CN106814061A - A kind of method for improving LIBS overlap peak accuracy of quantitative analysis - Google Patents
A kind of method for improving LIBS overlap peak accuracy of quantitative analysis Download PDFInfo
- Publication number
- CN106814061A CN106814061A CN201611143482.1A CN201611143482A CN106814061A CN 106814061 A CN106814061 A CN 106814061A CN 201611143482 A CN201611143482 A CN 201611143482A CN 106814061 A CN106814061 A CN 106814061A
- Authority
- CN
- China
- Prior art keywords
- spectral line
- interference
- overlap
- opt
- sample
- 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
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
- G01N21/71—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light thermally excited
- G01N21/718—Laser microanalysis, i.e. with formation of sample plasma
Abstract
The invention discloses a kind of method for improving LIBS overlap peak accuracy of quantitative analysis, it is detected using LIBS technologies to calibration sample known to constituent content, obtains the spectrogram of calibration sample；Integrated intensity combined with wavelet transformed is recycled to be corrected to overlapping interference and continuous background interference, characteristic spectral line integrated intensity and analytical element content after correction are set up into single-variable linear regression model as independent variable and dependent variable, according to calibration root-mean-square error optimization integral domain factor w1And w2, overlap interference factor α, Decomposition order l, wavelet function and continuous background and deduct scale factor γ；Finally Optimal Parameters are obtained using corresponding calibration sample to be corrected the spectrum of testing sample, the content of analytical element in testing sample is predicted by single-variable linear regression model.The present invention can simultaneously remove overlap and interference of the continuous background to spectral line, can effectively improve the precision of overlap peak quantitative analysis in matrix complex sample.
Description
Technical field
The invention belongs to material composition detection technique field, LIBS is improved more particularly, to one kind
The method of overlap peak accuracy of quantitative analysis.
Background technology
LIBS (Laser Induced Breakdown Spectroscopy, abbreviation LIBS) is to pass through
To measured matter surface, ablation produces high-temperature plasma to the pulse laser focusing of high-energy-density, and then by collection analysis
Emission spectrum in plasma determines the composition of each element in measured matter and the elemental analysis technology of content.LIBS
Technology because have the advantages that speed it is fast, without sample preparation, multielement, it is remote and it is online detect simultaneously, in industrial production, ring
The numerous areas such as border monitoring, biological medicine and deep space probing are with a wide range of applications.
However, the significant challenge that LIBS technologies are present at present is, because the atom and ion line in plasma are in reality
There is broadening of spectral lines (dopplerbroadening, collision broadening and natural broadening) in the measurement process of border, cause the spectrum that centre wavelength is close
Can there is the interference phenomenon that overlaps each other in line.Additionally, plasma produces the compound of electronics bremstrahlen, ion and the electronics at initial stage
The caused strongly continuous background of radiation is also the key factor for influenceing spectrum actual strength.Therefore, the LIBS of matrix complex sample
In spectrum easily there is the phenomenon such as overlap interference and continuous background interference in spectral line, thus be difficult to select weak jamming, particularly without dry
The analysis spectral line disturbed, and select overlap peak often led to when carrying out quantitative determination LIBS analysis precision it is relatively low.
For the problems of overlap peak analysis in LIBS, high-resolution spectrometer and time high-resolution can be used
Combinations of detectors improve the accuracy of quantitative analysis of overlap peak to a certain extent, however, this will be significantly increased instrument cost and
Size, and overlap interference and its influence of continuous background in LIBS can not be completely eliminated.Another kind removal overlap of spectral lines interference
The effective method disturbed with continuous background is data processing method, can not only be effectively increased the range of choice of LIBS spectrum
(often only selecting weak jamming or glitch-free independent spectral line in the past), but also the quantitative analysis essence of LIBS can be greatly improved
Degree.Chinese patent CN103076308A discloses a kind of based on rational spectral peak Mathematical Modeling, by unconstrained optimization algorithm
The method of the spectral peak relevant information being calculated after Resolving Overlapped Peaks, this method effectively can carry out swarming to overlap peak, drop
The low influence for overlapping interference, but quantitative analysis of the method not to overlap peak in LIBS further furtherd investigate.Text
Offer " Accuracy improvement of quantitative analysis in laser-induced breakdown
Spectroscopy using modified wavelet transform " (April 21 2014 publication date) disclose one
Plant carries out continuous background correction to LIBS spectrum using wavelet transformation, improves the side of micro- accuracy of quantitative analysis in sample
Method, but analysis spectral line needs to choose the spectral line of non-overlapping interference in this method, its interference correction to overlapping disturbance spectrum line
Not yet study.
The content of the invention
For the disadvantages described above or Improvement requirement of prior art, LIBS is improved the invention provides one kind
Integral domain factor W after the method for overlap peak accuracy of quantitative analysis, its optimization obtained using corresponding calibration sample1And W2、
Overlap interference factor αopt, Decomposition order L, wavelet function F and continuous background deduct scale factor γoptSpectrum is corrected,
Finally by the content of analytical element in single-variable linear regression model prediction testing sample, can effectively solve existing because of overlap peak
The decomposition degree of accuracy is not high and cannot accurately remove the relatively low problem of the caused overlap peak accuracy of quantitative analysis of continuous background interference, tool
There are easy analysis, accurate, it is adaptable to which the overlap peak of the matrix material with complex especially material such as steel, plastics and oil is determined
The occasions such as amount analysis.
To achieve the above object, the present invention proposes a kind of raising LIBS overlap peak accuracy of quantitative analysis
Method, it is characterised in that comprise the following steps：
(1) one group of calibration sample is detected using LIBS, obtains the spectrum of calibration sample
Figure；
(2) wavelet function and Decomposition order l are preset, the spectral signal in the spectrogram is entered using wavelet transformation then
Row is decomposed, and obtains top low-frequency approximation coefficient CJ, by the CJThe spectral signal that zero setting is reconstructed；
(3) spectral signal according to the reconstruct determines analytical element spectral line and interference element spectral line in the calibration sample
Centre wavelength, and default spectral line limit of integration w1And w2, according to the centre wavelength and spectral line limit of integration w1、w2Obtain respectively
The integrated intensity of element spectral line and interference element spectral line must be analyzed；
(4) integrated intensity according to the analytical element spectral line and interference element spectral line sets up the single argument time of analytical element
Return model, and default overlap interference factor α；
(5) calibration root-mean-square error RMSEC, and spectral line limit of integration when determining to make the RMSEC take minimum value are set up
w1And w2, overlap interference factor α and Decomposition order l optimum organization W1、W2、αoptAnd L；
(6) according to the spectral line limit of integration W tried to achieve in step (5)1、W2, overlap interference factor αoptWith Decomposition order L with institute
RMSEC is stated for optimisation criteria is optimized to wavelet function, Optimum wavelet function F is obtained；
(7) according to the spectral line limit of integration W tried to achieve in step (5) and (6)1、W2, overlap interference factor αopt, Decomposition order L
Scale factor γ is deducted with Optimum wavelet function F to continuous background as optimisation criteria with the RMSEC to optimize, it is determined that most preferably
Continuous background deducts scale factor γopt；
(8) according to the spectral line limit of integration W tried to achieve in step (5), (6) and (7)1、W2, overlap interference factor αopt, point
Solution number of plies L, Optimum wavelet function F and optimal continuous background deduct scale factor γoptParameter Conditions under to the light of testing sample
Spectrum is processed, and analysis elements in testing sample are obtained in the single argument regression model for extracting the integrated intensity substitution analytical element
The content of element.
As it is further preferred that the wavelet function in the step (2) is the db1 of Daubechies wavelets functions
Wavelet function, Decomposition order l is 4~20 layers.
As it is further preferred that the integrated intensity of the analytical element spectral line and interference element spectral line is by middle cardiac wave
The w of the weaker side of interference long1And w2Spectral line is integrated in wave-length coverage is obtained.
As it is further preferred that the overlap interference factor α is introduced by the following method：
Single variable linear regression equation group is changed into after the interference of continuous background is deducted using wavelet transformation：
Wherein, S1iAnd S2iThe integrated intensity of analytical element spectral line and interference element spectral line is represented respectively, and i represents that sample is compiled
Number, c1iAnd c2iAnalytical element and interference element content, a are represented respectively1Represent analytical element spectral line integrated intensity in analytical element
With the rate of change of constituent content, a in spectral line limit of integration21Represent that interference element spectral line integrated intensity is integrated in analytical element spectral line
In the range of with constituent content rate of change, a2Represent interference element spectral line integrated intensity in interference element spectral line limit of integration with
The rate of change of constituent content, a12Represent analytical element spectral line integrated intensity in interference element spectral line limit of integration with constituent content
Rate of change；
Make a21=α × a2, obtain the single variable linear regression equation on analytical element：
S1i-α×S2i=(a1-α×a12)c1i
Wherein, α represents interference of the interference element spectral line to analytical element spectral line to overlap interference factor.
As it is further preferred that the calibration root-mean-square error RMSEC is：
Wherein,And c1iThe prediction content and standard content of analytical element in i-th sample in calibration collection, n are represented respectively
Represent the sum of calibration collection sample.
As it is further preferred that the Optimum wavelet function F is obtained in the following manner：
In the limit of integration W of optimization1And W2, overlap interference factor αoptRMSEC is set to get minimum with Decomposition order L parameters
The wavelet function of value, that is, meet：
RMSEC(W1,W2,αopt,F,L,γini)=RMSECmin(W1,W2,αopt,f,L,γini)；
Wherein, wavelet function f takes db1~db20, γiniTake 1.
As it is further preferred that the optimal continuous background deducts scale factor γoptObtain in the following manner：
The limit of integration W of optimization1And W2, overlap interference factor αopt, under Decomposition order L and Optimum wavelet function F parameters, after decomposition
Top low-frequency approximation coefficient CJEffect continuous background deducts scale factor γ, i.e., amended top low-frequency approximation coefficient C 'J
=(1- γ) CJ, spectral signal is then reconstructed, RMSEC is reached the γ as γ of minimum value againopt, that is, meet：
RMSEC(W1,W2,αopt,F,L,γopt)=RMSECmin(W1,W2,αopt,F,L,γ)。
As it is further preferred that the testing sample material is steel, plastics or oil.
The present invention obtains the LIBS spectrum of testing sample using LIBS, then using corresponding fixed
Integral domain factor W after the optimization that standard specimen product are obtained1And W2, overlap interference factor αopt, Decomposition order L, wavelet function F and company
Continuous background deduction scale factor γoptSpectrum is corrected, finally by single-variable linear regression model prediction testing sample
The content of analytical element.In general, by the contemplated above technical scheme of the present invention compared with prior art, mainly possess
Following technological merit：
(1) the inventive method for spectral line be subject to overlap interference and continuous background interference problem, introduce overlap interference because
Sub- α combinations continuous background deducts scale factor γ and is corrected simultaneously to overlapping interference and continuous background, reaches while accurately going
Except the effect for overlapping interference and continuous background interference, effectively reduce and overlap interference and continuous background interference to overlap peak quantitative analysis
Influence.
(2) the inventive method does not need the priori of interference element, i.e. the content information of interference element, you can sets up and divides
The single-variable linear regression model of element is analysed, the process of interference correction is simplified.
(3) on the basis of generally only selecting weak jamming or noiseless independent spectral line in conventional LIBS analyses, present invention side
Method can be effectively increased the range of choice of LIBS spectrum, you can carry out quantitative analysis using overlap peak spectral line.
(4), used as a kind of data processing method, can reduce LIBS measurements will to the precision of hardware device for the inventive method
Ask, hardware cost is greatly reduced, with important application value.
In a word, because current LIBS technologies are to material (the especially complicated sample of the matrix such as steel, plastics and oil)
Accuracy of quantitative analysis needs further raising, and the present invention can improve the precision of material quantitative analysis.
Brief description of the drawings
Fig. 1 is a kind of side for improving LIBS overlap peak accuracy of quantitative analysis provided in an embodiment of the present invention
The FB(flow block) of method.
Specific embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.As long as additionally, technical characteristic involved in invention described below each implementation method
Not constituting conflict each other can just be mutually combined.
In plasma emission spectroscopy, causing the principal element of characteristic spectral line broadening has：Natural broadening, dopplerbroadening
And Stark broadening.Under common experiment condition, natural broadening and the corresponding spectral line halfwidth (FWHM) of dopplerbroadening
The order of magnitude is respectively 10-5Nm and 10-3Nm, and the spectral line FWHM determined in testing is 10-1The nm orders of magnitude, therefore can ignore certainly
Right broadening and dopplerbroadening.Stark broadening is in the broadening mechanism of high-density plasma radiation spectral line in occupation of predominantly
Position, thus general described broadening of spectral lines refer to due to caused by radiating system and plasma intermediate ion and electron interaction this
Plutarch broadening.The corresponding linear function of Stark broadening is Lorentz linear functions, therefore analytical element spectral line is without self-absorption
Meet Lorentz line styles, line style intensity I in the case of phenomenonLFor：
Wherein, y0Continuous background intensity is represented, A represents the constant of spectral line characteristic, and π is math constant, and λ represents spectral line ripple
It is long, λcCore wavelength is represented, ω represents spectral line FWHM.
The influence of background intensity, spectral line peak value (λ=λ are not consideredc) intensity is：
In a series of samples, spectral line peak strength meets basic calibration formula I0=aC, wherein a represent that element unit contains
The changing value of the spectral line peak strength that amount causes, C represents constituent content.At broadening of spectral lines range distance centre wavelength Δ λ
The intensity of spectral line is：
In a series of samples, because ω is definite value when without self-absorption phenomenon, when Δ λ is definite value, broadening of spectral lines model
Enclose interior any point intensity IΔLinear relationship I is there is also between constituent content CΔ=aΔC, wherein aΔRepresent broadening of spectral lines scope
Interior any point intensity with constituent content rate of change, therefore, the integrated intensity S in the range of spectral lineiBetween constituent content C
There is linear relationship Si=aΔiC, wherein aΔiRepresent rate of change of the integrated intensity in the range of spectral line with constituent content.
When spectrum is present overlaps interference, interference element spectral line can be removed to analytical element spectral line by correcting integrated intensity
Interference, in order to avoid the influence of other invisible interference of naked eyes, limit of integration W1、W2Drawn using optimized algorithm.
It is generally acknowledged that spectrum is made up of three kinds of frequency informations, including high-frequency noise, low-frequency background, and be in
The spectral peak signal of intermediate frequency.The method for deducting continuous background using wavelet transformation is first by signal decomposition into multilayer frequency point
Amount, it is contemplated that continuous background only occupies certain proportion in top low frequency part, thus superior in top approximation coefficient
Upper scale factor γ (0<γ<1), then according to the inverse process of decomposition algorithm, using amended top low frequency part and respectively
Layer HFS, gradually reconstruction signal, so as to obtain the spectrum after background correction.Spectral signal f (t) after background correction can be with
The linear combination of wavelet function (equivalent to high-pass filter) and scaling function (equivalent to low pass filter) is launched into, i.e.,
Wherein, Ψj,k(t) andIt is respectively small echo father function and wavelet mother function (scaling function), j represents yardstick,
Corresponding to frequency, k is relevant with translation, and corresponding to spectral wavelength, J represents the yardstick of highest decomposition layer；dj,kAnd cJ,kIt is respectively small
Ripple detail coefficients and wavelet approximation coefficients, two in formula (4) details (high frequency) for representing signal f (t) respectively are partly and approximately
(low frequency) part.
Specific embodiment of the invention is described further with reference to Fig. 1, as shown in figure 1, the present invention proposes one
Plant and remove the method that interference improves LIBS overlap peak accuracy of quantitative analysis with continuous background interference that overlaps simultaneously,
Wherein continuous background interference is removed using wavelet transformation, is related to wavelet decomposition and reconstruction result and is passed through Matlab
Wavelet analysis tool box in R2014a softwares is obtained, and the method is first with LIBS technologies to fixed known to analytical element content
Standard specimen product carry out detecting the LIBS spectroscopic datas for obtaining testing sample；Then carried out simultaneously using integrated intensity combined with wavelet transformed
Interference and continuous background interference correction are overlapped, using the characteristic spectral line integrated intensity and analytical element content after correction as certainly
Variable and dependent variable set up single-variable linear regression model, and with the calibration root-mean-square error (RMSEC) of single argument regression model
It is optimisation criteria optimization integral domain factor w1And w2, overlap interference factor α, Decomposition order l, wavelet function and continuous background button
Except scale factor γ；The Optimal Parameters for finally being obtained using corresponding calibration sample carry out school to the LIBS spectrum of testing sample
Just, the content of analytical element in testing sample is predicted by single-variable linear regression model.The method specifically includes following steps：
(1) calibration sample LIBS spectrum are obtained
Using n steel samples known to analytical element content as calibration sample, analytical element content is respectively c11,
c12,……,c1n, calibration sample is detected using LIBS technologies, obtain the spectrogram of calibration sample.
(2) wavelet decomposition is carried out to spectral signal, and reconstructs spectral signal
Spectral signal in spectrogram can be resolved into multilayer frequency component by wavelet transformation, for example, choose Daubechies
The db1 wavelet functions of wavelets function, Decomposition order is 4~20 layers, and spectral signal is decomposed, and obtains each decomposition layer
High frequency detail coefficient D1,D2,……,Dl, and top low-frequency approximation coefficient CJ, by top low-frequency approximation coefficient CJPut
Zero, deduct scale factor γ equivalent to continuous background is madeini=1, reconstruct spectral signal using other high fdrequency components, it is assumed that decompose
The number of plies is l.
(3) Overlapped spectral line parameter acquiring
Inquiry National Institute of Standards and Technology (National Institute of Standard and
Technology, NIST) standard atomic spectra database, with reference to reconstruct spectral signal, determine analytical element spectral line and interference
The centre wavelength of spectral line, in order to avoid the influence of other invisible interference of naked eyes, the w of weaker side is disturbed in centre wavelength1And w2
(wherein, w1And w2Positioned at the both sides of centre wavelength, because broadening of spectral lines is in 0.1nm or so, w1And w2Span be 0.01
~0.1nm) spectral line is integrated in wave-length coverage, respectively obtain the integrated intensity S of analytical element spectral line11, S12... ..., S1n
With the integrated intensity S of interference element spectral line21, S22... ..., S2n, wherein n is calibration sample number.
(4) the single argument regression model of analytical element and interference element is set up
According to analytical element c containing moment matrix1=[c11,c12,……,c1n]TWith interference element c containing moment matrix2=
[c21,c22,……,c2n]T, and spectral line integrated intensity matrix S1=[S11, S12... ..., S1n]TAnd S2=[S21, S22... ...,
S2n]TSingle variable linear regression equation group during in the presence of overlap interference and continuous background interference can be obtained is：
Wherein, i represents sample number into spectrum (1≤i≤n), and T is the transposition of matrix, a1Represent analytical element spectral line integrated intensity
With the rate of change of constituent content, a in analytical element spectral line limit of integration21Represent that interference element spectral line is accumulated in analytical element spectral line
Point in the range of its integrated intensity with constituent content rate of change, a2Represent interference element spectral line integrated intensity in interference element spectral line
With the rate of change of constituent content, a in limit of integration12Represent analytical element spectral line integrated intensity in interference element spectral line limit of integration
The interior rate of change with constituent content, b is continuous background intensity, S in experimentation1、S2With analytical element content c1For, it is known that its
His unknown parameters.
Single variable linear regression equation group using formula (5) after the interference of wavelet transformation deduction continuous background is changed into：
Make a21=α a2, equation group disappear unit obtain the single variable linear regression equation on analytical element：
S1i-α×S2i=(a1-α×a12)c1i (7)
Wherein, α represents interference of the interference spectral line to analysis spectral line to overlap interference factor；a21Represent interference element spectral line
In analytical element spectral line limit of integration its integrated intensity with constituent content rate of change, therefore generally a21﹤ a2, i.e. α ﹤ 1, so
As long as obtaining being set up on analytical element content c by simple regression analysis by overlapping interference factor α1Calibration equation.
(5) W is determined1、W2、αopt, L optimum organization
According to Decomposition order l, wavelet function db1, spectral line limit of integration w that step (2), (3), (4) are chosen1And w2, overlap
Interference factor α, continuous background deducts scale factor γ and takes definite value, i.e. γini=1, using these parameter processing calibration sample spectrum
Obtain spectral line integrated intensity S1iAnd S2i, set up the prediction content that regression equation calculation obtains analytical element in calibration sampleMeter
Calculate the calibration root-mean-square error RMSEC under these parameter combinations：
Wherein,And c1iThe prediction content and standard content of analytical element in i-th sample in calibration collection, n are represented respectively
Represent the sum of calibration collection sample.
Concretely comprise the following steps：Db1 is taken in wavelet function, Decomposition order is carried out under taking 4~20 parameter combination to spectral signal
Decompose, by top low-frequency approximation coefficient CJZero setting (γini=1) after, using other high fdrequency components reconstruct spectral signal, so
Afterwards in w1(taking 0.01~0.1nm, step-length is 0.01nm) and w2Integrated in the range of (taking 0.01~0.1nm, step-length is 0.01nm)
To analysis spectral line and the integrated intensity S of interference spectral line1iAnd S2i, (0~1 is taken, 0.01) step-length is in conjunction with interference factor α is overlapped
The single variable linear regression equation (7) of analytical element can be set up, analytical element prediction content can be obtained using formula (7)
Further according to standard content c1iRMSEC values can be obtained.
It is then determined that making RMSEC get the limit of integration w of minimum value1And w2, overlap interference factor α, Decomposition order l most
Excellent solution W1、W2、αopt, L combination, that is, meet：
RMSEC(W1,W2,αopt,db1,L,γini)=RMSECmin(w1,w2,α,db1,l,γini) (9)
(6) optimal wavelet function F is determined
The limit of integration W tried to achieve using step (5)1、W2, overlap interference factor αoptWith Decomposition order L parameters to small echo letter
Number is optimized, that is, meet：
RMSEC(W1,W2,αopt,F,L,γini)=RMSECmin(W1,W2,αopt,f,L,γini) (10)
Concretely comprise the following steps：Wavelet function f takes db1~db20, and limit of integration takes W1、W2, overlap interference factor take αopt, decompose
The number of plies takes L, is obtained by step (5), and continuous background deducts scale factor γ and takes 1, and the wavelet function for making RMSEC values minimum is
Optimum wavelet function F.
(7) optimal background deduction factor gamma is determinedopt
The limit of integration W tried to achieve using step (5)1、W2, overlap interference factor αopt, Decomposition order L, and step (6) asks
The Optimum wavelet function F for obtaining further deducts scale factor γ and optimizes to continuous background, solves the optimum value γ of γopt, make
RMSEC reaches minimum value again, that is, meet：
RMSEC(W1,W2,αopt,F,L,γopt)=RMSECmin(W1,W2,αopt,F,L,γ) (11)
Concretely comprise the following steps：F is taken in wavelet function, Decomposition order is decomposed under taking the parameter combination of L to spectral signal,
Top low-frequency approximation coefficient C after to decompositionJZero setting but effect continuous background deduct scale factor γ (take 0~1,
Step-length is 0.01) that is, amended top low-frequency approximation coefficient C 'J=(1- γ) CJ, using amended low frequency component and its
Its high fdrequency component reconstructs spectral signal, then in W1、W2In the range of integration obtain analysis spectral line and interference spectral line integrated intensity
S1iAnd S2i, in conjunction with overlap interference factor αopt, the γ of RMSEC values minimum is optimal continuous background and deduct scale factor γ
opt。
(8) (analytical element is prediction unknown sample (i.e. testing sample) analytical element with the analytical element of calibration sample
Identity element) content
Its LIBS spectrum is obtained first with LIBS technology for detection analytical element content unknown samples, then using step
(5), the limit of integration W that (6) and (7) determine1、W2, overlap interference factor αopt, Decomposition order L, wavelet function F and the continuous back of the body
Scape deducts scale factor γoptSpectrum to unknown sample is processed, specific steps similar step (7), i.e., taken in wavelet function
F, Decomposition order is decomposed under taking the parameter combination of L to spectral signal, to top low-frequency approximation coefficient CJEffect is continuous
Background deduction scale factor γopt, spectral signal is reconstructed using amended low frequency component and other high fdrequency components, then in W1、
W2In the range of integration obtain analysis spectral line and interference spectral line integrated intensity S1iAnd S2i, in conjunction with overlap interference factor αopt, utilize
Formula (7) can try to achieve the content of analytical element in testing sample.
Embodiment 1
(1) the LIBS spectroscopic datas of sample are obtained
Laboratory sample is from 6 kinds of micro alloyed steel standard samples (numbering GSB03-2453-2008, Iron and Steel Research Geueral Inst analysis
Testing research institute and Ma'anshan Iron and Steel Co., Ltd develop), micro nonmetalloid Si and Main elements Fe in sample
Content is as shown in table 1.
Table 1
Experimental provision is common laser-induced breakdown spectroscopy device, and experiment is carried out under air ambient, using Q-switch
Nd:YAG pulse lasers (Beamtech Nimma 400, wavelength 532nm, repetition rate 3Hz, pulse width 8ns), laser energy
Amount is 80mJ/ pulses.Laser focuses on sample surfaces by speculum and planoconvex spotlight (focal length 15cm).In order to prevent puncturing sky
Gas, focus is at 4mm below sample surfaces.Collection time delay is set to 2 μ s, and gate-width is set to 10 μ s, the plasma for inspiring
Radiant light is collected and is coupled in optical fiber by light collector, is transmitted to spectrometer (Andor Technology, Mechelle
5000, wave-length coverage 200-950nm, resolution lambda/Δ λ=5000) carry out light splitting, spectrometer is equipped with enhanced charge-coupled device
Part (ICCD) (pixels of Andor Technology, iStar DH-334T 1024 × 1024) is used for opto-electronic conversion.Digital delay
Generator (DG535) is used for triggering the signal of laser and ICCD, realizes the Synchronization Control between signal, and ICCD connects with computer
Connect, so as to spectroscopic data is obtained and analyzed.In order to reduce influence of the Laser Energy Change to spectral intensity, each sample
Repeated acquisition 6 times, every spectrum accumulates 30 pulses.
(2) Overlapped spectral line parameter acquiring, Optimal Decomposition number of plies l, spectral line limit of integration w1、w2With overlap interference factor α
Inquiry NIST standard atomic spectra databases, with reference to actual spectrum, determine analytical element spectral line Si and interference spectral line
The centre wavelength of Fe is respectively 250.69nm and 250.61nm.Db1 is taken in wavelet function, Decomposition order l takes 4~20, integrates model
Enclose w1、w20.01~0.1nm is taken, interference factor α is overlapped and is taken 0~1, continuous background deducts the Parameter Conditions that scale factor γ takes 1
Under, extract integrated intensity set up the single-variable linear regression model on analytical element, and with RMSEC be optimisation criteria to decomposition
Number of plies l, spectral line limit of integration w1、w2Optimized with interference factor α is overlapped, it is 13, W to obtain optimization knot result L1And W2Respectively
It is+0.03nm and -0.03nm, αoptIt is 0.66.
(3) Optimization of Wavelet function f
Db1~db20 is taken in wavelet function f, Decomposition order L is 13, and it is 1 that continuous background deducts scale factor γ, spectral line product
Divide scope W1And W2Respectively+0.03nm and -0.03nm, overlaps interference factor αoptUnder for 0.66 Parameter Conditions, it is with RMSEC
Optimisation criteria is optimized to wavelet function f, obtains optimum results F for db9.
(4) continuous background deducts scale factor γ optimizations
It is db9 in wavelet function F, Decomposition order L is 13, continuous background deducts scale factor γ and takes 0~1, spectral line integration
Scope W1And W2Respectively+0.03nm and -0.03nm, overlaps interference factor αoptUnder for 0.66 Parameter Conditions, with RMSEC as excellent
Change standard deducts scale factor γ and optimizes to continuous background, obtains optimum results γoptIt is 0.95, final optimum results
As shown in table 2.
Table 2
(W is+expression integral domain centre wavelength right side, be-represent wavelength left side centered on integral domain)
(5) quantitative analysis results
It is db9 in wavelet function F, Decomposition order L is 13, and it is 0.95 that continuous background deducts scale factor γ, spectral line integration
Scope W1And W2Respectively+0.03nm and -0.03nm, overlaps interference factor αoptFor under 0.66 Parameter Conditions to standard sample light
Spectrum is processed, and is extracted in integrated intensity substitution single-variable linear regression model formula (7), finally calibrates regression curve analysis result
As shown in table 3.The predictive ability of single-variable linear regression model is marked from 6 every time using a cross-verification method evaluation is gone
5 samples are selected in quasi- sample as calibration sample, a remaining sample finally goes a cross-verification to return as testing sample
Return tracing analysis result as shown in table 3.In table 3, R2Represent the degree of fitting between prediction content and standard content；MPE represents 6
Sample predicts the mean percent ratio error between content and standard content.R2It is smaller closer to 1, MPE, represent the analysis effect of model
Fruit is better.
Table 3
Nonmetalloid Si spectral lines Si I 250.69nm are due to by Main elements Fe spectral line Fe II250.61nm in steel
Interference, causes not high with traditional method internal standard method and Lorentz fitting quantitative analysis precisions, and the inventive method can be effective
The accuracy of quantitative analysis of nonmetalloid Si spectral line Si I 250.69nm is improved, comparing result is as shown in table 3.
Embodiment 2
(1) the LIBS spectroscopic datas of sample are obtained
Laboratory sample is from 6 kinds of micro alloyed steel standard samples (numbering GSB03-2453-2008, Iron and Steel Research Geueral Inst analysis
Testing research institute and Ma'anshan Iron and Steel Co., Ltd develop), the content of minor metallic element V and Main elements Fe in sample
As shown in table 4.Each sample repeated acquisition of collection sample LIBS spectroscopic datas 6 times, every spectrum under identical experiment condition
30 pulses of accumulation.
Table 4
(2) Overlapped spectral line parameter acquiring, Decomposition order l, spectral line limit of integration w1、w2With the optimization for overlapping interference factor α
Inquiry NIST standard atomic spectra databases, with reference to actual spectrum, determine analytical element spectral line V and interference spectral line Fe
Centre wavelength be respectively 292.40nm and 292.38nm.Db1 is taken in wavelet function, Decomposition order l takes 4~20, limit of integration
w1、w20.01~0.1nm is taken, interference factor α is overlapped and is taken 0~1, continuous background deducts scale factor γ and takes under 1 Parameter Conditions,
Extract integrated intensity and set up the single-variable linear regression model on analytical element, and with RMSEC be optimisation criteria to decomposition layer
Number l, spectral line limit of integration w1、w2Optimized with interference factor α is overlapped, it is 15, W to obtain optimization knot result L1And W2Respectively+
0.05nm and -0.05nm, αoptIt is 0.26.
(3) wavelet function f optimizations
Db1~db20 is taken in wavelet function f, Decomposition order L is 15, and it is 1 that continuous background deducts scale factor γ, spectral line product
Divide scope W1And W2Respectively+0.05nm and -0.05nm, overlaps interference factor αoptUnder for 0.26 Parameter Conditions, it is with RMSEC
Optimisation criteria is optimized to wavelet function f, obtains optimum results F for db13.
(4) continuous background deducts scale factor γ optimizations
It is db13 in wavelet function F, Decomposition order L is 15, continuous background deducts scale factor γ and takes 0~1, spectral line integration
Scope W1And W2Respectively+0.05nm and -0.05nm, overlaps interference factor αoptUnder for 0.26 Parameter Conditions, with RMSEC as excellent
Change standard deducts scale factor γ and optimizes to continuous background, obtains optimum results γoptIt is 1, final optimum results such as table
Shown in 5.
Table 5
(W is+expression integral domain centre wavelength right side, be-represent wavelength left side centered on integral domain)
(5) quantitative analysis results
It is db13 in wavelet function F, Decomposition order L is 15, and it is 1 that continuous background deducts scale factor γ, spectral line integration model
Enclose W1And W2Respectively+0.05nm and -0.05nm, overlaps interference factor αoptFor under 0.26 Parameter Conditions to standard sample spectral
Processed, extracted in integrated intensity substitution single-variable linear regression model formula (7), final calibration regression curve analysis result is such as
Shown in table 6.The predictive ability of single-variable linear regression model is used and goes a cross-verification method evaluation, i.e., every time from 6 standards
5 samples are selected in sample as calibration sample, a remaining sample finally goes a cross-verification to return as testing sample
Tracing analysis result is as shown in table 6.In table 6, R2Represent the degree of fitting between prediction content and standard content；MPE represents 6 samples
Product predict the mean percent ratio error between content and standard content.R2It is smaller closer to 1, MPE, represent the analytical effect of model
Better.
Metallic element V spectral line VII 292.40nm are due to dry by Main elements Fe spectral line Fe II 292.38nm in steel
Disturb, cause not high with traditional method internal standard method and Lorentz fitting quantitative analysis precisions, the inventive method can be carried effectively
The accuracy of quantitative analysis of high metal element V spectral line V II 292.40nm, comparing result is as shown in table 6.
Table 6
Embodiment 3
(1) the LIBS spectroscopic datas of sample are obtained
Laboratory sample is from 6 kinds of micro alloyed steel standard samples (numbering GSB03-2453-2008, Iron and Steel Research Geueral Inst analysis
Testing research institute and Ma'anshan Iron and Steel Co., Ltd develop), the content of the minor metallic element Mn and Cr such as institute of table 7 in sample
Show.Each sample repeated acquisition of collection sample LIBS spectroscopic datas 6 times under identical experiment condition, every spectrum accumulates 30
Pulse.
Table 7
(2) Overlapped spectral line parameter acquiring, Decomposition order l, spectral line limit of integration w1、w2With the optimization for overlapping interference factor α
Inquiry NIST standard atomic spectra databases, with reference to actual spectrum, determine analytical element spectral line Mn and interference spectral line
The centre wavelength of Cr is respectively 357.79nm and 357.87nm.Db1 is taken in wavelet function, Decomposition order l takes 4~20, integrates model
Enclose w1、w20.01~0.1nm is taken, interference factor α is overlapped and is taken 0~1, continuous background deducts the Parameter Conditions that scale factor γ takes 1
Under, extract integrated intensity set up the single-variable linear regression model on analytical element, and with RMSEC be optimisation criteria to decomposition
Number of plies l, spectral line limit of integration w1、w2Optimized with interference factor α is overlapped, it is 15, W to obtain optimization knot result L1And W2Respectively
It is -0.04nm and+0.03nm, αoptIt is 0.16.
(3) wavelet function f optimizations
Db1~db20 is taken in wavelet function f, Decomposition order L is 15, and it is 1 that continuous background deducts scale factor γ, spectral line product
Divide scope W1And W2Respectively -0.04nm and+0.03nm, overlaps interference factor αoptUnder for 0.16 Parameter Conditions, it is with RMSEC
Optimisation criteria is optimized to wavelet function f, obtains optimum results F for db5.
(4) continuous background deducts scale factor γ optimizations
It is db5 in wavelet function F, Decomposition order L is 15, continuous background deducts scale factor γ and takes 0~1, spectral line integration
Scope W1And W2Respectively -0.04nm and+0.03nm, overlaps interference factor αoptUnder for 0.16 Parameter Conditions, with RMSEC as excellent
Change standard deducts scale factor γ and optimizes to continuous background, obtains optimum results γoptIt is 1, final optimum results such as table
Shown in 8.
Table 8
(W is+expression integral domain centre wavelength right side, be-represent wavelength left side centered on integral domain)
(5) quantitative analysis results
It is db5 in wavelet function F, Decomposition order L is 15, and it is 1, spectral line limit of integration that continuous background deducts scale factor γ
W1And W2Respectively -0.04nm and+0.03nm, overlaps interference factor αoptTo enter to standard sample spectral under 0.16 Parameter Conditions
Row treatment, extraction integrated intensity is substituted into single-variable linear regression model formula (7), final calibration regression curve analysis result such as table
Shown in 9.The predictive ability of single-variable linear regression model is used and goes a cross-verification method evaluation, i.e., every time from 6 standard samples
5 samples are selected in product as calibration sample, a remaining sample finally goes a cross-verification to return bent as testing sample
Line analysis result is as shown in table 9.In table 9, R2Represent the degree of fitting between prediction content and standard content；MPE represents 6 samples
Mean percent ratio error between prediction content and standard content.R2It is smaller closer to 1, MPE, represent that the analytical effect of model is got over
It is good.
Table 9
Metallic element Mn spectral lines Mn I 357.79nm are due to dry by metallic element Cr spectral line Cr I357.87nm in steel
Disturb, cause not high with traditional method internal standard method and Lorentz fitting quantitative analysis precisions, the inventive method can be carried effectively
The accuracy of quantitative analysis of high metal element M n spectral line Mn I 357.79nm, comparing result is as shown in table 9, due to Mn I
357.79nm spectral lines are very weak relative to interference the intensity of spectral line, and spectral line serious interference causes Lorentz fittings not draw effective result.
As it will be easily appreciated by one skilled in the art that the foregoing is only presently preferred embodiments of the present invention, it is not used to
The limitation present invention, all any modification, equivalent and improvement made within the spirit and principles in the present invention etc., all should include
Within protection scope of the present invention.
Claims (8)
1. a kind of method for improving LIBS overlap peak accuracy of quantitative analysis, it is characterised in that including following step
Suddenly：
(1) one group of calibration sample is detected using LIBS, obtains the spectrogram of calibration sample；
(2) wavelet function and Decomposition order l are preset, the spectral signal in the spectrogram is divided using wavelet transformation then
Solution, obtains top low-frequency approximation coefficient CJ, by the CJThe spectral signal that zero setting is reconstructed；
(3) spectral signal according to the reconstruct is determined in the calibration sample in analytical element spectral line and interference element spectral line
Cardiac wave is long, and default spectral line limit of integration w1And w2, according to the centre wavelength and spectral line limit of integration w1、w2Divided respectively
The integrated intensity of analysis element spectral line and interference element spectral line；
(4) integrated intensity according to the analytical element spectral line and interference element spectral line sets up the single argument recurrence mould of analytical element
Type, and default overlap interference factor α；
(5) calibration root-mean-square error RMSEC, and spectral line limit of integration w when determining to make the RMSEC take minimum value are set up1With
w2, overlap interference factor α and Decomposition order l optimum organization W1、W2、αoptAnd L；
(6) according to the spectral line limit of integration W tried to achieve in step (5)1、W2, overlap interference factor αoptWith Decomposition order L with described
RMSEC is optimized for optimisation criteria to wavelet function, obtains Optimum wavelet function F；
(7) according to the spectral line limit of integration W tried to achieve in step (5) and (6)1、W2, overlap interference factor αopt, Decomposition order L and most
Good wavelet function F deducts scale factor γ and optimizes as optimisation criteria with the RMSEC to continuous background, it is determined that optimal continuous
Background deduction scale factor γopt；
(8) according to the spectral line limit of integration W tried to achieve in step (5), (6) and (7)1、W2, overlap interference factor αopt, decomposition layer
Number L, Optimum wavelet function F and optimal continuous background deduct scale factor γoptParameter Conditions under the spectrum of testing sample is entered
Row treatment, analytical element in acquisition testing sample in the single argument regression model of the extraction integrated intensity substitution analytical element
Content.
2. it is according to claim 1 improve LIBS overlap peak accuracy of quantitative analysis method, its feature
It is that the wavelet function in the step (2) is the db1 wavelet functions of Daubechies wavelets functions, Decomposition order l is
4~20 layers.
3. the method for improving LIBS overlap peak accuracy of quantitative analysis according to claim 1 and 2, it is special
Levy and be, the integrated intensity of the analytical element spectral line and interference element spectral line is by the w in the weaker side of centre wavelength interference1
And w2Spectral line is integrated in wave-length coverage is obtained.
4. it is according to claim 3 improve LIBS overlap peak accuracy of quantitative analysis method, its feature
It is that the overlap interference factor α is introduced by the following method：
Single variable linear regression equation group is changed into after the interference of continuous background is deducted using wavelet transformation：
Wherein, S1iAnd S2iThe integrated intensity of analytical element spectral line and interference element spectral line is represented respectively, and i represents sample number into spectrum, c1i
And c2iAnalytical element and interference element content, a are represented respectively1Represent that analytical element spectral line integrated intensity is accumulated in analytical element spectral line
With the rate of change of constituent content, a in the range of point21Represent interference element spectral line integrated intensity in analytical element spectral line limit of integration
With the rate of change of constituent content, a2Represent that interference element spectral line integrated intensity contains in interference element spectral line limit of integration with element
The rate of change of amount, a12Represent analytical element spectral line integrated intensity in interference element spectral line limit of integration with the change of constituent content
Rate；
Make a21=α × a2, obtain the single variable linear regression equation on analytical element：
S1i-α×S2i=(a1-α×a12)c1i
Wherein, α represents interference of the interference element spectral line to analytical element spectral line to overlap interference factor.
5. it is according to claim 4 improve LIBS overlap peak accuracy of quantitative analysis method, its feature
It is that the calibration root-mean-square error RMSEC is：
Wherein,And c1iThe prediction content and standard content of analytical element in i-th sample in calibration collection are represented respectively, and n is represented
The sum of calibration collection sample.
6. it is according to claim 5 improve LIBS overlap peak accuracy of quantitative analysis method, its feature
It is that the Optimum wavelet function F is obtained in the following manner：
In the limit of integration W of optimization1And W2, overlap interference factor αoptRMSEC is set to get minimum value with Decomposition order L parameters
Wavelet function, that is, meet：
RMSEC(W1,W2,αopt,F,L,γini)=RMSECmin(W1,W2,αopt,f,L,γini)；
Wherein, wavelet function f takes db1~db20, γiniTake 1.
7. it is according to claim 6 improve LIBS overlap peak accuracy of quantitative analysis method, its feature
It is that the optimal continuous background deducts scale factor γoptObtain in the following manner：
In the limit of integration W of optimization1And W2, overlap interference factor αopt, under Decomposition order L and Optimum wavelet function F parameters, to point
Top low-frequency approximation coefficient C after solutionJEffect continuous background deducts scale factor γ, i.e., amended top low-frequency approximation
Coefficient C 'J=(1- γ) CJ, spectral signal is then reconstructed, RMSEC is reached the γ as γ of minimum value againopt, that is, meet：
RMSEC(W1,W2,αopt,F,L,γopt)=RMSECmin(W1,W2,αopt,F,L,γ)。
8. according to claim any one of 1-7 raising LIBS overlap peak accuracy of quantitative analysis side
Method, it is characterised in that the testing sample material is steel, plastics or oil.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611143482.1A CN106814061B (en) | 2016-12-13 | 2016-12-13 | A method of improving laser induced breakdown spectroscopy overlap peak accuracy of quantitative analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611143482.1A CN106814061B (en) | 2016-12-13 | 2016-12-13 | A method of improving laser induced breakdown spectroscopy overlap peak accuracy of quantitative analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106814061A true CN106814061A (en) | 2017-06-09 |
| CN106814061B CN106814061B (en) | 2019-05-21 |
Family
ID=59108962
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611143482.1A Active CN106814061B (en) | 2016-12-13 | 2016-12-13 | A method of improving laser induced breakdown spectroscopy overlap peak accuracy of quantitative analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106814061B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107677640A (en) * | 2017-09-29 | 2018-02-09 | 吉林大学 | A kind of plastic sample elemental recognition method based on LIBS |
| CN110208251A (en) * | 2019-06-20 | 2019-09-06 | 安徽创谱仪器科技有限公司 | Plasma emission spectroscopy interference correction method |
| CN110398488A (en) * | 2019-06-11 | 2019-11-01 | 中国科学院合肥物质科学研究院 | Quantitative analysis of nonlinear method applied to laser induced breakdown spectroscopy |
| CN111982949A (en) * | 2020-08-19 | 2020-11-24 | 东华理工大学 | Method for separating EDXRF spectrum overlapping peak by combining fourth derivative with three-spline wavelet transform |
| CN112700822A (en) * | 2020-12-03 | 2021-04-23 | 西南交通大学 | Laser-induced breakdown spectroscopy concentration extraction method for online monitoring of trace gas impurities |
| CN113092448A (en) * | 2021-03-24 | 2021-07-09 | 中南大学 | Online detection method and system for silicon content of blast furnace molten iron |
| CN114199856A (en) * | 2021-12-16 | 2022-03-18 | 福建师范大学 | Method for rapidly calibrating element content ratio of laser-induced breakdown spectroscopy |
| CN114322762A (en) * | 2021-12-16 | 2022-04-12 | 上海精测半导体技术有限公司 | Optical parameter measuring method and device |
| CN114894781A (en) * | 2022-05-25 | 2022-08-12 | 中国海洋大学 | Underwater LIBS spectrum standardization generalization method and system based on external signal |
| CN114894782A (en) * | 2022-07-14 | 2022-08-12 | 合肥金星智控科技股份有限公司 | LIBS quantitative analysis method, system and electronic equipment |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103076308A (en) * | 2011-10-25 | 2013-05-01 | 中国科学院沈阳自动化研究所 | Laser-induced breakdown spectroscopy overlapped peak resolution method |
| CN103983617A (en) * | 2014-05-04 | 2014-08-13 | 华中科技大学 | Improved laser probe quantitative analysis method based on wavelet transform |
-
2016
- 2016-12-13 CN CN201611143482.1A patent/CN106814061B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103076308A (en) * | 2011-10-25 | 2013-05-01 | 中国科学院沈阳自动化研究所 | Laser-induced breakdown spectroscopy overlapped peak resolution method |
| CN103983617A (en) * | 2014-05-04 | 2014-08-13 | 华中科技大学 | Improved laser probe quantitative analysis method based on wavelet transform |
Non-Patent Citations (1)
| Title |
|---|
| 张天龙等: "化学计量学在激光诱导击穿光谱分析中的研究进展", 《分析化学》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107677640B (en) * | 2017-09-29 | 2018-07-27 | 吉林大学 | A kind of plastic sample elemental recognition method based on laser induced breakdown spectroscopy |
| CN107677640A (en) * | 2017-09-29 | 2018-02-09 | 吉林大学 | A kind of plastic sample elemental recognition method based on LIBS |
| CN110398488B (en) * | 2019-06-11 | 2021-09-07 | 中国科学院合肥物质科学研究院 | Nonlinear quantitative analysis method applied to laser-induced breakdown spectroscopy |
| CN110398488A (en) * | 2019-06-11 | 2019-11-01 | 中国科学院合肥物质科学研究院 | Quantitative analysis of nonlinear method applied to laser induced breakdown spectroscopy |
| CN110208251A (en) * | 2019-06-20 | 2019-09-06 | 安徽创谱仪器科技有限公司 | Plasma emission spectroscopy interference correction method |
| CN111982949A (en) * | 2020-08-19 | 2020-11-24 | 东华理工大学 | Method for separating EDXRF spectrum overlapping peak by combining fourth derivative with three-spline wavelet transform |
| CN112700822A (en) * | 2020-12-03 | 2021-04-23 | 西南交通大学 | Laser-induced breakdown spectroscopy concentration extraction method for online monitoring of trace gas impurities |
| CN113092448A (en) * | 2021-03-24 | 2021-07-09 | 中南大学 | Online detection method and system for silicon content of blast furnace molten iron |
| CN114199856A (en) * | 2021-12-16 | 2022-03-18 | 福建师范大学 | Method for rapidly calibrating element content ratio of laser-induced breakdown spectroscopy |
| CN114322762A (en) * | 2021-12-16 | 2022-04-12 | 上海精测半导体技术有限公司 | Optical parameter measuring method and device |
| CN114199856B (en) * | 2021-12-16 | 2023-06-23 | 福建师范大学 | Rapid calibration method for element content ratio of laser-induced breakdown spectroscopy |
| CN114322762B (en) * | 2021-12-16 | 2024-03-29 | 上海精测半导体技术有限公司 | Optical parameter measuring method and device |
| CN114894781A (en) * | 2022-05-25 | 2022-08-12 | 中国海洋大学 | Underwater LIBS spectrum standardization generalization method and system based on external signal |
| CN114894782A (en) * | 2022-07-14 | 2022-08-12 | 合肥金星智控科技股份有限公司 | LIBS quantitative analysis method, system and electronic equipment |
| CN114894782B (en) * | 2022-07-14 | 2022-12-23 | 合肥金星智控科技股份有限公司 | LIBS quantitative analysis method, system and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106814061B (en) | 2019-05-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106814061B (en) | A method of improving laser induced breakdown spectroscopy overlap peak accuracy of quantitative analysis | |
| Syvilay et al. | Evaluation of the standard normal variate method for Laser-Induced Breakdown Spectroscopy data treatment applied to the discrimination of painting layers | |
| Naes et al. | A comparison of laser ablation inductively coupled plasma mass spectrometry, micro X-ray fluorescence spectroscopy, and laser induced breakdown spectroscopy for the discrimination of automotive glass | |
| Lazic et al. | Self-absorption model in quantitative laser induced breakdown spectroscopy measurements on soils and sediments | |
| Shi et al. | Quantitative analysis of sedimentary rocks using laser-induced breakdown spectroscopy: comparison of support vector regression and partial least squares regression chemometric methods | |
| Dyar et al. | Comparison of univariate and multivariate models for prediction of major and minor elements from laser-induced breakdown spectra with and without masking | |
| Lasheras et al. | Quantitative analysis of oxide materials by laser-induced breakdown spectroscopy with argon as an internal standard | |
| CN103983617A (en) | Improved laser probe quantitative analysis method based on wavelet transform | |
| Yu et al. | Principles, developments and applications of laser-induced breakdown spectroscopy in agriculture: A review | |
| Dyar et al. | Comparison of baseline removal methods for laser-induced breakdown spectroscopy of geological samples | |
| Bhatt et al. | LIBS development methodology for forensic nuclear materials analysis | |
| EP2525213A1 (en) | Spectroscopic apparatus and methods for determining components present in a sample | |
| Shah et al. | Quantitative elemental analysis of steel using calibration-free laser-induced breakdown spectroscopy | |
| CN105319198A (en) | Gasoline benzene content prediction method based on Raman spectrum analysis technology | |
| Müller et al. | Enhanced spectral analysis of C-TOF Aerosol Mass Spectrometer data: Iterative residual analysis and cumulative peak fitting | |
| Xiu et al. | Quantitative determination of manganese in aqueous solutions and seawater by laser-induced breakdown spectroscopy (LIBS) using paper substrates | |
| US20170191929A1 (en) | Spectral modeling for complex absorption spectrum interpretation | |
| CN103543132B (en) | A kind of coal characteristic measuring method based on wavelet transformation | |
| CN109781711A (en) | A kind of laser induced breakdown spectroscopy quantitative analysis method based on the calibration of single standard specimen | |
| CN112834485B (en) | Non-calibration method for quantitative analysis of laser-induced breakdown spectroscopy elements | |
| Horňáčková et al. | Determination of Si/Al molar ratios in microporous zeolites using calibration-free laser induced breakdown spectroscopy | |
| Goltz | A review of instrumental approaches for studying historical inks | |
| Takahashi et al. | Partial least squares regression calculation for quantitative analysis of metals submerged in water measured using laser-induced breakdown spectroscopy | |
| Yu et al. | Response surface methodology for optimizing LIBS testing parameters: A case to conduct the elemental contents analysis in soil | |
| CN103792214A (en) | Method for improving carbon content measuring accuracy in steel |
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 | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Guo Lianbo Inventor after: Guo Yangmin Inventor after: Yang Xinyan Inventor after: Zhu Zhihao Inventor after: Li Kuohu Inventor after: Li Xiangyou Inventor after: Zeng Xiaoyan Inventor before: Guo Lianbo Inventor before: Guo Yangmin Inventor before: Yang Xinyan Inventor before: Zhu Zhihao Inventor before: Li Kuohu Inventor before: Li Xiangyou Inventor before: Zeng Xiaoyan Inventor before: Lu Yongfeng |