CN108444976B - A kind of heating value of natural gas measurement method based on Raman spectrum - Google Patents
A kind of heating value of natural gas measurement method based on Raman spectrum Download PDFInfo
- Publication number
- CN108444976B CN108444976B CN201810382163.9A CN201810382163A CN108444976B CN 108444976 B CN108444976 B CN 108444976B CN 201810382163 A CN201810382163 A CN 201810382163A CN 108444976 B CN108444976 B CN 108444976B
- Authority
- CN
- China
- Prior art keywords
- spectrum
- peak
- gas
- natural gas
- raman
- 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.)
- Active
Links
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 title claims abstract description 184
- 239000003345 natural gas Substances 0.000 title claims abstract description 80
- 238000010438 heat treatment Methods 0.000 title claims abstract description 37
- 238000001237 Raman spectrum Methods 0.000 title claims abstract description 34
- 238000000691 measurement method Methods 0.000 title claims abstract description 15
- 238000001228 spectrum Methods 0.000 claims abstract description 107
- 239000007789 gas Substances 0.000 claims abstract description 92
- 238000001069 Raman spectroscopy Methods 0.000 claims abstract description 50
- 238000000034 method Methods 0.000 claims abstract description 39
- 230000003595 spectral effect Effects 0.000 claims abstract description 39
- 238000004458 analytical method Methods 0.000 claims abstract description 24
- 238000001514 detection method Methods 0.000 claims abstract description 13
- 238000012549 training Methods 0.000 claims description 42
- 238000010606 normalization Methods 0.000 claims description 36
- 238000005259 measurement Methods 0.000 claims description 27
- 238000004817 gas chromatography Methods 0.000 claims description 13
- 239000000203 mixture Substances 0.000 claims description 13
- 229930195733 hydrocarbon Natural products 0.000 claims description 11
- 150000002430 hydrocarbons Chemical class 0.000 claims description 11
- 239000004215 Carbon black (E152) Substances 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 6
- 239000000470 constituent Substances 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 2
- 150000001335 aliphatic alkanes Chemical class 0.000 abstract description 6
- 239000008246 gaseous mixture Substances 0.000 abstract description 5
- 238000004422 calculation algorithm Methods 0.000 abstract description 3
- 239000001273 butane Substances 0.000 abstract description 2
- 238000000354 decomposition reaction Methods 0.000 abstract description 2
- IJDNQMDRQITEOD-UHFFFAOYSA-N n-butane Chemical compound CCCC IJDNQMDRQITEOD-UHFFFAOYSA-N 0.000 abstract description 2
- OFBQJSOFQDEBGM-UHFFFAOYSA-N n-pentane Natural products CCCCC OFBQJSOFQDEBGM-UHFFFAOYSA-N 0.000 abstract description 2
- 239000000523 sample Substances 0.000 description 73
- 238000012360 testing method Methods 0.000 description 8
- 238000003841 Raman measurement Methods 0.000 description 6
- 238000004587 chromatography analysis Methods 0.000 description 5
- 238000004566 IR spectroscopy Methods 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 3
- 239000000835 fiber Substances 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000000737 periodic effect Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 239000002737 fuel gas Substances 0.000 description 2
- 238000004868 gas analysis Methods 0.000 description 2
- 239000012535 impurity Substances 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000005033 Fourier transform infrared spectroscopy Methods 0.000 description 1
- 238000003705 background correction Methods 0.000 description 1
- 239000012159 carrier gas Substances 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 239000000567 combustion gas Substances 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 239000005350 fused silica glass Substances 0.000 description 1
- 239000011261 inert gas Substances 0.000 description 1
- 230000001678 irradiating effect Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012067 mathematical method Methods 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 230000003472 neutralizing effect Effects 0.000 description 1
- 239000003209 petroleum derivative Substances 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 238000005057 refrigeration Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
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
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/63—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light optically excited
- G01N21/65—Raman scattering
Landscapes
- Health & Medical Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The heating value of natural gas measurement method based on Raman spectrum that the invention discloses a kind of.In natural gas may comprising a small amount of butane and its more than alkane (conjunction is denoted as particular components C4 +).The present invention utilizes a kind of spectral decomposition algorithm, the Raman spectrum of any native gas gaseous mixture is decomposed into the weighted sum form of known pure component spectrum and several peak functions, wherein several peak functions describe particular components C4 +Spectral components；Based on each pure component and particular components C decomposited4 +Spectrum, establish the functional relation between calorific value and relative spectral area；By the natural gas spectral resolution of unknown calorific value it is the spectral components of each basic component and several peak functions, and calculates the ratio of each group spectral area Yu methane area under spectrum, finds out gas heating value using the heating value of natural gas Raman analysis model of foundation.The method of the present invention has many advantages, such as that detection speed is fast, accuracy is higher, reproducible, non-contact, live non-maintaining with sample gas.
Description
Technical field
The invention belongs to Chemical Measurement fields, are related to a kind of admixture of gas heat decomposed based on gaseous mixture Raman spectrum
Value measurement method especially contains more multi-component heating value of natural gas measurement method.
Background technique
With the sustainable growth of China's natural gas usage amount, gas metering is an important link.Gas metering
Method can be divided into volume metering and energy meter method, and China uses volume metering method substantially at present.Not due to neutralizing gas
Together, there may be larger differences for gas component, lead to the heating value of natural gas significant difference of different compositions, using simple volume
Measurement Law is lost just.And energy meter method is used for the calorific value that measurement unit volume natural gas contains, and directly embodies natural gas
Value, can eliminate deficiency existing for volume metering method.
Directly measurement and measurement two major classes indirectly are broadly divided into for heating value of natural gas measurement at present.The direct method of measurement passes through
The heat that measurement fuel gas buring generates determines calorific value.In standard GB/T/T 12206-2006, city gas calorific value is using appearance
Gram formula flow type calorimeter measures, and using the water flow of continuous-stable absorbs the heat that combustion gas completely burned releases, according to reaching
Parameters when to stable state calculate the fuel gases calorific value under metering reference condition.The standard is less than suitable for higher calorific value
62.8MJ/m3City gas, test permission repetition relative error be 1%.But measurement device includes that equipment is various, behaviour
Make complicated and more demanding for determination of the environment, it is difficult to be applied to on-line checking field.
The indirect method of measurement usually first detects the component of natural gas, then its calorific value is calculated.China is at present for natural
Gas calorific value measures mainly based on gas chromatography, and has issued standard GB/T/T 13610-2014 " composition of natural gas
Analytical gas chromatography "；The each component concentration obtained using measurement, according to standard GB/T 11062-2014, " natural gas generates heat
Amount, density, the calculation method of relative density and wobbe index " in formula the calorific value of natural gas is calculated.Gas chromatography
Multicomponent gas can be measured simultaneously, and accuracy is higher, for natural gas main component, the repetition deviation of permission 0.20% it
It is interior.However, there are analytical cycle length, the multiple chromatographic columns of needs and carrier gas, needs periodically to mark for gas chromatography for on-line checking
The limitation such as fixed, periodic maintenance.
Patent CN 201120375483.5 proposes a kind of measurement method based on Fourier transform infrared spectroscopy, and real
Specific measuring device is showed.The patent carries out constituent analysis to natural gas using infrared spectroscopy, then calculates with mathematical method
Its calorific value.This method can be realized to heating value of natural gas nondestructive measurement, but be limited by infrared spectrometry principle, this method without
Method measures homonuclear diatomic molecule, including H2、N2, moreover, steam can generate very big interference to measurement.
Raman spectrum is as a kind of novel detection method, the problem of can effectively solve the problem that infrared spectrometry.Patent
CN201120165694.6 proposes a kind of petroleum gas gas analyzer based on Raman spectrum, is used for oil-gas exploration process
The detection of the component of middle hydro carbons or non-hydrocarbon gases, measurement accuracy are ± the 0.25% of full scale.Sample gas is introduced into sharp by this method
In optical cavity, excites the Raman signal of generation to pass through the copped wave optical filter in 8 channels, 8 kinds of components can be detected simultaneously.The party
Method high sensitivity, but it is high to require (including steam) to require the impurity of sample, and a small amount of impurity will significantly reduce Raman analysis
Precision and efficiency.In addition, this method is for independent gas peak (such as CO, N2) there is very high detection accuracy, but working as has
When the case where component spectra is overlapped occurs, method failure.
J Kiefer et al. has designed and developed a set of LR laser raman analysis system [J Kiefer, T for gas analysis
Seeger,S Steuer,S Schorsch,et al.Design and Characterization of a Raman-
Sattering-Based Sensor System for Temporally Resolved Gas Analysis and its
Application in a Gas Tubine Power Plant[J].Measurement science&Technology,
2008,19 (8): 1-9.], laser is separated with gas measurement room with fused silica glass, biogas and natural gas are determined
Analysis, measurement result and chromatography have good consistency.However, by the limit of existing laser Raman spectroscopy detection sensitivity
System, cannot detect the lower various hydrocarbon components of concentration.For the practical natural gas comprising a small amount of heavy hydrocarbon component, heat is calculated
It is worth less than normal.
Summary of the invention
The shortcomings that present invention is measured for existing gas chromatography and Raman spectroscopy are difficult to detect low concentration component
Deficiency proposes a kind of new heating value of natural gas measurement method based on Raman spectrum, natural gas Raman spectrum is decomposed into respectively
Pure component (CH4、C2H6、C3H6、CO2、N2、H2, CO) spectrum and several peak function weighted sums form, wherein several peak function tables
Show particular components C4 +Spectrum, and the functional relation between relative peak area and heating value of natural gas is established, using actually measured
Natural gas Raman spectrum passes through its actual calorific value of the heating value of natural gas Raman analysis model prediction of foundation.
In view of the above-mentioned deficiencies in the prior art, it is an object of the present invention to provide a kind of heating value of natural gas survey based on Raman spectrum
Amount method, this method comprises the following steps:
(1) acquisition and pretreatment of natural gas correlation pure component Raman spectrums, specifically includes following sub-step:
(1.1) the main pure component gas for including in natural gas is CH4、C2H6、C3H6、CO2、N2、H2,CO；Obtain natural gas
The original spectrum of pure component；
(1.2) background spectrum deduction, baseline correction and area normalization are carried out to the original spectrum of pure component, obtains pure group
Divide normalization spectrum；
(2) foundation of heating value of natural gas Raman analysis model specifically includes following sub-step:
(2.1) the natural gas aggregate sample gas of several typical case's compositions is prepared as training sample gas, obtains each trained sample gas
Calorific value；Raman spectroscopy measurement is carried out to training sample gas, obtains training sample gas original spectrum；
(2.2) background spectrum deduction, baseline correction and area normalization are carried out to training sample gas original spectrum, is trained
Sample gas normalizes spectrum；
(2.3) CH will be removed in training sample gas4、C2H6、C3H6、CO2、N2、H2, that other hydrocarbon components outside CO are expressed as is special
Component C4 +, training sample gas normalization spectrum Y (v) is decomposed into the weighted sum of known pure component normalization spectrum and N number of peak function
Form, wherein the weighted sum form of several peak functions describes particular components C4 +Spectral components, it may be assumed that
Wherein, α=(α1,…,α7)T, αiIndicate the model parameter of pure component known to i-th kind, Pi(v) i-th kind pure group is indicated
Divide normalization spectrum, f (v, βj) indicate j-th of peak function (such as Gaussian function, Lorentz peak function etc.), βjIndicate j-th of peak
Function parameter；The acquisition modes of peak function number N are as follows: be fitted training sample gas normalization spectrum Y (v) first with pure component and obtain
Normalization spectrum R after fitting, then by Y (v)-R highest point add a peak function, recycle the peak function of addition with it is pure
Component is fitted training sample gas normalization spectrum Y (v), the normalization spectrum R after being fitted together, and continuous iteration is missed until optimization
Difference is less than setting value or N reaches maximum set value；
(2.4) for k-th of training sample, pure component C is sought2H6、C3H8、CO2、N2、H2, CO modeling region in light
Area under spectrum { Sj(k), j=2 ..., 7 } relative to methane spectra area S1(k) ratio Rj(k):
For particular components C4 +, the deformation peak C-H is chosen as its characteristic peak, seeks C-H deformation peak area relative to methane
Spectrum area S1(k) ratio S8(k), here it is possible to which choosing the C-H deformation area Feng Pu is 1420-1540cm-1；
(2.5) the calorific value area under spectrum opposite with each component based on whole training samples, establishes heating value of natural gas Raman analysis
Model:
WhereinFor the calorific value of k-th of training sample, aj, b indicate model parameter, obtained using linear regression；
(3) by the original spectrum of the gas samples of calorific value to be measured, each component is obtained according to step (2.2)-(2.4)
Area ratio of the spectrum area relative to methane spectra substitutes into the heating value of natural gas Raman analysis of foundation using area ratio as input
In model, the calorific value of the sample is obtained.
Further, in the step (1.2), area normalization is carried out in modeling region to each pure component original spectrum
Processing, modeling region are 500-2500cm-1。
Further, in the step (1.2), full constituent detection is carried out to training sample with gas chromatography, and utilize
The calorific value of each training sample is calculated in standard GB/T/T 11062-2014.
Further, in the step (2.3), the peak function is selected from Lorentz peak function, it may be assumed that
Wherein, βj=(Sj cj wj)T, v expression wave number, SjFor the area of j-th of Lorentz spectral peak, cjFor j-th of long-range navigation
The hereby center wave number of spectral peak, wjFor the Lorentz half-breadth of j-th of Lorentz spectral peak.
Further, Lorentz spectral peak number N mends peak method by following iteration and obtains:
(a) do not consider C first first4 +Component optimizes pure component coefficient, so that:
Wherein,Y (v) is that training sample gas normalizes spectrum, [v1,v2] it is spectral resolution model
It encloses；
(b) training sample gas normalization spectrum Y (v) is subtracted into R (v, α) as error spectrum Yc4(v), Y is selectedc4(v) spectrogram
At middle highest point, a Lorentz peak is added, and to pure component factor alpha=(α1,…,α7)TWith Lorentz peak parameter beta=(S1 c1
w1) optimize again, so that
(c) training sample gas normalization spectrum Y (v) is subtracted into fit-spectra R (v, α, β) as error spectrum again, selects it
At middle highest point, then a Lorentz peak is added, at this time Lorentz peak parameter beta=(S1 c1 w1 S2 c2 w2), optimization is joined again
Number, such iteration mend peak, until the Lorentz peak height or area that meet maximum number of iterations or addition are less than error.
The present invention effectively can overcome gas chromatography to need periodic cleaning as a kind of novel Raman measurement of caloric value method
The long disadvantage of maintenance, measurement period, and have many advantages, such as it is reproducible, it is non-contact with sample gas；And in contrast to existing Raman
Measurement method, measurement method simple and flexible of the present invention, for the hydrocarbon component of low concentration, as an alanysis, reasonably
The deformation peak C-H is chosen to indicate its influence to calorific value, improves the accuracy of measurement.Meanwhile also in sample gas contain unknown group
The Raman analysis divided provides a kind of thinking.It is embodied in:
1, natural gas is divided into 8 kinds of component (CH by this method4、C2H6、C3H6、CO2、N2、H2、CO、C4 +), content is lower
Butane more than a variety of hydrocarbon components be merged into one kind and analyzed (C4 +Component).
2, it for the obtained original Raman spectrum of natural gas of detection, is first pre-processed, then by the normalizing of gas samples
Change the weighted sum form that spectral resolution is known pure component spectrum and several peak functions, wherein the weighted sum shape of several peak functions
Formula describes particular components C4 +Spectral components.
3, according to each spectral components after decomposition, with each pure component spectrum relative to methane spectra area ratio come between
The reversed influence for reflecting each pure component to calorific value；And for C4 +Component spectra is then chosen the deformation peak C-H as its characteristic peak, is come comprehensive
Close reflection C4 +Influence of the component to calorific value.
4, according to each pure component and C4 +The opposite Raman peak area of component is based on established heating value of natural gas Raman analysis
Model, it is predictable to obtain the calorific value of any native gas sample gas.
Detailed description of the invention
The Raman spectrum measurement system that Fig. 1 present invention uses；
Fig. 2 standard sample gas 1#, 2# and 3# original spectrum；
Fig. 3 standard sample gas 4# and Ar original spectrum；
Standard sample gas Raman spectrum after Fig. 4 pretreatment；
Raman spectrum after the pretreatment of Fig. 5 pure component modeling sample；
The area normalization Raman spectrum of Fig. 6 natural gas main component；
The natural original Raman spectrum of gas sample gas 3# of Fig. 7；
Raman spectrum after the natural gas sample gas 3# pretreatment of Fig. 8；
The natural gas sample gas 3# area normalization Raman spectrum of Fig. 9；
All trained sample gas of Figure 10 normalize Raman spectrum；
The natural gas sample gas 3# Raman spectrum of Figure 11 decomposes；
The partial enlargement of Figure 12 whole sample gas unknown component spectrum；
Figure 13 heating value of natural gas Raman models fitting result；
All test specimens gas of Figure 14 normalize Raman spectrum；
The Raman spectroscopy of Figure 15 natural gas test specimens gas Lower heat value is compared with chromatography.
Specific embodiment
Invention is further described in detail in the following with reference to the drawings and specific embodiments.
The Raman test macro that the present embodiment uses is as shown in Figure 1, major optical component are as follows: laser, optical fiber, Raman are visited
Head and spectrometer；The laser is the laser that central wavelength is 532nm；The Raman probe is that 532nm fiber Raman is visited
Head；The spectrometer is the TEC refrigeration fiber spectrometer of Ocean Optics production.
The measurement process of the system is as follows: natural gas sample gas is passed through in closed sampling pipe；Laser issues laser, warp
It crosses excitation fiber to conduct to Raman probe, the irradiating sample after Raman probe focuses, the Raman diffused light of generation is by Raman probe
It collects, passes spectrometer back using optical fiber is collected；Raman diffused light turns by spectrometer grating beam splitting, ccd array detection, A/D
After changing, spectral digital signal is transmitted to PC machine.
Heating value of natural gas measurement method of the present invention based on Raman spectrum, comprising the following steps:
(1) acquisition and pretreatment of natural gas correlation pure component Raman spectrums, specifically includes following sub-step:
(1.1) the main pure component gas for including in natural gas is CH4、C2H6、C3H6、CO2、N2、H2, CO, to obtain these
Pure component spectrum, matches sample preparation gas, and concrete composition is as shown in table 1.Using Raman Measurement system shown in FIG. 1 to above-mentioned sample gas and
Inert gas Ar is measured, and original spectrum difference is as shown in Figure 2,3.
Table 1 is used to obtain the standard sample gas composition of pure component Raman spectrums
(1.2) sample gas spectrum is directly subtracted into the background spectrum under identical integral condition, it can background correction spectrum；It adopts again
With iteration polynomials fit-spectra baseline, subtracting baseline can be obtained the sample gas spectrum after correction, specific as shown in Figure 4.
Although being located at 2500-3500cm-1C-H spectral peak of stretching it is very strong, but the peak overlap of different alkane molecule is serious；And
Positioned at 500-2000cm-1Fingerprint region Raman signal although weaker, different alkane molecule spectral peaks are relatively independent.In view of H2It is special
Sign peak is located at 590cm-1Near, therefore, select 500-2500cm-1Area is composed as modeling.
Because of N2Raman peaks are independent, for containing N2Above-mentioned hydro carbons sample gas, deduct N2Pure component C can be obtained behind peak2H6、
C3H6Spectrum；And the above-mentioned sample gas for not containing hydrocarbon molecules, due to CO2、N2、H2, the Raman peaks of CO it is independent, can directly mention
Take the spectrum of these components.7 kinds of pure component Raman spectrums are as shown in Figure 5 after pretreatment.Area is carried out in modeling spectrum area to return
One changes, and thus obtained main component normalization spectrum is as shown in Figure 6.
(2) foundation of heating value of natural gas Raman analysis model specifically includes following sub-step:
(2.1) the natural gas mixing sample of several typical case's compositions is prepared as training sample.With gas chromatography to instruction
Practice sample and carry out full constituent detection, and utilizes standard GB/T 11062-2014 " calorific capacity of natural gas, density, relative density and fertile
Moor the calculation method of index " in formula the calorific value of each sample is calculated, the concrete composition and calorific value of training sample are shown in Table 2.
And above-mentioned sample gas is measured using Raman Measurement system shown in FIG. 1.
The composition and calorific value of 2 training sample of table
(2.2) background deduction, baseline correction and area normalization are carried out to above-mentioned sample gas Raman spectrum, specific method is synchronous
Suddenly (1.2).By taking training sample sample gas #3 as an example, the original spectrum measured is as shown in Figure 7；The pretreated spectrum of sample gas #3 is as schemed
Shown in 8, through the result of normalized as shown in figure 9, the preprocessed spectrum such as Figure 10 with after normalization of whole training samples
It is shown.
(2.3) training sample gas normalization light spectral factorization is normalized to the weighting of spectrum and N number of peak function for known pure component
And form, wherein the weighted sum form of several peak functions describes particular components C4 +Spectral components, specifically include the following steps:
(2.3.1) is here using Lorentz peak function as analytic function, it is assumed that a certain spectral mixture is represented by known
The weighted sum of pure component spectrum and N number of Lorentz spectral peak:
Wherein, α=(α1,…,αM)T, β=(S1 c1 w1 S2 c2 w2…SN cN wN)T, v expression wave number, M is known pure
The number (M=7) of component, αiIndicate the model parameter of pure component known to i-th kind, Pi(v) face of pure component known to i-th kind is indicated
Product normalization spectrum, N are the number of Lorentz spectral peak, SjFor the area of j-th of Lorentz spectral peak, cjFor j-th of Lorentz spectral peak
Center wave number, wjFor the Lorentz half-breadth of j-th of Lorentz spectral peak.
(2.3.2) for natural gas gaseous mixture spectral resolution problem equivalent in following non-linear least square parameter optimization
Problem:
Wherein, Y (v) is the gaseous mixture area normalization spectrum that actual measurement obtains, [v1,v2] it is spectral resolution range.This example
In, v1=500cm-1, v2=2500cm-1。
The solution of (2.3.3) gaseous mixture spectral resolution problem.Since unknown component may be contained in mixture, it is possible to have
Several Lorentz spectral peaks, and Lorentz spectral peak number N is unknown.It is proposed to this end that following iteration mends peak algorithm:
Step1: do not consider C first first4 +Component optimizes pure component coefficient, so that
Wherein,
Step2: normalization spectrum Y (v) is subtracted into pure component fit-spectra R (v, α) as error spectrum Yc4(v), it selects
Yc4(v) in spectrogram at highest point, a Lorentz peak is added, and to pure component factor alpha=(α1,…,αM)TJoin with Lorentz peak
Number β=(S1 c1 w1) optimize again, so that
Step3: normalization spectrum Y (v) is subtracted into fit-spectra R (v, α, β) as error spectrum again, selects wherein highest
At point, then a Lorentz peak is added, at this time Lorentz peak parameter beta=(S1 c1 w1 S2 c2 w2), Optimal Parameters again.Such as
This iteration mends peak, until the Lorentz peak height or area that meet maximum number of iterations or addition are less than error.
Still by taking above-mentioned sample gas 3# as an example, it is based on CH4、C2H6、C3H8、CO2、N2、H2, CO pure component spectrum, in combination with upper
The iteration for stating Lorentz spectral peak mends peak algorithm, and the spectrogram component decomposed is as shown in figure 11, uses mathematical description wherein containing
Unknown component spectrum.And for whole training samples, the unknown component spectrum decomposed is as shown in figure 12.
(2.4) for k-th of training sample, each pure component (C is sought2H6、C3H8、CO2、N2、H2, CO) spectrum area Sj
(k), j=2 ..., 7 relative to methane spectra area S1(k) ratio Rj(k), see formula (2)；And for C4The above hydrocarbon component
C4 +, choose 1420-1540cm-1C-H deformation peak as its characteristic peak, and seek face of the peak area relative to methane spectra
Product compares S8(k)。
(2.5) the calorific value area under spectrum opposite with each component based on whole training samples is established using statistical regression formula (3)
Heating value of natural gas Quantitative Analysis Model, fitting result analysis are shown in Table 3.1, table 3.2 and Figure 13.
3.1 natural gas Raman calorific value models fitting result of table
| b | k2 | k3 | k4 | k5 | k6 | k7 | k8 |
| 33.65 | 0.0301 | 0.0550 | -0.0711 | -0.1822 | -0.0557 | -0.1111 | 0.2009 |
3.2 natural gas Raman calorific value models fitting interpretation of result of table
(3) application of heating value of natural gas Raman analysis model
(3.1) for the gas samples of unknown calorific value, its original spectrum is measured using Raman system shown in Fig. 1.
(3.2) pretreatment and normalized are carried out to natural gas original spectrum, specific method is the same as (2.2), all tests
Normalization spectrum is as shown in figure 14 after natural gas sample gas pretreatment.
It (3.3) is the weighting of known pure component spectrum and several peak functions by the normalization light spectral factorization of gas samples
And form, specific method is the same as (2.3).
(3.4) according to the heating value of natural gas Raman analysis model having built up, each group spectral area is sought relative to first
The area ratio of alkane spectrum, in this, as input, the calorific value of Natural Gas Prediction sample, as a result such as table 4.
4 heating value of natural gas Raman analysis model prediction result of table
| Sample gas number | Raman method | Sample gas number | Raman method | Sample gas number | Raman method |
| 6# | 38.753 | 17# | 35.981 | 27# | 34.441 |
| 7# | 36.917 | 18# | 29.618 | 28# | 35.838 |
| 8# | 37.657 | 19# | 32.085 | 29# | 33.694 |
| 16# | 33.536 | 20# | 39.195 | 30# | 33.341 |
(4) Raman analysis model is compared with chromatography
The calorific value of each test sample is measured as comparative result with gas chromatography, and natural gas test sample gas-chromatography is complete
Composition analysis result is specifically shown in Table 5.The comparison result of Raman analysis method and gas chromatography is respectively such as table 6.1, table 6.2 and Figure 15
It is shown.
5 heating value of natural gas Raman analysis model measurement sample of table composition
6.1 natural gas Lower heat value Raman spectroscopy (unit: MJ/m compared with red, orange, green, blue, yellow (ROGBY) of table3)
| Sample gas number | Chromatography | Raman method | Error | Sample gas number | Chromatography | Raman method | Error |
| 1# | 38.560 | 38.753 | -0.194 | 7# | 32.003 | 32.085 | -0.082 |
| 2# | 37.292 | 36.917 | 0.375 | 8# | 39.228 | 39.195 | 0.033 |
| 3# | 37.618 | 37.657 | -0.039 | 9# | 34.754 | 34.441 | 0.313 |
| 4# | 33.299 | 33.536 | -0.237 | 10# | 36.067 | 35.838 | 0.230 |
| 5# | 36.122 | 35.981 | 0.141 | 11# | 33.860 | 33.694 | 0.167 |
| 6# | 30.077 | 29.618 | 0.459 | 12# | 33.176 | 33.341 | -0.166 |
6.2 heating value of natural gas Raman analysis model error of table compares (unit: MJ/m3)
It can be seen that novel heating value of natural gas Raman Measurement method proposed by the present invention, can effectively overcome gas-chromatography
The disadvantage that method needs periodic cleaning maintenance, measurement period long, and have many advantages, such as it is reproducible, it is non-contact with sample gas.And with
Existing Raman Measurement method is compared, measuring device simple and flexible of the present invention, and fourth lower for concentration and that type is more
Alkane and the above alkane component, are merged into one kind and are analyzed, and choose the deformation peak C-H reasonably to indicate its shadow to calorific value
It rings.
Application example shows: the heating value of natural gas measurement method detection accuracy with higher, for natural gas low level heat
The detection test problems of value, mean square error are about 0.238MJ/m3, the precision with the heating value of natural gas direct method of measurement is same
In level；In addition Raman spectroscopy has the characteristics that detection speed is fast, non-contact with sample gas, the method for the present invention is particularly suitable for day
The on-line checking of right gas calorific value.
Claims (5)
1. a kind of heating value of natural gas measurement method based on Raman spectrum, which is characterized in that this method comprises the following steps:
(1) acquisition and pretreatment of natural gas correlation pure component Raman spectrums, specifically includes following sub-step:
(1.1) natural gas pure component CH is obtained4、C2H6、C3H6、CO2、N2、H2, CO original spectrum；
(1.2) background spectrum deduction, baseline correction and area normalization are carried out to the original spectrum of pure component, obtains pure component and returns
One changes spectrum:
(2) foundation of heating value of natural gas Raman analysis model specifically includes following sub-step:
(2.1) the natural gas aggregate sample gas of several typical case's compositions is prepared as training sample gas, obtains the calorific value of each trained sample gas；
Raman spectroscopy measurement is carried out to training sample gas, obtains training sample gas original spectrum；
(2.2) background spectrum deduction, baseline correction and area normalization are carried out to training sample gas original spectrum, obtains training sample gas
Normalize spectrum；
(2.3) CH will be removed in training sample gas4、C2H6、C3H6、CO2、N2、H2, other hydrocarbon components outside CO be expressed as particular components
C4 +, training sample gas normalization spectrum Y (v) is decomposed into the weighted sum form of known pure component normalization spectrum and N number of peak function,
That is:
Wherein, α=(α1..., α7)T, αiIndicate the model parameter of pure component known to i-th kind, Pi(v) indicate that i-th kind of pure component is returned
One changes spectrum, f (v, βj) indicate peak function, βjIndicate jth peak function parameter；The acquisition modes of peak function number N are as follows: sharp first
It is fitted the normalization spectrum R after training sample gas normalization spectrum Y (v) is fitted with pure component, then by error in Y (v)-R
A peak function is added in highest point, and the peak function of addition is recycled to be fitted training sample gas normalization spectrum Y together with pure component
(v), the normalization spectrum R after being fitted, continuous iteration reach maximum set value less than setting value or N until optimization error；
(2.4) for k-th of training sample, pure component C is sought2H6、C3H8、CO2、N2、H2, CO modeling region in spectrum face
Product Sj(k), j=2 ..., 7, relative to methane spectra area S1(k) ratio Rj(k):
For particular components C4 +, the deformation peak C-H is chosen as its characteristic peak, seeks C-H deformation peak area relative to methane spectra
Area S1(k) ratio S8(k)；
(2.5) the calorific value area under spectrum opposite with each component based on whole training samples, establishes heating value of natural gas Raman analysis model:
WhereinFor the calorific value of k-th of training sample, aj, b indicate model parameter；
(3) by the original spectrum of the gas samples of calorific value to be measured, the spectrum of each component is obtained according to step (2.2)-(2.4)
Area ratio of the area relative to methane spectra substitutes into the heating value of natural gas Raman analysis model of foundation using area ratio as input
In, obtain the calorific value of the sample.
2. a kind of heating value of natural gas measurement method based on Raman spectrum according to claim 1, which is characterized in that described
In step (1.2), area normalization processing is carried out in modeling region to each pure component original spectrum, modeling region is 500-
2500cm-1。
3. a kind of heating value of natural gas measurement method based on Raman spectrum according to claim 1, which is characterized in that described
In step (1.2), full constituent detection is carried out to training sample with gas chromatography, and utilize standard GB/T/T 11062-
2014 are calculated the calorific value of each training sample.
4. a kind of heating value of natural gas measurement method based on Raman spectrum according to claim 1, which is characterized in that described
In step (2.3), the peak function is selected from Lorentz peak function, it may be assumed that
Wherein, βj=(Sj cj wj)T, v expression wave number, SjFor the area of j-th of Lorentz spectral peak, cjFor j-th of Lorentz spectral peak
Center wave number, wjFor the Lorentz half-breadth of j-th of Lorentz spectral peak.
5. a kind of heating value of natural gas measurement method based on Raman spectrum according to claim 4, which is characterized in that long-range navigation
Hereby spectral peak number N mends peak method by following iteration and obtains:
(a) do not consider C first first4 +Component optimizes pure component coefficient, so that:
Wherein,Y (v) is that training sample gas normalizes spectrum, [v1, v2] it is spectral resolution range；
(b) training sample gas normalization spectrum Y (v) is subtracted into R (v, α) as error spectrum Yc4(v), Y is selectedc4(v) in spectrogram most
At high point, a Lorentz peak is added, and to pure component factor alpha=(α1..., α7)TWith Lorentz peak parameter beta=(S1 c1 w1)
It optimizes again, so that
(c) training sample gas normalization spectrum Y (v) is subtracted into fit-spectra R (v, α, β) as error spectrum again, selected wherein most
At high point, then a Lorentz peak is added, at this time Lorentz peak parameter beta=(S1 c1 w1 S2 c2 w2), Optimal Parameters again,
Such iteration mends peak, until the Lorentz peak height or area that meet maximum number of iterations or addition are less than error.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810382163.9A CN108444976B (en) | 2018-04-26 | 2018-04-26 | A kind of heating value of natural gas measurement method based on Raman spectrum |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810382163.9A CN108444976B (en) | 2018-04-26 | 2018-04-26 | A kind of heating value of natural gas measurement method based on Raman spectrum |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108444976A CN108444976A (en) | 2018-08-24 |
| CN108444976B true CN108444976B (en) | 2019-10-22 |
Family
ID=63201210
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810382163.9A Active CN108444976B (en) | 2018-04-26 | 2018-04-26 | A kind of heating value of natural gas measurement method based on Raman spectrum |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108444976B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109405995A (en) * | 2018-09-17 | 2019-03-01 | 中国电子科技集团公司第五十五研究所 | A kind of analysis method improving the junction temperature of chip measuring accuracy based on Raman spectroscopy |
| EP3961195A1 (en) * | 2020-08-28 | 2022-03-02 | Siemens Aktiengesellschaft | Measuring device for determining the calorific value of a hydrocarbon-containing fuel gas |
| CN113552110A (en) * | 2021-07-16 | 2021-10-26 | 中国民航大学 | Raman spectrum-based dynamic early warning system and method for thermal runaway of lithium ion battery |
| CN114324294A (en) * | 2021-12-31 | 2022-04-12 | 中国海洋大学 | Multi-probe gas Raman spectrum system and mixed gas quantitative analysis method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101403696A (en) * | 2008-10-21 | 2009-04-08 | 浙江大学 | Method for measuring gasoline olefin content based on Raman spectrum |
| CN102590175A (en) * | 2012-02-21 | 2012-07-18 | 浙江大学 | Raman spectrum superposition-based method for quickly determining content of methanol in methanol gasoline |
| CN104777149A (en) * | 2015-04-17 | 2015-07-15 | 浙江大学 | Method for rapidly measuring content of trace methylbenzene in benzene based on Raman spectrum |
| CN105319198A (en) * | 2014-07-15 | 2016-02-10 | 中国石油化工股份有限公司 | Gasoline benzene content prediction method based on Raman spectrum analysis technology |
| CN105806825A (en) * | 2016-05-17 | 2016-07-27 | 浙江大学 | On-line gas Raman analysis method for natural gas components |
| CN106932378A (en) * | 2017-03-29 | 2017-07-07 | 浙江大学 | The on-line detecting system and method for a kind of sour gas composition based on Raman spectrum |
| CN107110776A (en) * | 2014-11-11 | 2017-08-29 | 光谱传感器公司 | Target analysis analyte detection and quantization in sample gas with complex background composition |
-
2018
- 2018-04-26 CN CN201810382163.9A patent/CN108444976B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101403696A (en) * | 2008-10-21 | 2009-04-08 | 浙江大学 | Method for measuring gasoline olefin content based on Raman spectrum |
| CN102590175A (en) * | 2012-02-21 | 2012-07-18 | 浙江大学 | Raman spectrum superposition-based method for quickly determining content of methanol in methanol gasoline |
| CN105319198A (en) * | 2014-07-15 | 2016-02-10 | 中国石油化工股份有限公司 | Gasoline benzene content prediction method based on Raman spectrum analysis technology |
| CN107110776A (en) * | 2014-11-11 | 2017-08-29 | 光谱传感器公司 | Target analysis analyte detection and quantization in sample gas with complex background composition |
| CN104777149A (en) * | 2015-04-17 | 2015-07-15 | 浙江大学 | Method for rapidly measuring content of trace methylbenzene in benzene based on Raman spectrum |
| CN105806825A (en) * | 2016-05-17 | 2016-07-27 | 浙江大学 | On-line gas Raman analysis method for natural gas components |
| CN106932378A (en) * | 2017-03-29 | 2017-07-07 | 浙江大学 | The on-line detecting system and method for a kind of sour gas composition based on Raman spectrum |
Non-Patent Citations (4)
| Title |
|---|
| Design and characterization of a Raman-scattering-based sensor system for temporally resolved gas analysis and its application in a gas turbine power plant;J Kiefer et al;《MEASUREMENT SCIENCE AND TECHNOLOGY》;20080710;第1-9页 * |
| Identification of unknown pure component spectra by indirect hard modeling;E. Kriesten et al;《Chemometrics and Intelligent Laboratory Systems》;20080505;第93卷;第108–119页 * |
| 在线拉曼分析系统关键技术研究与工业应用;阮华;《中国博士学位论文全文数据库 工程科技Ⅰ辑》;20130815;正文第1-125页 * |
| 拉曼光谱的数学解析及其在定量分析;李津蓉;《中国博士学位论文全文数据库 工程科技Ⅰ辑》;20130815;正文第1-118页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108444976A (en) | 2018-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108444976B (en) | A kind of heating value of natural gas measurement method based on Raman spectrum | |
| CN105806825B (en) | A kind of online gas Raman analysis method of gas component | |
| Ying et al. | Quantitative analysis of main components of natural gas based on Raman spectroscopy | |
| CN109983338B (en) | Measurement of Hydrocarbon Fuel gas composition and Properties | |
| Sieburg et al. | Characterization of fuel gases with fiber-enhanced Raman spectroscopy | |
| EP2994742B1 (en) | Method and apparatus for real-time measurement of fuel gas compositions heating values | |
| Knebl et al. | Hydrogen and C2–C6 alkane sensing in complex fuel gas mixtures with fiber-enhanced Raman spectroscopy | |
| CN102539374B (en) | Method for measuring coal gas component and calorific value | |
| CN106932378A (en) | The on-line detecting system and method for a kind of sour gas composition based on Raman spectrum | |
| CN101403696A (en) | Method for measuring gasoline olefin content based on Raman spectrum | |
| Jiang et al. | Tracing methane dissolved in transformer oil by tunable diode laser absorption spectrum | |
| CN112782146B (en) | Gasoline olefin content analysis method based on Raman spectrum | |
| US7973926B1 (en) | Methods for determining olefin concentrations in olefin-containing fuels | |
| CN110414169B (en) | Fourier infrared gas logging method and device thereof | |
| Li et al. | Robust cepstral analysis at variable wavelength scan depth for narrowband tunable laser absorption spectroscopy | |
| Li et al. | A hard modeling approach to determine methanol concentration in methanol gasoline by Raman spectroscopy | |
| CN103134764A (en) | Method for predicting crude oil true boiling point distillation curve by transmission infrared spectroscopy | |
| Yang et al. | Study on Temperature Influence of Infrared Gas Composition Detection | |
| Fan et al. | Nitric oxide detection using principal component analysis spectral structure matching to the UV derivative spectrum | |
| Pereira et al. | A flow system for generation of concentration perturbation in two-dimensional correlation near-infrared spectroscopy: application to variable selection in multivariate calibration | |
| Kim et al. | Quantitative measurement of ethane using mid-infrared cavity ring-down spectroscopy | |
| NL2030689B1 (en) | Analysis method for online monitoring of biomass blended-combustion ratio based on stable carbon isotopes | |
| Fan et al. | DEVELOPMENT OF METHANE CARBON ISOTOPE BASED ON MID-INFRARED ABSORPTION SPECTRUM TECHNOLOGY APPLIED IN LOGGING PLATFORM | |
| Jangale et al. | A real-time method for determining the composition and heating value of opportunity fuel blends | |
| Zelepouga et al. | Gas Quality Sensor to Improve Biogas-Fueled CHP/DG |
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 |