111007033A true CN111007033A (en) | 2020-04-14 |
CN111007033B CN111007033B (en) | 2022-08-30 |
Family
ID=70114264
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911248926.1A Active CN111007033B (en) | 2019-12-09 | 2019-12-09 | Trace acetylene gas concentration detection method based on spectrum and power spectrum feature fusion |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111007033B (en) |
Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0257004A (en) * | 1988-08-23 | 1990-02-26 | Matsushita Electric Ind Co Ltd | Control device |
US5305076A (en) * | 1991-02-16 | 1994-04-19 | Horiba, Ltd. | Quantitative analytical method and apparatus for spectrometric analysis using wave number data points |
US5351198A (en) * | 1991-02-15 | 1994-09-27 | Horiba, Ltd. | Quantitative analytical method and apparatus for determining a plurality of ingredients with spectrometric analysis |
HU9800510D0 (en) * | 1998-03-09 | 1998-04-28 | Ladanyi Jozsef | Method and device for detecting colour vision deficiencies based on critical fusion frequency spectral scanning |
US6894297B1 (en) * | 1999-06-04 | 2005-05-17 | Horiba, Ltd. | Multi-component gas analyzing method using FTIR |
US20120065948A1 (en) * | 2010-09-13 | 2012-03-15 | Mks Instruments, Inc. | Monitoring, Detecting and Quantifying Chemical Compounds in a Sample |
WO2013011253A1 (en) * | 2011-07-15 | 2013-01-24 | The Secretary Of State For Defence | Method and apparatus for gas monitoring and detection |
US20130054603A1 (en) * | 2010-06-25 | 2013-02-28 | U.S. Govt. As Repr. By The Secretary Of The Army | Method and apparatus for classifying known specimens and media using spectral properties and identifying unknown specimens and media |
GB201419267D0 (en) * | 2014-10-29 | 2014-12-10 | Univ Aston | Methods and devices incorporating surface nanoscale axial photonics |
US20170059477A1 (en) * | 2015-08-03 | 2017-03-02 | Spectrasensors, Inc. | Reconstruction of frequency registration for quantitative spectroscopy |
CN108195791A (en) * | 2017-12-18 | 2018-06-22 | 武汉中科志康生物科技有限公司 | It is a kind of that based on too core, hereby technology detects pesticide residue and the method for illegal additive |
CN109409350A (en) * | 2018-10-23 | 2019-03-01 | 桂林理工大学 | A kind of Wavelength selecting method based on PCA modeling reaction type load weighting |
CN109766909A (en) * | 2018-11-29 | 2019-05-17 | 温州大学 | The micro- ageing of plastics behavior analytic method of coastal environment based on spectrogram fusion |
-
2019
- 2019-12-09 CN CN201911248926.1A patent/CN111007033B/en active Active
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0257004A (en) * | 1988-08-23 | 1990-02-26 | Matsushita Electric Ind Co Ltd | Control device |
US5351198A (en) * | 1991-02-15 | 1994-09-27 | Horiba, Ltd. | Quantitative analytical method and apparatus for determining a plurality of ingredients with spectrometric analysis |
US5305076A (en) * | 1991-02-16 | 1994-04-19 | Horiba, Ltd. | Quantitative analytical method and apparatus for spectrometric analysis using wave number data points |
HU9800510D0 (en) * | 1998-03-09 | 1998-04-28 | Ladanyi Jozsef | Method and device for detecting colour vision deficiencies based on critical fusion frequency spectral scanning |
US6894297B1 (en) * | 1999-06-04 | 2005-05-17 | Horiba, Ltd. | Multi-component gas analyzing method using FTIR |
US20130054603A1 (en) * | 2010-06-25 | 2013-02-28 | U.S. Govt. As Repr. By The Secretary Of The Army | Method and apparatus for classifying known specimens and media using spectral properties and identifying unknown specimens and media |
US20120065948A1 (en) * | 2010-09-13 | 2012-03-15 | Mks Instruments, Inc. | Monitoring, Detecting and Quantifying Chemical Compounds in a Sample |
WO2013011253A1 (en) * | 2011-07-15 | 2013-01-24 | The Secretary Of State For Defence | Method and apparatus for gas monitoring and detection |
GB201419267D0 (en) * | 2014-10-29 | 2014-12-10 | Univ Aston | Methods and devices incorporating surface nanoscale axial photonics |
US20170059477A1 (en) * | 2015-08-03 | 2017-03-02 | Spectrasensors, Inc. | Reconstruction of frequency registration for quantitative spectroscopy |
CN108195791A (en) * | 2017-12-18 | 2018-06-22 | 武汉中科志康生物科技有限公司 | It is a kind of that based on too core, hereby technology detects pesticide residue and the method for illegal additive |
CN109409350A (en) * | 2018-10-23 | 2019-03-01 | 桂林理工大学 | A kind of Wavelength selecting method based on PCA modeling reaction type load weighting |
CN109766909A (en) * | 2018-11-29 | 2019-05-17 | 温州大学 | The micro- ageing of plastics behavior analytic method of coastal environment based on spectrogram fusion |
Non-Patent Citations (7)
Title |
---|
XI CHEN: "A Simple Algorithm for C2H2 Real-Time", 《IEEE PHOTONICS TECHNOLOGY LETTERS》 * |
余段辉: ""基于红外光谱的气体识别与浓度检测算法研究"", 《中国优秀博硕士学位论文全文数据库(硕士) 信息科技辑,2018年第8期》 * |
刘燕德: "基于GA-LSSVM的苹果糖度近红外光谱检测", 《西北农林科技大学学报》 * |
田博: "基于协变光谱随机场理论的激光引信背景模型", 《探测与控制学报》 * |
秦善知: "基于便携式近红外光谱仪检测梨可溶性固形物", 《江苏农业科学》 * |
郁敏捷: ""基于傅里叶变换的差分吸收光谱法测量NH3和"", 《中国激光》 * |
郑林: ""基于近红外光谱吸收原理的气体传感研究"", 《中国优秀博硕士学位论文全文数据库(硕士) 信息科技辑,2012年第8期》 * |
Also Published As
Publication number | Publication date |
---|---|
CN111007033B (en) | 2022-08-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103105441B (en) | Qualitative and quantitative analysis method for sulfur hexafluoride (SF6) gas discharge decomposition product | |
CN108181266B (en) | TD L AS gas concentration detection method | |
CN107643252B (en) | Real-time background-deduction nonlinear correction method for WMS detection of oxygen concentration in bottle | |
CN101105446A (en) | Differential optical absorption spectroscopy air quality detection system | |
CN105486663B (en) | A method of detecting the stable carbon isotope ratio of soil using near infrared spectrum | |
CN111272696A (en) | Method for rapidly detecting essence doped in Pu' er tea | |
CN109557045A (en) | It is a kind of for extracting the approximating method of direct absorption line in TDLAS technology | |
CN103743703A (en) | Method for detecting main components in tea leaves by adopting near infrared spectrum | |
CN110736719A (en) | gas concentration calibration and measurement method based on TDLAS direct absorption spectrum | |
CN106716109A (en) | Estimation of water interference for spectral correction | |
Griffiths et al. | Self-weighted correlation coefficients and their application to measure spectral similarity | |
CN113310922A (en) | NO concentration measuring system based on ultraviolet differential absorption spectrum and concentration calculating method | |
CN115728259A (en) | Multi-component gas analysis method | |
CN111007033B (en) | Trace acetylene gas concentration detection method based on spectrum and power spectrum feature fusion | |
CN111707728A (en) | Method for identifying white peony tea with different grades based on HS-PTR-TOF-MS | |
CN109709060B (en) | Method for measuring asphalt softening point, penetration degree and mass loss | |
CN105784629B (en) | The method that the stable carbon isotope ratio of soil is quickly detected using middle infrared spectrum | |
CN105699314B (en) | A method of detecting soil stabilization carbon isotope ratio using middle infrared spectrum | |
CN106644980A (en) | Method for detecting oil content in sewage | |
CN106323888B (en) | Ultra-low emission flue gas measuring method | |
CN103163100B (en) | Gaseous element mercury detection method | |
CN110346346B (en) | Raman gas detection method based on compressed sensing correlation algorithm | |
CN105784630B (en) | A kind of silicon rubber glue detection method of content based on infrared spectrum analysis | |
Cheng et al. | Recognition of MIR data of semen armeniacae amarum and semen persicae using discrete wavelet transformation and BP-artificial neural network | |
CN111220571B (en) | Second harmonic signal fitting method and system based on amplitude dispersion |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |