CN104020124B - Based on absorbance light splitting wavelength screening technique preferentially - Google Patents
Based on absorbance light splitting wavelength screening technique preferentially Download PDFInfo
- Publication number
- CN104020124B CN104020124B CN201410233964.0A CN201410233964A CN104020124B CN 104020124 B CN104020124 B CN 104020124B CN 201410233964 A CN201410233964 A CN 201410233964A CN 104020124 B CN104020124 B CN 104020124B
- Authority
- CN
- China
- Prior art keywords
- absorbance
- wavelength
- interval
- light splitting
- screening technique
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Abstract
The invention discloses a kind of based on absorbance light splitting wavelength screening technique preferentially, including: S1, test sample obtain spectroscopic data and index determining value；S2, selection, for the range delta of wavelength screening, determine absorption maximum AmaxWith minima Amin；S3, absorbance step-length ε is set, by (Amin,Amax) n decile；S4, from AminThe starting point of correspondence, AmaxCorresponding terminal and 1 Along ent of n are appointed and takes at 2 and be combined, obtain the interval (A of an absorbance*,A*)；S5, determine (A*,A*Wavelength combinations corresponding to)；S6, according to above-mentioned steps S4, S5, the interval (A of exhaustive all of absorbance*,A*), the wavelength combinations that each absorbance interval is corresponding is set up calibration forecast model, calculates root-mean-square error or correlation coefficient；S7, finding absorbance corresponding to root-mean-square error minima or correlation coefficient maximum interval, wavelength combinations corresponding to this absorbance interval is the selection result of light splitting wavelength.The present invention has the advantage that amount of calculation is few, prediction effect is good.
Description
Technical field
The present invention relates to the wavelength triage techniques field in beam splitting system design, be specifically related to a kind of based on absorption
Rate light splitting wavelength screening technique preferentially.
Background technology
Infrared spectrum analysis is to differentiate material and the method determining its chemical composition and content, and it can need not
Biochemical reagents, have easy quick, non-destructive and are prone to the features such as real-time analysis, having in a lot of fields
The advantage of application.Develop the technology the most comparative maturity of all band universal infrared spectrum instrument at present,
But it has instrument shortcoming huge, expensive, is unsuitable for popularization and application.Therefore, research and development low price
Small, dedicated infrared analytical instrument has application prospect.
When infrared spectrum analysis, high s/n ratio light splitting wavelength screening technique is a key technology, and it is for building
The sides such as the beam splitting system of vertical high accuracy analysis model, reduction model complexity and small design dedicated spectral instrument
Mask is significant.But, the number of wavelengths of infrared band is a lot, if built respectively with any random combine
The mode of mould screens wavelength, and the existing operational speed of a computer far from disclosure satisfy that.Therefore, at infrared light
The selection of analysis of spectrum wavelength, the aspect such as beam splitting system design of small, dedicated spectral instrument there is also difficulty, lack
Weary effective light splitting wavelength screening technique.
Summary of the invention
Present invention is primarily targeted at the shortcoming overcoming prior art with not enough, it is provided that a kind of based on absorbance
Light splitting wavelength screening technique preferentially, the method can filter out the high noise analyzed corresponding to object effectively
Ratio wavelength combinations, has the advantages such as applied range, model is simple, amount of calculation is few, prediction effect is good, for
In small, dedicated analytical tool, the design of beam splitting system proposes effective solution.
The purpose of the present invention is realized by following technical scheme: screen based on absorbance light splitting wavelength preferentially
Method, comprises the following steps:
S1, test sample, obtain the index determining value of spectroscopic data and sample；
S2, in measured spectral band, according to measuring the physics of object, chemical characteristic and spectrum
The performance of instrument, selects low noise and covers the range delta for wavelength screening of indication information, determining this simultaneously
Absorption maximum A that in wave-length coverage, sample average spectrum is correspondingmaxWith minima Amin；
S3, suitable absorbance step-length ε is set, by hypersorption rate scope (Amin,Amax) n decile, obtain
N-1 Along ent；
S4, from terminal corresponding to starting point corresponding to absorbance minima, absorption maximum and n-1 etc.
Branch arbitrarily takes 2 be combined, obtain the interval (A of an absorbance*,A*), wherein
S5, corresponding relation according to the wavelength of spectroscopic data with absorbance, in wavelength screening range delta, really
Interval (the A of this absorbance fixed*,A*Wavelength combinations corresponding to)；
S6, according to above-mentioned steps S4, S5, the interval (A of exhaustive all of absorbance*,A*), each is inhaled
The wavelength combinations foundation calibration forecast model that yield interval is corresponding, calculating spectral predicted value is mean square with measured value
Root error (RMSEP) or correlation coefficient；
S7, find absorbance corresponding to root-mean-square error minima or correlation coefficient maximum interval, by it
Be defined as optimal absorption rate interval, and and then find wavelength combinations corresponding to this optimal absorption rate interval, complete
The screening of light splitting wavelength.
Preferably, in described step S2, wavelength screening range delta is based on measuring the physics of object, chemistry
Characteristic sets for the absorption region of infrared light, and gets rid of the noise wavelength band that instrument operation causes.
Preferably, in described step S3, absorbance step-length ε is based on the spectrum entirety suction that spectrum experiment obtains
Yield value (including absorbance minimum and maximum), model accuracy and modeling operational efficiency are set.
Preferably, in described step S6, calibration forecast model uses offset minimum binary (PLS), multiple linear
Return in the method such as (MLR), principal component analysis (PCA) any one.
Preferably, in described step S7, optimal absorption rate is interval, and corresponding wavelength combinations is based on
The wavelength model being most appropriate for quantitative analysis that the effect of spectral calibration prediction obtains.
The present invention compared with prior art, has the advantage that and beneficial effect:
1, the present invention is by carrying out segmentation by absorbance, then utilizes offset minimum binary (PLS), multiple linear
The calibration forecast model of the mode such as (MLR), principal component analysis (PCA) of recurrence selects optimal absorption rate district
Between, and then effectively filter out and analyze high s/n ratio wavelength combinations corresponding to object, have applied range,
The advantages such as model is simple, amount of calculation is few, prediction effect is good, for beam splitting system in small, dedicated analytical tool
Design proposes effective solution.
2, confirm through many experiments result: the wavelength combinations that the present invention filters out is sieved than the wavelength at they places
Select range delta, it was predicted that effect is obviously improved, owing to the wavelength number used in the present invention greatly reduces, right
In setting up high accuracy analysis model, reducing model complexity and the beam splitting system of small design dedicated spectral instrument
Etc. aspect significant.
3, the present invention is according to the size screening wavelength model of absorbance, it is to avoid the noise of high-absorbility wavelength is big,
The shortcoming that the information of low absorptivity wavelength is weak, has obvious physics, chemical sense.Which overcome routine
Wavelength basis size sequence screen single continuous wave band deficiency, multiple high noise can be chosen according to absorbance
Ratio wave band, has the broader scope of application.
Accompanying drawing explanation
Fig. 1 is the method flow diagram of embodiment 1.
Fig. 2 be in embodiment 1 optimal absorption rate as a example by human serum cholesterol near-infrared analysis interval (0.42,
1.00) schematic diagram of interior wavelength combinations.
Fig. 3 is the wave band one in the wavelength combinations that Fig. 2 chooses.
Fig. 4 is the wave band two in the wavelength combinations that Fig. 2 chooses.
Detailed description of the invention
Below in conjunction with embodiment and accompanying drawing, the present invention is described in further detail, but the embodiment party of the present invention
Formula is not limited to this.
Embodiment 1
The present embodiment, as a example by the NIR transmittance spectroscopy of human serum cholesterol is analyzed, in conjunction with Fig. 1, illustrates this
Bright proposed concrete steps based on absorbance light splitting wavelength screening technique preferentially.
S1, sample is tested, obtain the clinical assays value of spectroscopic data and sample cholesterol index.
S2, the present embodiment are chosen full near infrared band (780～2498nm) and screen range delta as wavelength,
In the range of this wavelength screens, the absorption maximum of the averaged spectrum of all samples, minima are respectively close to 5
With 0, therefore by Amax, AminIt is set to 5 and 0.
S3, to arrange absorbance step-length ε be 0.01, hypersorption rate scope (0,5) is done 500 deciles, obtains
499 Along ents, absorbance corresponding on each Along ent is respectively 0.01, and 0.02 ..., 4.99.
S4, from 0,0.01,0.02 ..., arbitrarily select at 2 in 4.99,5 and be combined, obtain an absorption
Rate is interval, such as shown in Fig. 2, one absorbance interval of selection 0.42 and 1.00 the two absorbances composition (0.42,
1.00)。
S5 is from fig. 2 it can be seen that corresponding two wave band A and B in this absorbance interval, respectively such as Fig. 3
Shown in Fig. 4, in absorbance interval (0.42,1.00), the wave band A having wavelength to be 1374～1392nm,
And the wave band B that wavelength is 1544～1852nm.It is interval (0.42,1.00) that the two wave band is absorbance
Corresponding wavelength combinations.
S6, according to above-mentioned steps S4, S5, the interval (A of exhaustive all of absorbance*,A*), such as can be first
Fix a starting point, the most exhaustive other all Along ents, beginning and ends, change one the most successively and rise
Point, i.e. first from the beginning of 0, takes (0,0.01), (0,0.02), (0,0.03) ..., (0,5), the most again from
0.01 this Along ent starts, and takes (0.01,0.02), (0.01,0.03) ..., (0.01,5), choose successively, directly
Terminate after being combined the most two-by-two to institute.
The wavelength combinations that each absorbance interval above is corresponding is set up offset minimum binary (PLS) calibration prediction
Model, current offset minimum binary (PLS) method is a kind of being widely used and the building of effective infrared spectrum analysis
Mould method, calculates the root-mean-square error of this absorbance interval spectral predicted value and measured value by this method
And correlation coefficient (RMSEP).
S7, by step S6 can obtain totally 500 × 499 × ... × 2 × 1 RMSEP value and correlation coefficient,
From these all RMSEP values, select minima, or from all correlation coefficienies, select maximum, then look for
Interval to its corresponding absorbance, it is defined as optimal absorption rate interval, and and then finds this optimum to inhale
The wavelength combinations that yield interval is corresponding, completes the screening of light splitting wavelength.
The wavelength screening of the present invention is combined the present embodiment PLS method and full near-infrared spectra district PLS method is carried out
Relatively, comparative result is as follows:
Full near-infrared spectra district PLS method: full spectrum wave band is 780～2498nm, uses unknown sample survey to obtain
Predicted root mean square error, correlation coefficient be respectively 0.835 (mmol L-1)、0.677。
The screening of the absorbance of the present invention preferentially wavelength combines PLS method: the optimal absorption rate interval determined is
(0.42,1.00), corresponding band combination is 1374～1392 ∪ 1544～1852, uses unknown sample survey
The predicted root mean square error, the correlation coefficient that obtain are respectively 0.181 (mmol L-1)、0.988。
Experimental result confirms: the absorbance wavelength combinations that preferentially wavelength screening technique filters out based on the present invention
Significantly being better than the prediction effect in full near-infrared spectra district, number of wavelengths significantly reduces, and the method can be according to suction
Yield chooses multiple high s/n ratio wave band, has the broader scope of application, for set up high accuracy analysis model,
The aspects such as the beam splitting system of reduction model complexity and small design dedicated spectral instrument are significant.
Above-described embodiment is the present invention preferably embodiment, but embodiments of the present invention are not by above-mentioned reality
Execute the restriction of example, the change made under other any spirit without departing from the present invention and principle, modification,
Substitute, combine, simplify, all should be the substitute mode of equivalence, within being included in protection scope of the present invention.
Claims (4)
1. based on absorbance light splitting wavelength screening technique preferentially, it is characterised in that comprise the following steps:
S1, test sample, obtain the index determining value of spectroscopic data and sample；
S2, in measured spectral band, according to measuring the physics of object, chemical characteristic and spectrum
The performance of instrument, selects low noise and covers the range delta for wavelength screening of indication information, determining this simultaneously
Absorption maximum A that in wave-length coverage, sample average spectrum is correspondingmaxWith minima Amin；
S3, suitable absorbance step-length ε is set, by hypersorption rate scope (Amin,Amax) n decile, obtain
N-1 Along ent；
S4, from terminal corresponding to starting point corresponding to absorbance minima, absorption maximum and n-1 etc.
Branch arbitrarily takes 2 be combined, obtain the interval (A of an absorbance*,A*), wherein
S5, corresponding relation according to the wavelength of spectroscopic data with absorbance, in wavelength screening range delta, really
Interval (the A of this absorbance fixed*,A*Wavelength combinations corresponding to)；
S6, according to above-mentioned steps S4, S5, the interval (A of exhaustive all of absorbance*,A*), each is inhaled
The wavelength combinations foundation calibration forecast model that yield interval is corresponding, calculating spectral predicted value is mean square with measured value
Root error or correlation coefficient；
S7, find absorbance corresponding to root-mean-square error minima or correlation coefficient maximum interval, by it
Be defined as optimal absorption rate interval, and and then find wavelength combinations corresponding to this optimal absorption rate interval, complete
The screening of light splitting wavelength.
The most according to claim 1 based on absorbance light splitting wavelength screening technique preferentially, its feature exists
In, in described step S2, wavelength screening range delta be based on measure the physics of object, chemical characteristic for
The absorption region of infrared light sets, and gets rid of the noise wavelength band that instrument operation causes.
The most according to claim 1 based on absorbance light splitting wavelength screening technique preferentially, its feature exists
In, in described step S3, absorbance step-length ε be based on spectrum experiment obtain spectrum overall absorption rate value,
Model accuracy and modeling operational efficiency are set.
The most according to claim 1 based on absorbance light splitting wavelength screening technique preferentially, its feature exists
In, in described step S6, calibration forecast model uses offset minimum binary, multiple linear regression, main constituent to divide
Any one in analysis method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410233964.0A CN104020124B (en) | 2014-05-29 | 2014-05-29 | Based on absorbance light splitting wavelength screening technique preferentially |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410233964.0A CN104020124B (en) | 2014-05-29 | 2014-05-29 | Based on absorbance light splitting wavelength screening technique preferentially |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104020124A CN104020124A (en) | 2014-09-03 |
| CN104020124B true CN104020124B (en) | 2016-08-24 |
Family
ID=51437002
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410233964.0A Active CN104020124B (en) | 2014-05-29 | 2014-05-29 | Based on absorbance light splitting wavelength screening technique preferentially |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104020124B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105334181B (en) * | 2014-10-22 | 2018-05-04 | 北京市农林科学院 | The quick determination method of Irradiation On Peanut |
| CN106918567B (en) * | 2017-03-27 | 2019-05-28 | 中南大学 | A kind of method and apparatus measuring trace metal ion concentration |
| CN112461782B (en) * | 2019-10-17 | 2022-11-01 | 山东金璋隆祥智能科技有限责任公司 | Spectrum correction technology based on GSA near-infrared spectrometer |
| CN111474128B (en) * | 2020-05-12 | 2023-03-14 | 暨南大学 | Spectral wavelength combination method based on spectral separation degree |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3768634B2 (en) * | 1997-01-31 | 2006-04-19 | 日本分光株式会社 | Calculation wavelength range selection method for secondary structure analysis of proteins |
| JPH11173915A (en) * | 1997-12-10 | 1999-07-02 | Sumitomo Metal Ind Ltd | Infrared absorbance measuring method |
| CN101806728B (en) * | 2010-03-12 | 2011-12-21 | 江苏大学 | Method for selecting characteristic wavelength of near-infrared spectrum based on simulated annealing algorithm |
| CN101825567A (en) * | 2010-04-02 | 2010-09-08 | 南开大学 | Screening method for near infrared spectrum wavelength and Raman spectrum wavelength |
-
2014
- 2014-05-29 CN CN201410233964.0A patent/CN104020124B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN104020124A (en) | 2014-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104062257B (en) | A kind of based on the method for general flavone content near infrared ray solution | |
| Han et al. | Determination of glycated hemoglobin using near-infrared spectroscopy combined with equidistant combination partial least squares | |
| CN104020124B (en) | Based on absorbance light splitting wavelength screening technique preferentially | |
| CN104062258B (en) | Method for rapid determination of soluble solids in compound ass-hide glue pulp by near infrared spectroscopy | |
| CN106932378A (en) | The on-line detecting system and method for a kind of sour gas composition based on Raman spectrum | |
| CN106018331B (en) | The method for estimating stability and pretreatment optimization method of multi-channel spectral system | |
| CN105203498A (en) | Near infrared spectrum variable selection method based on LASSO | |
| CN106770058A (en) | The quick special purpose device and its application method of the soil nitrate-N based on infrared spectrum | |
| CN104062259A (en) | Method for rapid determination of total saponin content in compound ass-hide glue pulp by near infrared spectroscopy | |
| CN105044025A (en) | Method for fast recognizing sesame oil and sesame oil doped with soybean oil through near infrared | |
| CN101408501A (en) | Method for quantitatively detecting DNA base by using near-infrared spectrum-partial least squares method | |
| CN101349638A (en) | Optical spectrum rapid nondestructive detection method of fruit and vegetable vitamin C content | |
| CN105136682B (en) | Arene content is quick in a kind of fuel oil, the new method of Accurate Determining | |
| Li et al. | Fast determination of four active compounds in Sanqi Panax Notoginseng Injection samples by high‐performance liquid chromatography with a chemometric method | |
| CN101788459B (en) | Quasi-continuous spectroscopic wavelength combination method | |
| CN103105369B (en) | Fluent meterial spectrum baseline corrects quantitative analysis method | |
| Nishii et al. | Selection of the NIR region for a regression model of the ethanol concentration in fermentation process by an online NIR and mid-IR dual-region spectrometer and 2D heterospectral correlation spectroscopy | |
| CN103134770B (en) | Eliminate moisture detects total nitrogen content of soil impact method near infrared spectrum | |
| CN106525755A (en) | Oil-sand pH value testing method based on near infrared spectroscopy technology | |
| CN109100315B (en) | Wavelength selection method based on noise-signal ratio | |
| CN113686811B (en) | Spectral data processing method based on double sensors | |
| Zhang et al. | Optimal wavelength selection for visible diffuse reflectance spectroscopy discriminating human and nonhuman blood species | |
| CN105806803A (en) | Multi-index collaborative analysis wavelength combination and selection method thereof | |
| Sleiman et al. | Qualitative and quantitative analysis of explosives by terahertz time-domain spectroscopy: application to imaging | |
| CN108489928B (en) | Method for detecting textile fiber components by short-wave infrared extinction spectrum |
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 |