CN102103080A - Spectrum analyzing method - Google Patents
Spectrum analyzing method Download PDFInfo
- Publication number
- CN102103080A CN102103080A CN 201010622421 CN201010622421A CN102103080A CN 102103080 A CN102103080 A CN 102103080A CN 201010622421 CN201010622421 CN 201010622421 CN 201010622421 A CN201010622421 A CN 201010622421A CN 102103080 A CN102103080 A CN 102103080A
- Authority
- CN
- China
- Prior art keywords
- matrix
- spectral line
- characteristic spectral
- analytical
- line
- 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
Images
Abstract
The invention relates to a spectrum analyzing method, comprising the following steps of: A, building an analysis model, namely exciting a standard sample to obtain atomic spectrum information, selecting m characteristic spectrum lines of elements to be analyzed from the spectrum information, and building an analyzing model according to spectrum line data corresponding to the m characteristic spectrum lines and the element content/induction content of the elements to be analyzed, wherein m is more than or equal to 2; and B, analyzing an unknown sample, namely obtaining spectrum line data of corresponding elements to be analyzed in the unknown sample according to the method in the step A, and substituting the spectrum line data into the analysis model to obtain the contents of the elements to be analyzed in the unknown sample. The invention has the advantages of steady analysis model, accurate analysis and good repeatability.
Description
Technical field
The present invention relates to a kind of spectroscopic analysis methods, the content of element in the sample is detected.
Background technology
Atomic emission spectrometry is to utilize atom or ion excited by external condition and the characteristic spectral line launched comes the analytical approach of amalyzing substances chemical composition.Mainly be that line strength according to atomic emission spectrum becomes certain relation to carry out quantitative test with the concentration of element to be analyzed in the analyzed sample.
Direct-reading spectrometer is a kind of of Atomic Emission Spectrometer AES, is mainly used in the detection of metal ingredient, and application comprises material evaluation, on-the-spot sample analysis etc.The device of direct-reading spectrometer comprises excitation apparatus 1, electrode 2, metal sample 3, light-dividing device 4, measurement mechanism 5 and data processing equipment 6 as shown in Figure 1.Its principle of work is: excitation apparatus 1 produces a high-voltage pulse, be applied between electrode 2 and the metal sample 3, a part of vaporised atom of metal sample, high-voltage pulse forms the lower ion plasma of resistance with the gas between electrode 2 and the metal sample 3 (air or blanket gas) ionization simultaneously; Excitation apparatus 1 is to the ion plasma output current, and the metal sample atom that evaporates in the ion plasma is excited under the effect of electric current, and the internal electron transition gives off a large amount of atomic spectral lines; The half-breadth of spectral line only is the micromicron magnitude; Line wavelength is relevant with the kind of element, and intensity is relevant with the content of element.
Light-dividing device 4 is pressed the wavelength order separately with spectral line, and the intensity that measurement mechanism 5 is surveyed spectral line obtains line strength.Data processing equipment 6 is set up the constituent content of analytical model or calculation sample according to line strength.
But the analytical model approximate representation between constituent content and the spectral line data is a following formula:
In the formula, c representation element cellulose content (or the content ratio of analytical element and matrix element, claim " inducing content "), a
mThe expression coefficient, I represents spectral line data, is line strength of analytical line or the analytical line ratio with line strength of internal standard line, or the ratio of internal standard line and line strength of analytical line, n represents high order power; In the direct-reading spectrometer, n generally equals 2, is 1 or 3 under a few cases.
In actual application, need at first utilize the known standard model of constituent content to set up quantitative relationship between line strength of constituent content and analytical element characteristic spectral line, and then carry out the analysis of constituent content.
But element to be analyzed generally has many characteristic spectral lines in scanning optical spectrum (full spectrum) scope.For example: after the analyzed sample that with the aluminium element is matrix element is subjected to external condition and excites, in the full spectral coverage scope of ultraviolet-visible, aluminium element can the obvious characteristics peak occur at 4 wavelength places such as 265.25nm, 266.04nm, 305.71nm and 358.67nm, and we are called the characteristic spectral line of aluminium element these characteristic peaks.
In the actual analysis process, factors such as analyzed sample physicochemical property will be brought influence in various degree to each characteristic spectral line in the spectral coverage scope.As, be in the analyzed sample of matrix element with the aluminium element, the variation that excites equivalent temperature that the element silicon of different content can directly influence analyzed sample when being subjected to light source activation, thus influence line strength of each elemental characteristic spectral line; And the content of element silicon also can directly influence the physical characteristics of analyzed sample, thereby the degree that is stimulated of the analyzed sample of remote effect, and then influence line strength of each characteristic spectral line, and final influence is according to the accuracy of constituent content with the analytical model of line strength foundation.
Traditional analytical approach relies on artificial experience to select the single characteristic spectral line of respective element to set up analytical model, can realize the mensuration of sample constituent content in certain content range like this, but still have following problem:
1, analytical model is set up the process complexity
Corresponding many characteristic spectral lines of element to be analyzed possibility in the spectral scan scope, as, may there be the characteristic spectral line of Mg285.2nm and 382.9nm in magnesium elements, when analyzing the magnesium elements of low content, selects Mg285.2nm as characteristic spectral line usually; When analyzing the magnesium elements of high-load, generally select Mg382.9nm as characteristic spectral line; For this situation, need set up corresponding analytical model at every characteristic spectral line; The modeling process complexity is loaded down with trivial details, and the versatility of analytical model is relatively poor;
2, analytic process needs manual intervention
Need set up different analytical models for different characteristic spectral lines; Then, need set up a plurality of analytical models for same element; And in actual use, when different samples to be analyzed is analyzed, need constantly change the analytical model that is fit to analyze sample to be analyzed in the mode of manual intervention according to constituent content scope and analytical effect; The analytic process trouble, and can't determine which analytical model analysis result is more accurate; Analysis result is relatively subjective;
3, analysis result poor stability
When using selected single characteristic spectral line to set up analytical model, influence factors such as the state parameter of instrument and environmental baseline when only having considered to set up model, but the fine difference of influence factors such as the state parameter of the instrument in the actual use and environmental baseline when setting up model do not obtain embodying in model, and these small differences will have a strong impact on precision of analysis, even cause analytical model can't continue to use in subsequent analysis;
For example: the general standard model of fixing that uses when setting up analytical model, and be the process sample of smelting at the sample of on-the-spot sample analysis, have notable difference between standard model when then setting up analytical model and the actual sample to be analyzed, when adopting when according to the analytical model of single characteristic spectral line exploitation the sample constituent content being analyzed, these differences will be brought in various degree influence to precision of analysis;
Simultaneously, the traditional analysis method has been lost the information of further feature spectral line in the scanning optical spectrum scope, only adopts single characteristic spectral line to set up analytical model, and the versatility and the robustness of analytical model are relatively poor, precision of analysis and poor repeatability.
Summary of the invention
In order to solve above-mentioned deficiency of the prior art, the invention provides a kind of analytical model sane, analyze accurately, the spectroscopic analysis methods of favorable reproducibility.
A kind of spectroscopic analysis methods may further comprise the steps:
A, set up analytical model
Excite standard model, obtain atomic spectrum information;
From described spectral information, select the m bar characteristic spectral line of analytical element, m 〉=2;
According to the spectral line data of above-mentioned m bar characteristic spectral line correspondence and the constituent content of described analytical element/induce content, set up analytical model;
B, analysis unknown sample
Obtain the spectral line data of unknown sample corresponding analysis element according to the method for steps A, and the above-mentioned analytical model of substitution, obtain the content of unknown sample corresponding analysis element.
Further, described analytical model is Y
P * 1=X
P * t* F
T * 1, wherein Y is a constituent content/induce to contain moment matrix, and X is the spectral line data matrix, and F is a matrix of coefficients, and p is the sample number, p 〉=2, t 〉=m.
Further, a1, the line strength of getting m bar characteristic spectral line are formed the vectorial I of row;
With the vectorial I combination of the row of p piece sample, set up intensity matrix A
P * q, q=m;
A2, utilize matrix A
P * qMake up s order polynomial matrix B
P * t, t=s * q+1, s=1,2,
A3, with matrix B
P * tAs spectral line data matrix X
P * t
Further, by matrix A
P * qMake up matrix B
P * tMethod be:
Matrix B
P * t(s * i-j+1) be listed as each matrix element is a matrix A
P * qI is listed as the j power of corresponding matrix element, i=1, and 2 ..., q; J=1,2 ..., s;
Matrix B
P * t(s * q+1) be listed as each matrix element to be constant.
As preferably, matrix B
P * t(s * q+1) be listed as each matrix element to be 1.
Further, in steps A, from described spectral information, select the characteristic spectral line of n bar internal standard element again, n 〉=1;
Step a1 is: with the characteristic spectral line of m bar analytical element and the characteristic spectral line cross match of n bar internal standard element, and get pairing line strength ratio of individual features spectral line afterwards, form the vectorial I of row;
With the vectorial I combination of the row of p piece sample, set up intensity matrix A
P * q, q=m * n.
Further, among the step a1, it is line strength vector of r that each wavelength points intensity groups of characteristic spectral line becomes length, line strength Vector Groups of m bar characteristic spectral line vectorial I that embarks on journey, r 〉=1;
With the vectorial I combination of the row of p piece sample, set up intensity matrix A
P * q, q=m * n * r.
Further, in steps A, select the characteristic spectral line of n bar internal standard element again from described spectral information, it is line strength vector of l that interior each wavelength points intensity groups of mark spectral line characteristic spectral line becomes length, l 〉=1;
Step a1 is: with the characteristic spectral line of m bar analytical element and the characteristic spectral line cross match of n bar internal standard element, and to the vector element cross match in line strength vector of pairing, and get the ratio of vector element behind the cross match, form the vectorial I of row;
With the vectorial I combination of the row of p piece sample, set up intensity matrix A
P * q, q=m * n * r * l.
Further, do not comprise step a2;
Step a3 is: with matrix A
P * qAs spectral line data matrix X
P * t, q=t.
Further, internal standard element is the sample matrices element.
As preferably, adopt the offset minimum binary algorithm, set up analytical model.
The present invention compared with prior art has following beneficial effect:
1, utilized the useful information of a plurality of characteristic spectral lines of scanning optical spectrum scope interior element, fully extracted the characteristic information of identity element at the different wave length place, avoided of the influence of external interference factor, made that the analytical model of being developed is more sane the single analysis spectral line of certain bar;
2, adopt analysis channel and interior mark passage separately integral area ratio or the absolute light intensity value at multi-wavelength place as relative light intensity, further extracted the correlationship of analyzed element and relative light intensity; Owing to be the accumulation contribution of a plurality of characteristic value datas, avoided the influence of extraneous disturbing factor to relative light intensity, make the relative light intensity of extraction more stable, avoided the errors of analytical results or the mistake that cause by abnormal data;
3, a plurality of analysis channels and interior mark passage combined crosswise are extracted effective information, made full use of the accumulation contribution advantage of many characteristic spectral lines;
4, adopt the multivariable nonlinearity regression algorithm to consider the match weight of a plurality of variablees, avoided the influence of interfering channel, improved the antijamming capability of precision of analysis and analytical model;
5, same type sample only need be set up the analytical model of selected spectral line in the scanning optical spectrum scope, need not the single characteristic spectral line of artificial selection bar sets up analytical model and carries out the sample size analysis, avoided the trouble of artificial switching model back and forth, made analysis result more accurately, reliably; Same owing to avoided the influence of external interference factor, make analytical model between instrument, transplant and also become very convenient.
Description of drawings
Fig. 1 is the structural representation of direct-reading spectrometer;
Fig. 2 is that analysis channel and interior mark passage combined crosswise are extracted the relative light intensity access diagram;
Fig. 3 is the effect contrast figure that spectroscopic analysis methods of the present invention and existing method are analyzed same sample.
Embodiment
A kind of spectroscopic analysis methods may further comprise the steps:
A, set up analytical model, specific as follows:
A1, excite the p piece to be used for the standard model of modeling, obtain spectral information;
A2, from spectral information, select the characteristic spectral line of analytical element and internal standard element respectively; In the present embodiment, internal standard element is a matrix element; In the selected characteristic spectral line, at least two of the characteristic spectral lines of analytical element; The characteristic spectral line of described internal standard element is one or more;
A3, with each characteristic spectral line of selected internal standard element respectively with each characteristic spectral line cross match of analytical element, and get the spectral line strength ratio of pairing back individual features spectral line correspondence; Make up spectral line data according to spectral line strength ratio;
The ratio of described line strength is the ratio of analytical element line strength corresponding with the characteristic spectral line of internal standard element, or the ratio of internal standard element line strength corresponding with the characteristic spectral line of analytical element; The ratio of line strength is the ratio of analytical element line strength corresponding with the characteristic spectral line of internal standard element in the present embodiment;
A4, set up constituent content/the induce analytical model between content and the spectral line data;
Described element induces content to be meant that the ratio of analytical element content and matrix element content multiply by 100;
Present embodiment is set up is analytical model between analytical element content and the spectral line data;
B, analysis unknown sample
Obtain the spectral line data of unknown sample according to the method for steps A 1～A3, and the analytical model of setting up in the substitution steps A 4, obtain the constituent content to be analyzed of unknown sample.
See also Fig. 1～Fig. 3, present embodiment is to adopt CCD type Atomic Emission Spectrometer AES collected specimens spectrum;
1) steps A 1 realizes as follows:
Adopt photo-electric direct reading spectrometer, provide energy to make sample partly evaporate, form gaseous atom, and gaseous atom is excited and produce complex light by light source, through optical device such as gratings this complex light is resolved into by the tactic a series of monochromatic light of wavelength then, by photoelectricity testing part light signal is converted to the spectral line that electric signal also finally converts a series of different wave length correspondences to then, thereby forms sample spectra; This spectrum has comprised the characteristic spectral line of different elements and identity element diverse location; This step is the prior art in atomic spectroscopic analysis field, does not repeat them here;
2) steps A 3 realizes as follows:
The characteristic spectral line of selected analytical element is the m bar in the steps A 2, m 〉=2, and the characteristic spectral line of internal standard element is the n bar, n 〉=1;
A1, utilize line strength to set up matrix A
P * q
The m bar characteristic spectral line of analytical element is A1, A2...Am;
N bar characteristic spectral line B1, the B2...Bn of internal standard element;
Utilize Atomic Emission Spectrometer AES to obtain line strength of analysis spectral line correspondence, be respectively I
A1, I
A2... I
AmAnd the intensity of interior mark spectral line correspondence, be respectively I
B1, I
B2... I
Bn
The method that obtains line strength is the prior art in atomic spectroscopic analysis field, repeats no more herein;
See also Fig. 2, analysis spectral line and interior mark spectral line are intersected, make each analysis spectral line and interior mark spectral line carry out the cross match combination, generate A1B1, A2B1, AmB1 ... m such as A1Bn, A2Bn, AmBn * n pairing;
Analysis spectral line line strength corresponding with interior mark spectral line in the pairing compared, and the strength ratio that obtains corresponding pairing is I
A1/ I
B1, I
A2/ I
B1, I
Am/ I
B1I
A1/ I
Bn, I
A2/ I
Bn, I
Am/ I
Bn, the strength ratio of all pairings is formed the vectorial I of row
A/B
The vectorial I of row that the p piece is used for the standard model of modeling
A/BBe combined to form matrix A
P * q, q=m * n;
The B of a2, structure s order polynomial
P * t, t=q * s+1, s=1,2, S=2 in the present embodiment;
Matrix B
P * t(s * i-2+1) be listed as each matrix element is a matrix A
P * qI is listed as the quadratic power of corresponding matrix element, i=1, and 2 ..., q;
Matrix B
P * t(s * q+1) being listed as each matrix element is 1;
A3, with matrix B
P * tAs spectral line data matrix X
P * t
3) steps A 4 is set up the analytical model Y between constituent content and the spectral line data
P * 1=X
P * t* F
T * 1:
I) to matrix X
P * tWith contain moment matrix Y
P * 1Carry out main gene and decompose, extract X
P * tAnd Y
P * 1Feature vector, X '
P * sAnd Y '
P * s, s is a score main cause subnumber;
Ii) to X '
P * sAnd Y '
P * sCarry out regression Calculation, ask for the regression coefficient matrix F
T * 1
Iii) preserve and write down regression coefficient F
T * 1, output analytical model Y
P * 1=X
P * t* F
T * 1
Step I) and step I i) be the calculation procedure of partial least squares regression algorithm, described regression algorithm is a prior art, has only provided process prescription herein, does not do specific explanations.Adopt the partial least squares regression algorithm to consider the match weight of a plurality of variablees, avoided disturbing the influence of spectral line, improved the antijamming capability of precision of analysis and analytical model;
According to the described spectroscopic analysis methods of present embodiment, the many covers standard model that uses different trade mark multicomponent alloys (low-alloy aluminium, almag, Al-Si-Cu alloy, alumin(i)um zinc alloy and almag etc.), various criterion sample manufacturer (southwestern Aluminum and Fushun Aluminum) to produce, amount to 30 aluminium base samples, set up the analytical model of Mg element, and compared with traditional conventional method.
Present embodiment is in step a2, the characteristic spectral line of selected analytical element Mg is four of 279.55nm, 285.2nm, 383.2nm and 383.83nm etc. from spectral information, the characteristic spectral line of selected internal standard element is three of 266.04nm, 305.71nm and 358.67nm etc., m=4 then, n=3, p=30; Internal standard element is a matrix element aluminium;
In step a4, the matrix that constructs is respectively: the spectral line data matrix A
30 * 12With contain moment matrix Y
30 * 1
To A
30 * 12Decompose, construct 2 order polynomial matrix B
30 * 25With matrix B
30 * 25Be made as X
30 * 25, to X
30 * 25Extract the score X ' of 3 main genes
30 * 3And Y '
30 * 3, and to X '
30 * 3And Y '
30 * 3Return, preserve X '
30 * 3And Y '
30 * 3Fitting coefficient F
25 * 1, the output analytical model;
Graph of relation between the constituent content of drawing the element real content of analytic sample and being calculated by analytical model is shown in Fig. 3 (a);
In the traditional analysis method, rule of thumb select the Mg383.83nm characteristic spectral line to carry out the foundation of the full aluminium analytical model of Mg element, and the graph of relation between the real content of drawing standard sample and the constituent content that calculates by analytical model, shown in Fig. 3 (b);
The analysis result of two kinds of methods of contrast can find out obviously that the analytical model that adopts method of the present invention to set up is significantly improved on the analysis result accuracy than traditional analysis method, and the goodness of fit of analytical model and sample real content is higher; And in whole M g constituent content scope obvious discreteness deviation does not appear, the phenomenon that the independent correlativity of single cover standard model also do not occur, preferably resolve that other interference elements and extraneous shooting conditions have improved precision of analysis significantly to the influence of analysis result in the different samples.
Adopt method of the present invention that sample is analyzed, utilized the useful information of many characteristic spectral lines of respective element in the scanning optical spectrum scope, fully extracted the characteristic information of identity element at the different wave length place, avoided of the influence of external interference factor, made that the analytical model of being set up is more stable the single characteristic spectral line of certain bar;
Simultaneously, same type sample only need be set up the analytical model in the scanning optical spectrum scope, need not artificial selection and carry out the sample size analysis, avoided artificial switching model, make analysis result more accurately, reliably according to the analytical model of certain the bar characteristic spectral line foundation in the spectral coverage; Owing to got rid of of the influence of external interference factor, made analytical model between instrument, transplant and also become very convenient to analytical model.
A kind of spectroscopic analysis methods, different with embodiment 1 described spectroscopic analysis methods is:
In the steps A 3, present embodiment will reflect that one group of light intensity value of characteristic spectral line intensity distributions of the characteristic spectral line of analytical element or internal standard element is as line strength vector of corresponding spectral line, the length of line strength vector of the characteristic spectral line of analytical element is r, the length of line strength vector of the characteristic spectral line of internal standard element is l, r 〉=1, l 〉=1;
Line strength vector of the characteristic spectral line correspondence of m bar analytical element is I
A1 (1 * r), I
A2 (1 * r)... I
Am (1 * r), line strength vector of the characteristic spectral line correspondence of internal standard element is I
B1 (1 * r), I
B2 (1 * r)... I
Bn (1 * r)
In the present embodiment, choose the characteristic spectral line of 2 analytical elements and the characteristic spectral line of 1 internal standard element, m=2 then, n=1 sets r=l=5; Then line strength of analysis spectral line correspondence vector is I
A1 (1 * 5), I
A2 (1 * 5), line strength vector of interior mark spectral line correspondence is I
B1 (1 * 5)
A1, utilize line strength to set up matrix A
P * q
See also Fig. 2, each analysis spectral line and the combination of interior mark spectral line cross match generate A1B1, A2B1, AmB1 ... m * n pairings such as A1Bn, A2Bn, AmBn; Respectively the analysis spectral line in each pairing and line strength vector of interior mark spectral line are compared, obtained the intensity rate set (I of each combinations of pairs
A11/ I
B11, I
A12/ I
B11... I
A13/ I
B11), (I
A21/ I
B11, I
A22/ I
B11... I
A2s/ I
B11) ... (I
Am1/ I
Bn1, I
Am2/ I
Bn1... I
Ams/ I
Bns), the vectorial I of row is formed in all intensity rate set
A/B
The vectorial I of row that the p piece is used for the standard model of modeling
A/BBe combined to form matrix A
P * q, q=m * n * r
2
In the concrete present embodiment, in step a4, the matrix that constructs is respectively: matrix A
30 * 50With matrix Y
30 * 1
To A
30 * 50Decompose, construct 2 order polynomial matrix B
30 * 101With polynomial matrix B
30 * 101As matrix X
30 * 101, to X
30 * 101Extract the score X ' of 6 main genes
30 * 6And Y '
30 * 6, and to X '
30 * 6And Y '
30 * 6Return, preserve X '
30 * 6And Y '
30 * 6Fitting coefficient F
101 * 1, the output analytical model;
Set up element and induce analytical model between content and the relative light intensity, the analytical model that this analytical model and traditional analysis method are drawn compares, and comparing result is as shown in the table:
The actual light intensity value that adopts analysis channel and interior mark passage multi-wavelength place has further been extracted the correlationship of analyzed element and relative light intensity as relative light intensity; Because be the accumulation contribution of a plurality of characteristic value datas, so the variable that extracts is more stable, analysis result is significantly improved than the traditional analysis method.
A kind of spectroscopic analysis methods, different with embodiment 2 described spectroscopic analysis methods is:
1, in steps A 2, from spectral information, only selects the characteristic spectral line of m bar analytical element;
2, in steps A 3, step I) only obtain line strength of each characteristic spectral line of analytical element, line strength vector of m bar analysis spectral line correspondence is I
A1 (1 * r), I
A2 (1 * r)... I
Am (1 * r)
3, in step a1, line strength of analytical element m bar characteristic spectral line constitutes the vectorial I of row
AThe p piece is used for the vectorial I of row of the standard model of modeling
ABe combined to form matrix A
P * q, q=m * r.
Present embodiment in step a2, m=4, r=3, p=25, the characteristic spectral line of selected element M g to be analyzed is four of 279.55nm, 285.2nm, 383.2nm and 383.83nm etc. from spectral information;
In step a1, the matrix that constructs is respectively: matrix A
25 * 36With contain moment matrix y
25 * 1
4, there is not step a2;
5, step a3 is: with matrix A
25 * 36Directly as X
25 * 36
Extract the score of 4 main genes, the proper vector of extraction is X '
25 * 4And Y '
25 * 4, and to X '
25 * 4And Y '
25 * 4Return, preserve X '
25 * 4And Y '
25 * 4Fitting coefficient F
36 * 1, the output analytical model.
Above-mentioned embodiment should not be construed as limiting the scope of the invention.Key of the present invention is: a kind of many characteristic spectral line tectonic analysis models that make full use of in the particular range of wavelengths are provided, have avoided the influence of disturbing factor to analytical model, improved precision of analysis and repeatability.Under the situation that does not break away from spirit of the present invention, any type of change that the present invention is made all should fall within protection scope of the present invention.
Claims (11)
1. spectroscopic analysis methods may further comprise the steps:
A, set up analytical model
Excite standard model, obtain atomic spectrum information;
From described spectral information, select the m bar characteristic spectral line of analytical element, m 〉=2;
According to the spectral line data of above-mentioned m bar characteristic spectral line correspondence and the constituent content of described analytical element/induce content, set up analytical model;
B, analysis unknown sample
Obtain the spectral line data of unknown sample corresponding analysis element according to the method for steps A, and the above-mentioned analytical model of substitution, obtain the content of unknown sample corresponding analysis element.
2. spectroscopic analysis methods according to claim 1 is characterized in that: described analytical model is Y
P * 1=X
P * t* F
T * 1, wherein Y is a constituent content/induce to contain moment matrix, and X is the spectral line data matrix, and F is a matrix of coefficients, and p is the sample number, p 〉=2, t 〉=m.
3. spectroscopic analysis methods according to claim 2 is characterized in that:
A1, the line strength of getting m bar characteristic spectral line are formed the vectorial I of row;
With the vectorial I combination of the row of p piece sample, set up intensity matrix A
P * q, q=m;
A2, utilize matrix A
P * qMake up s order polynomial matrix B
P * t, t=s * q+1, s=1,2,
A3, with matrix B
P * tAs spectral line data matrix X
P * t
4. spectroscopic analysis methods according to claim 3 is characterized in that: by matrix A
P * qMake up matrix B
P * tMethod be:
Matrix B
P * t(s * i-j+1) be listed as each matrix element is a matrix A
P * qI is listed as the j power of corresponding matrix element, i=1, and 2 ..., q; J=1,2 ..., s;
Matrix B
P * t(s * q+1) be listed as each matrix element to be constant.
5. spectroscopic analysis methods according to claim 4 is characterized in that: matrix B
P * t(s * q+1) be listed as each matrix element to be 1.
6. spectroscopic analysis methods according to claim 3 is characterized in that:
In steps A, from described spectral information, select the characteristic spectral line of n bar internal standard element again, n 〉=1;
Step a1 is: with the characteristic spectral line of m bar analytical element and the characteristic spectral line cross match of n bar internal standard element, and get pairing line strength ratio of individual features spectral line afterwards, form the vectorial I of row;
With the vectorial I combination of the row of p piece sample, set up intensity matrix A
P * q, q=m * n.
7. spectroscopic analysis methods according to claim 3 is characterized in that:
Among the step a1, it is line strength vector of r that each wavelength points intensity groups of characteristic spectral line becomes length, line strength Vector Groups of m bar characteristic spectral line vectorial I that embarks on journey, r 〉=1;
With the vectorial I combination of the row of p piece sample, set up intensity matrix A
P * q, q=m * n * r.
8. spectroscopic analysis methods according to claim 7 is characterized in that:
In steps A, from described spectral information, select the characteristic spectral line of n bar internal standard element again, it is line strength vector of l that interior each wavelength points intensity groups of mark spectral line characteristic spectral line becomes length, l 〉=1;
Step a1 is: with the characteristic spectral line of m bar analytical element and the characteristic spectral line cross match of n bar internal standard element, and to the vector element cross match in line strength vector of pairing, and get the ratio of vector element behind the cross match, form the vectorial I of row;
With the vectorial I combination of the row of p piece sample, set up intensity matrix A
P * q, q=m * n * r * l.
9. according to claim 3 or 6 or 7 or 8 described spectroscopic analysis methods, it is characterized in that:
Do not comprise step a2;
Step a3 is: with matrix A
P * qAs spectral line data matrix X
P * t, q=t.
10. according to claim 6 or 8 described spectroscopic analysis methods, it is characterized in that: internal standard element is the sample matrices element.
11. spectroscopic analysis methods according to claim 1 is characterized in that: adopt the offset minimum binary algorithm, set up analytical model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010622421 CN102103080B (en) | 2010-12-31 | 2010-12-31 | Spectrum analyzing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010622421 CN102103080B (en) | 2010-12-31 | 2010-12-31 | Spectrum analyzing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102103080A true CN102103080A (en) | 2011-06-22 |
| CN102103080B CN102103080B (en) | 2013-09-04 |
Family
ID=44156012
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010622421 Expired - Fee Related CN102103080B (en) | 2010-12-31 | 2010-12-31 | Spectrum analyzing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102103080B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103217404A (en) * | 2013-03-30 | 2013-07-24 | 中国科学院安徽光学精密机械研究所 | Method for identifying affiliations of spectrum lines of elements by laser-induced breakdown spectroscopy |
| CN104406920A (en) * | 2014-12-12 | 2015-03-11 | 西南铝业（集团）有限责任公司 | 6082 aluminium alloy spectral standard sample and 6082 aluminium alloy composition detection method |
| CN104458663A (en) * | 2014-12-12 | 2015-03-25 | 西南铝业（集团）有限责任公司 | AHS aluminum alloy spectral standard sample and detection method of AHS aluminum alloy component |
| CN104865228A (en) * | 2015-06-02 | 2015-08-26 | 中国科学院上海技术物理研究所 | Quantitative laser-induced breakdown spectroscopy (LIBS) detecting method based on fusion entropy optimization |
| CN105021546A (en) * | 2015-07-09 | 2015-11-04 | 无锡创想分析仪器有限公司 | Method for measuring chemical elements by whole-spectrum direct-reading spectrometer |
| CN105157953A (en) * | 2015-08-10 | 2015-12-16 | 北京市计量检测科学研究院 | Non-dispersive atomic fluorescence excitation light source purity detection method |
| CN111351778A (en) * | 2020-03-27 | 2020-06-30 | 合肥金星机电科技发展有限公司 | LIBS spectrum-based component analysis method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0535700A2 (en) * | 1991-10-04 | 1993-04-07 | The Perkin-Elmer Corporation | Method and apparatus for comparing spectra |
| CN1880944A (en) * | 2005-09-27 | 2006-12-20 | 重庆大学 | Multi-component analysis method for spectrum |
| CN101846630A (en) * | 2010-04-30 | 2010-09-29 | 浙江出入境检验检疫局检验检疫技术中心 | Method for detecting soluble zinc salt in cosmetic by ultrasound centrifugal-ICP-OES |
-
2010
- 2010-12-31 CN CN 201010622421 patent/CN102103080B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0535700A2 (en) * | 1991-10-04 | 1993-04-07 | The Perkin-Elmer Corporation | Method and apparatus for comparing spectra |
| CN1880944A (en) * | 2005-09-27 | 2006-12-20 | 重庆大学 | Multi-component analysis method for spectrum |
| CN101846630A (en) * | 2010-04-30 | 2010-09-29 | 浙江出入境检验检疫局检验检疫技术中心 | Method for detecting soluble zinc salt in cosmetic by ultrasound centrifugal-ICP-OES |
Non-Patent Citations (2)
| Title |
|---|
| 《冶金分析》 20070430 高红波等 电感耦合等离子体原子发射光谱法测定铌合金中常量及微量元素 第42-45页 1 第27卷, 第4期 * |
| 《分析试验室》 20070430 郭岚等 电感耦合等离子体发射光谱法测定植物油中的磷 58-61 1-11 第26卷, 第4期 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103217404A (en) * | 2013-03-30 | 2013-07-24 | 中国科学院安徽光学精密机械研究所 | Method for identifying affiliations of spectrum lines of elements by laser-induced breakdown spectroscopy |
| CN104406920A (en) * | 2014-12-12 | 2015-03-11 | 西南铝业（集团）有限责任公司 | 6082 aluminium alloy spectral standard sample and 6082 aluminium alloy composition detection method |
| CN104458663A (en) * | 2014-12-12 | 2015-03-25 | 西南铝业（集团）有限责任公司 | AHS aluminum alloy spectral standard sample and detection method of AHS aluminum alloy component |
| CN104865228A (en) * | 2015-06-02 | 2015-08-26 | 中国科学院上海技术物理研究所 | Quantitative laser-induced breakdown spectroscopy (LIBS) detecting method based on fusion entropy optimization |
| CN104865228B (en) * | 2015-06-02 | 2017-08-15 | 中国科学院上海技术物理研究所 | The quantitative LIBS detection method solved based on fusion entropy optimization |
| CN105021546A (en) * | 2015-07-09 | 2015-11-04 | 无锡创想分析仪器有限公司 | Method for measuring chemical elements by whole-spectrum direct-reading spectrometer |
| CN105157953A (en) * | 2015-08-10 | 2015-12-16 | 北京市计量检测科学研究院 | Non-dispersive atomic fluorescence excitation light source purity detection method |
| CN111351778A (en) * | 2020-03-27 | 2020-06-30 | 合肥金星机电科技发展有限公司 | LIBS spectrum-based component analysis method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102103080B (en) | 2013-09-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102103080B (en) | Spectrum analyzing method | |
| Mohamed | Improved LIBS limit of detection of Be, Mg, Si, Mn, Fe and Cu in aluminum alloy samples using a portable Echelle spectrometer with ICCD camera | |
| CN102262076B (en) | Method for laser-induced breakdown spectroscopy element concentration determination based on spectral line combination | |
| Quarles et al. | Fluorine analysis using laser induced breakdown spectroscopy (LIBS) | |
| US9816934B2 (en) | Laser induced breakdown spectroscopy (LIBS) apparatus with automatic wavelength calibration | |
| EP2165168B1 (en) | High-speed spectrographic sensor for internal combustion engines | |
| US20120033212A1 (en) | Laser induced breakdown spectroscopy instrumentation for real-time elemental analysis | |
| US10557792B2 (en) | Spectral modeling for complex absorption spectrum interpretation | |
| CN102410993B (en) | Element measurement method based on laser-induced plasma emission spectral standardization | |
| CN104458701B (en) | Raman spectrum explosive substance identifier automatic calibrating method | |
| Kiser et al. | Mass spectral metastable transitions determined by electric sector variation | |
| Hou et al. | Femtosecond laser ablation molecular isotopic spectrometry for zirconium isotope analysis | |
| CN102410992B (en) | Simplified element measurement method through laser-induced plasma spectral standardization | |
| Johnson | Abundances of 30 elements in 23 metal-poor stars | |
| CN102103079B (en) | Spectrum analysis method | |
| Tempesta et al. | Crystal Chemical Characterisation of Red Beryl by ‘Standardless’ Laser‐Induced Breakdown Spectroscopy and Single‐Crystal Refinement by X‐Ray Diffraction: An Example of Validation of an Innovative Method for the Chemical Analysis of Minerals | |
| Sorrentino et al. | Fast analysis of complex metallic alloys by double-pulse time-integrated laser-induced breakdown spectroscopy | |
| JP2008008801A5 (en) | ||
| JP5204655B2 (en) | Methods for qualitative and quantitative analysis by emission spectroscopy | |
| CN102906851A (en) | Method, computer program, and system to analyze mass spectra | |
| Lymer et al. | UV laser‐induced fluorescence spectroscopy as a non‐destructive technique for mineral and organic detection in carbonaceous chondrites | |
| Wiens et al. | Calibration of the MSL/ChemCam/LIBS remote sensing composition instrument | |
| CN113588597A (en) | Method for improving analysis precision of furnace slag | |
| Singh et al. | Optimization of conditions for determination of Cr and Ni in steel by the method of laser-induced breakdown spectroscopy with the use of partial least squares regression | |
| Pickering et al. | Recent advances in experimental laboratory astrophysics for stellar astrophysics applications and future data needs |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130904 Termination date: 20151231 |
|
| EXPY | Termination of patent right or utility model |