CN108061724A - The lossless rapid detection method of Xinjiang coloured silk cotton interior quality near infrared spectrum - Google Patents
The lossless rapid detection method of Xinjiang coloured silk cotton interior quality near infrared spectrum Download PDFInfo
- Publication number
- CN108061724A CN108061724A CN201711267876.2A CN201711267876A CN108061724A CN 108061724 A CN108061724 A CN 108061724A CN 201711267876 A CN201711267876 A CN 201711267876A CN 108061724 A CN108061724 A CN 108061724A
- Authority
- CN
- China
- Prior art keywords
- calibration
- sample
- coloured silk
- cotton
- quantitative
- 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.)
- Pending
Links
- 244000146553 Ceiba pentandra Species 0.000 title claims abstract description 77
- 238000002329 infrared spectrum Methods 0.000 title claims abstract description 45
- 238000001514 detection method Methods 0.000 title claims abstract description 21
- 229920000742 Cotton Polymers 0.000 claims abstract description 69
- 238000001228 spectrum Methods 0.000 claims abstract description 29
- 238000000034 method Methods 0.000 claims abstract description 27
- 238000005457 optimization Methods 0.000 claims abstract description 4
- 238000012795 verification Methods 0.000 claims description 46
- 238000004497 NIR spectroscopy Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 239000012535 impurity Substances 0.000 claims description 3
- 238000004164 analytical calibration Methods 0.000 claims description 2
- 238000013459 approach Methods 0.000 claims description 2
- 238000010238 partial least squares regression Methods 0.000 claims description 2
- 239000000835 fiber Substances 0.000 abstract description 16
- 239000000463 material Substances 0.000 abstract description 7
- 238000012545 processing Methods 0.000 abstract description 5
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000012512 characterization method Methods 0.000 abstract description 3
- 230000001737 promoting effect Effects 0.000 abstract description 2
- 241000219146 Gossypium Species 0.000 description 51
- 238000012360 testing method Methods 0.000 description 11
- 238000004458 analytical method Methods 0.000 description 10
- 238000009395 breeding Methods 0.000 description 10
- 230000001488 breeding effect Effects 0.000 description 10
- 238000009826 distribution Methods 0.000 description 7
- 238000005259 measurement Methods 0.000 description 7
- 238000004445 quantitative analysis Methods 0.000 description 7
- 238000011160 research Methods 0.000 description 5
- 230000003595 spectral effect Effects 0.000 description 4
- 239000000126 substance Substances 0.000 description 4
- 229920002678 cellulose Polymers 0.000 description 3
- 239000001913 cellulose Substances 0.000 description 3
- 235000012343 cottonseed oil Nutrition 0.000 description 3
- 239000002385 cottonseed oil Substances 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 244000241235 Citrullus lanatus Species 0.000 description 2
- 235000012828 Citrullus lanatus var citroides Nutrition 0.000 description 2
- 240000008042 Zea mays Species 0.000 description 2
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 2
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 2
- 230000004888 barrier function Effects 0.000 description 2
- 235000005822 corn Nutrition 0.000 description 2
- 230000008676 import Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 108090000623 proteins and genes Proteins 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 244000068988 Glycine max Species 0.000 description 1
- 235000010469 Glycine max Nutrition 0.000 description 1
- 229920002488 Hemicellulose Polymers 0.000 description 1
- 244000131316 Panax pseudoginseng Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 244000046052 Phaseolus vulgaris Species 0.000 description 1
- 235000010627 Phaseolus vulgaris Nutrition 0.000 description 1
- 235000021307 Triticum Nutrition 0.000 description 1
- 244000098338 Triticum aestivum Species 0.000 description 1
- 238000002835 absorbance Methods 0.000 description 1
- 238000000862 absorption spectrum Methods 0.000 description 1
- AEMOLEFTQBMNLQ-BKBMJHBISA-N alpha-D-galacturonic acid Chemical group O[C@H]1O[C@H](C(O)=O)[C@H](O)[C@H](O)[C@H]1O AEMOLEFTQBMNLQ-BKBMJHBISA-N 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 150000001720 carbohydrates Chemical group 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 239000000460 chlorine Chemical group 0.000 description 1
- 229910052801 chlorine Inorganic materials 0.000 description 1
- 125000001309 chloro group Chemical group Cl* 0.000 description 1
- 239000000084 colloidal system Substances 0.000 description 1
- 238000004040 coloring Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000000205 computational method Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 235000021393 food security Nutrition 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 241000411851 herbal medicine Species 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 102000004169 proteins and genes Human genes 0.000 description 1
- 238000004451 qualitative analysis Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000004611 spectroscopical analysis Methods 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 230000009261 transgenic effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3563—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing solids; Preparation of samples therefor
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/359—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light using near infrared light
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Treatment Of Fiber Materials (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The present invention relates to a kind of lossless rapid detection methods of Xinjiang coloured silk cotton interior quality near infrared spectrum, this method is by the foundation of the measure of sample parameters value, sample sets classification, sample near infrared spectrum scanning and sample sets, editor's brown coloured silk cotton each sample collection, establishes quantitative calibration models, optimization quantitative calibration models and the completion of quick detecting step, pass through computerized near infrared scan cotton fiber, the spectrum based on content of material is obtained, feasible means are provided for characterization fiber quality.It solves the key technology difficulty that color cotton runs into terms of purchasing with quality identification in processing link at present, is worth promoting and applying in color cotton quality identification.
Description
Technical field
The present invention relates to a kind of lossless rapid detection methods of Xinjiang coloured silk cotton interior quality near infrared spectrum.
Background technology
Xinjiang is carried out in the leading position in color cotton research and development field, many color cotton research institutions in Xinjiang and enterprise
A large amount of fruitful work such as conventional breeding, molecular mark, transgenic breeding, end 2012 years, Xinjiang is common
Color cotton new varieties 23 are authorized, the fiber quality of new varieties is all significantly improved.But colorful cotton fibre quality-improving according to
It is so the hot and difficult issue problem in current color cotton breeding work.
Because fiber quality is the quantitative character of controlled by multiple genes, quality can not intuitively embody, it is necessary to specific instrument
Device；Expensive, the service requirement constant temperature and humidity currently used for color cotton fiber attribute test instrument, and, inspection low in my area's ownership
It surveys costly；Therefore many breeders have to the quantity for significantly compressing species test material during the selection and breeding of field,
Just it is present with leakage sieve during this, wastes hard-earned intermediate materials.Therefore there is an urgent need to one kind for Breeding of Color Cotton work
Simply, fast, low cost detection fiber quality method, solve fiber quality detection common problem.
The problem of existing detection means is insufficient during colorful cotton fibre quality breeding and many are engaged in wheat, big
The same problems that the researcher of the crop breedings such as beans, corn, watermelon work faces.
For limitation and deficiency existing for current color cotton fiber quality determination instrument, propose to utilize near infrared spectroscopy instrument
The method that fast quantitative analysis is carried out at the same time to a variety of qualities of color cotton.It this method solve current Xinjiang coloured silk cotton Quality Detection side
The problems such as tester is expensive existing for method, test laboratory environment requirement constant temperature and humidity, testing cost is high.
Near-infrared spectral analytical method is that the frequency multiplication of X-H groups (X=H, C, N, O) or sum of fundamental frequencies absorb in recorded matter,
Qualitative and quantitative analysis can be carried out by the method processing of Chemical Measurement.Different from general analysis method, its application needs
Quantitative or qualitative mathematics model is established in advance.
The flow of Near-Infrared Spectra for Quantitative Analysis is divided into two big steps：That is founding mathematical models (analysis method, prediction side
Journey) and examine, the stability of Optimized model；And applied mathematical model, using the near infrared spectrum of unknown sample, prediction is not
Know content or property in relation to component in sample.
The near infrared spectrum of measured matter depends on the Nomenclature Composition and Structure of Complexes of sample in near-infrared spectrum analysis.The group of sample
Into there is certain functional relation between structure and near infrared spectrum.Determine that these are important using chemometrics method
Functional relation, i.e., it is corrected, it is possible to according to the near infrared spectrum of sample, quickly to calculate the content of various components.
Near-infrared spectral analytical method has compared with other analysis methods：Analyze speed is fast, and measurement process mostly can be complete in 1min
Into；Analysis efficiency is high, can be simultaneously to multiple components of sample by a spectral measurement and established corresponding calibration model
Or property is measured, and provides qualitative, quantitative result；Applicable sample scope is wide, can be direct by corresponding test sample device
The sample spectra measurement for measuring the different states of matters such as liquid, solid, semisolid and colloid is convenient；Sample need not generally be located in advance
Reason, need not use the test conditions such as chemical reagent or high temperature, high pressure, high current, will not be generated after analysis chemistry, biology or
Electromagnetic pollution；Analysis cost it is relatively low (without numerous and diverse pretreatment, can multicomponent detect simultaneously, test favorable reproducibility；To sample without
Damage does not influence the two time selling of sample；Requirement to operating personnel is not harsh, can be competent by simply training
Etc. exclusive technical advantage, therefore near infrared spectrum, in food security, the fields such as food quality and Chinese herbal medicine quality identification obtain
Extensive use is arrived.
Up to the present, the document using the lossless quick detection coloured silk cotton interior quality of nearly red spectral is had no, but in spectroscopy
With finding one " cotton seed oil content near-infrared Nondestructive Detection model and application " on spectrum analysis magazine, the document is ground
The object studied carefully is the cottonseed oil of land white cotton, and the content of land white cottonseed oil is carried out using near-infrared spectrum technique
Lossless, quick detection.And the present invention relates to the brown color cotton fiber index of quality, the object domain and document of research are completely
It is different.
Numerous researchs show：Colorful cotton fibre is mainly by cellulose, hemicellulose, cellulose accompaniment and coloring matter
Composition, cellulose accompaniment grade including soluble saccharide, wax, fat, protein, pectic substance, chlorine-bearing compound and ash, remove
Outside ash content, containing X-H groups (X=H, C, N, O) in other materials, this is just to predict this in fiber using near infrared spectrum
The content of a little substances has established material base.
What is more important, by numerous studies, fiber quality is also closely related with the content of these substances.By near red
Outer optical scanning cotton fiber, obtains the spectrum based on content of material, and feasible means are provided for characterization fiber quality.
Research shows that the qualitative character of crop and inherent content of predetermined substance are closely related, no matter watermelon, corn and soybean
Or cotton considers that, all containing X-H groups (X=H, C, N, O), these groups can be reflected as from the most small structure of matter
The difference of near infrared spectrum, therefore, Different Crop quality trait near-infrared assistant breeding for the angle of near infrared spectrum,
With general character, can use for reference.
The content of the invention
Present invention aims at, provide a kind of Xinjiang coloured silk cotton interior quality near infrared spectrum lossless rapid detection method, it should
Method be classified by the measure of sample parameters value, sample sets, foundation, the editor palm fibre of sample near infrared spectrum scanning and sample sets
Color cotton each sample collection establishes quantitative calibration models, optimization quantitative calibration models and the completion of quick detecting step, by near red
Outer optical scanning cotton fiber, obtains the spectrum based on content of material, and feasible means are provided for characterization fiber quality.Solution
Determined current color cotton purchase and processing link in the key technology difficulty that runs into terms of quality identification, be worth in color cotton quality
It is promoted and applied in identification.
A kind of lossless rapid detection method of Xinjiang coloured silk cotton interior quality near infrared spectrum of the present invention, follows these steps
It carries out：
A, the measure of sample parameters value：The brown coloured silk cotton that impurity removing is cleaned out, is numbered, with fibrometer system point
The other interior quality long to mic value, intensity, elongation and the upper half of each brown cotton sample product is analyzed, and obtains palm fibre
The detection data of color cotton interior quality, i.e. quality determination value；
B, sample sets are classified：The sample of the brown coloured silk cotton interior quality measured value obtained in step a is divided into correcting sample
Collection, verification sample collection and Prediction, Calibration and verification sample collection are used for establishing quantitative calibration models, pre- test sample
Product collection is used for evaluating the accuracy and validity of the model, and wherein Calibration accounts for the 60% of sample total amount, verification sample collection
For 20%, Prediction 20%；
C, the foundation of sample near infrared spectrum scanning and sample sets：Step b is obtained respectively near infrared spectroscopy instrument
Calibration, verification sample collection and the Prediction of brown coloured silk cotton carry out near infrared spectrum scanning, respectively obtain brown coloured silk
Cotton sample product Calibration, verification sample collection and Prediction near infrared spectrum scanning figure；
D, brown coloured silk cotton each sample collection is edited：By the sample sets obtained in step c since Calibration, by all schools
The collection of illustrative plates of positive sample sets infrared diaphanoscopy according to Calibration serial number order, is imported near infrared spectrometer in batches
" model and sample sets management " section, establishes brown cotton Calibration in this section, and to the Calibration into edlin：
Calibration is opened, the mic value of brown coloured silk cotton sample product, intensity, elongation in step a are added on Calibration section
The title and its corresponding quality determination value of rate and the long interior quality of upper half, the correcting sample editted are concentrated, each sample
Include atlas of near infrared spectra and corresponding four kinds of quality determination values；
It establishes brown coloured silk cotton verification sample collection and Prediction is carried out as the step of Calibration, by all samples
After collection establishes, in " sample sets management " section, it is seen that established brown coloured silk cotton Calibration, verification sample collection and pre-
Sample collection；
E, quantitative calibration models are established：In near infrared spectrometer " model management " section, by established correction in step d
Sample sets and verification sample collection are imported into model foundation software, with Partial Least Squares Regression quantitative approach, selection
1100nm-2400nm bands are modeling wave band, and under the conditions of the modeling parameters of system default and Pretreated spectra, it is fixed to establish
Measure calibration model, modeling software using import and the modeling configuration condition of selection under calculate quantitative calibration models automatically, and give
Go out the prediction result figure of quantitative calibration models and corresponding model parameter, four kinds of brown coloured silk cotton is obtained in " model management " section
The quantitative calibration models of quality；
F, quantitative calibration models are optimized：Using modeling software editting function, to the four kinds of product of brown coloured silk cotton obtained in step e
Matter quantitative calibration models are modified and are optimized, and model optimization process is：Only change modeling wave band every time, other modeling configurations
It is remained unchanged with Pretreated spectra mode, establishes the quantitative calibration models of different-waveband, quantified by comparing each modeling wave band
The evaluating of calibration model：Calibration set standard deviation, cross-verification collection standard deviation and forecast set standard deviation and reference are quantitative
Calibration model calibration set predicted value and measured value related coefficient, select calibration set standard deviation, K=fold cross-verification collection standards
Difference and forecast set standard deviation are minimum, the modeling ripple of quantitative calibration models calibration set predicted value and measured value related coefficient maximum
Two bands of section 1300nm-1700nm and 1800nm-2400nm model wave band for four kinds of qualities of brown coloured silk cotton；Then select
It is constant to select best modeled wave band, only changes Pretreated spectra mode, establishes the quantitative correction mould under different Pretreated spectra modes
Type, it is similary to select calibration set standard deviation by comparing the evaluating of quantitative calibration models under each Pretreated spectra mode,
Cross-verification collection standard deviation and forecast set standard deviation are minimum, quantitative calibration models calibration set predicted value and measured value related coefficient
Maximum Savitzky-Golay is smooth, and Savitzky-Golay derivative spectrums pretreatment mode is four kinds of qualities of brown coloured silk cotton
Pretreated spectra mode is modeled, finally with two region modeling wave bands of 1300nm-1700nm and 1800nm-2400nm and light
Savitzky-Golay is smooth, and Savitzky-Golay derivatives spectrum pretreatment mode establishes quantitative calibration models, which is
Four kinds of quantitative calibration models best in quality of brown coloured silk cotton；
G, brown coloured silk cotton Prediction 0.1g-1.5g in step b is placed near infrared spectrometer sample preheated in advance
In product disk, through near infrared spectrum scanning, the optimal quantitative calibration models that are obtained in steps for importing f are to get to brown coloured silk cotton sample product
The prediction result of corresponding interior quality.
The lossless rapid detection method of Xinjiang coloured silk cotton interior quality near infrared spectrum of the present invention, brown is color in this method
Cotton interior quality index：
Upper half is long：Brown coloured silk cotton upper half mean length upper-half-mean length brown color cotton fiber length
During measure, weight accounts for the average length of the radical of the longer fibers part of fibre bundle half (50%), and it is equal to be commonly abbreviated as upper half
Long, unit is millimeter；
Mic value：The a certain amount of brown color cotton fibers of mic value micronaire value under prescribed conditions ventilative
The measurement of property, is the overall target of cotton fiber fineness and maturity and determines that this index does not have unit.
Regularity：Uniformity index uniformity index, during brown color cotton fiber length measurment, average length with it is upper
The ratio between half portion average length represents that unit is represented with percentage (%) with the percentage of upper half mean length.
Elongation：Length when color cotton fiber is stretched to straight line, unit are represented with (%).
These four above-mentioned indexs are that must examine to refer in the most important index of quality of color cotton and color cotton quality grading inspection
Mark is purchased in color cotton, and processing and color cotton breeding field will carry out color cotton fiber quality according to the measured value of these indexs
Classification.Therefore a kind of simple and direct, quick, colleges and universities are established, be suitble to the quantitative detecting method of the high-volume index of quality color cotton breeding,
Purchase and manufacture field have Great significance.
The quality parameter value of color cotton sample product (brown coloured silk cotton) be with USTER fibrometer systems the laboratory of constant temperature and humidity into
What row measured, all coloured silk cotton sample product and its quality parameter measured value are provided by Xinjiang Cai Mian research institutes.Brown coloured silk cotton sample number
It measures as 7356.
Description of the drawings
Fig. 1 is the original atlas of near infrared spectra of brown coloured silk cotton calibration set of the present invention；
Fig. 2 is the optimal main cause subnumber distribution map of the long quantitative calibration models of brown coloured silk cotton upper half of the present invention；
Fig. 3 for PLS calibration models of the present invention to the long prediction result figure of brown cotton upper half, wherein-О-correcting sample
Collection ,-Δ-verification sample collection；
Fig. 4 is the optimal main cause subnumber distribution map of brown coloured silk cotton mic value quantitative calibration models of the present invention；
Fig. 5 for PLS calibration models of the present invention to brown cotton mic value prediction result figure, wherein-О-correcting sample
Collection ,-Δ-verification sample collection；
Fig. 6 is the optimal main cause subnumber distribution map of brown coloured silk cotton elongation quantitative calibration models of the present invention；
Fig. 7 is the prediction result figure that PLS calibration models of the present invention extend brown cotton, wherein-О-Calibration ,-
Δ-verification sample collection；
Fig. 8 is the optimal main cause subnumber distribution map of brown coloured silk cotton regularity quantitative calibration models of the present invention；
Fig. 9 is prediction result figure of the PLS calibration models of the present invention to brown cotton regularity, wherein-О-correcting sample
Collection ,-Δ-verification sample collection.
Specific embodiment
Embodiment
The measure of sample parameters value：The brown coloured silk cotton sample product that impurity removing is cleaned out, each weigh 0.1g-1.5g, into
Row number, with fibrometer system respectively in mic value, intensity, elongation and the upper half length of each brown cotton sample product
It is analyzed in the index of quality, obtains the detection data i.e. product of the long interior quality of mic value, intensity, elongation and upper half
Matter index determining value；
Sample sets are classified：The interior quality measured value sample obtained in step a is divided into Calibration, verification sample collection
And Prediction, Calibration and verification sample collection are used for establishing quantitative calibration models, Prediction is used for evaluating this
The accuracy and validity of model, wherein Calibration account for the 60% of sample total amount, and verification sample collection is 20%, pre- test sample
Product collection is 20%；Brown coloured silk cotton Calibration, verification sample concentrate the distribution situation distribution of each index of quality measured value to be shown in Table
1：
1 brown coloured silk cotton calibration set of table and verification collect the distribution of each quality content
The foundation of sample near infrared spectrum scanning and sample sets：It (is produced near infrared spectroscopy instrument in Chinese optically focused science and technology
Near infrared spectrometer SUPNIR 2700) respectively to the Calibration of the obtained brown coloured silk cottons of step b, verification sample collection and pre-
Sample collection carries out near infrared spectrum scanning, takes brown coloured silk cotton sample product 0.1g-1.5g every time, is placed in the rotation of near infrared spectrum
In diffusing reflection sampler tray, using background built in instrument as reference, with Rotating with Uniform pattern, color cotton sample product near infrared spectrum is gathered, often
Twice, sweep parameter is set a Sample Scan：Spectra collection scope 1000nm-2400nm, scanning times 32 times, resolution ratio 8cm-
1, a data point is gathered per 4cm, gathers 1500 data altogether, scans 22 DEG C of temperature, brown coloured silk cotton Calibration and verification
For sample sets infrared spectrum stacking chart as shown in Figure 1, abscissa is wavelength (nm), ordinate is absorbance (A)；
Edit brown coloured silk cotton each sample collection：Since calibration set, the collection of illustrative plates of all calibration set sample infrared diaphanoscopies is pressed
According to Calibration serial number order, " model and sample sets management " section is imported near infrared spectrometer in batches, in this section
Brown cotton Calibration is established in face, and to the Calibration into edlin：Calibration is opened, is cut in Calibration
The long interior quality of mic value, intensity, elongation and upper half of brown cotton sample product and its corresponding product in step a are added on face
Matter measured value, the correcting sample editted are concentrated, and each sample should include atlas of near infrared spectra and corresponding quality determination value；
It establishes brown coloured silk cotton verification sample collection and Prediction is carried out as the step of Calibration, cut in " sample sets management "
See brown cotton Calibration, verification sample collection and Prediction in face；
Establish quantitative calibration models：The RIMP.P003.V01B.001 softwares carried using near infrared spectrometer, from brown
In the near infrared spectrum scanning collection of illustrative plates of color cotton Calibration it can be seen that, it is evident that characteristic absorption peak：Exist respectively
At 1300nm-1670nm and at 1870-2220nm and at 2220nm-2430nm, chemometrics method-minimum two is utilized
Multiply recurrence computational methods, by these near infrared spectrum characteristic absorption peak datas of brown coloured silk cotton and its interior quality mic value, on
Partly long, intensity establishes corresponding relation between elongation, in the Pretreated spectra side of near infrared spectrometer modeling software acquiescence
Method and selection selection all band, extract the information of efficient association between color cotton interior quality and its near-infrared absorption spectrum, establish
The quantitative calibration models of corresponding quality parameter；
Optimize quantitative calibration models：In color cotton near infrared spectrum containing random noise, baseline drift, sample be uneven, light
Interference caused by the factors such as scattering can eliminate various noises and interference with rational Pretreated spectra means, and it is near to extract color cotton
The characteristic information of infrared spectrum improves the stability and precision of prediction of calibration model, utilizes the light for changing near infrared spectrum and carrying
Preprocess method modeling parameters are composed, edit-modify are carried out to the quantitative calibration models built up respectively, according to fixed in editing process
Measure the calibration set standard deviation (SEC) of calibration model, K=fold cross-verification collection standard deviations (SECV) and verification collection standard deviation
(SEP) size of value and degree close to each other continuously edit calibration model, wherein the calibration set standard of calibration model
Poor (SEC), K=fold cross-verification collection standard deviations (SECV) and verification collection standard deviation (SEP) value smaller (closer to zero) and
It is closer better between each other simultaneously, when calibration set predicted value and measured value related coefficient (RC) closer 1, the school established
The prediction effect of positive model is preferable, according to mentioned above principle, only changes modeling wave band every time, other modeling configurations are located in advance with spectrum
Reason mode remains unchanged, and establishes the quantitative calibration models of different-waveband, by comparing each modeling wave band quantitative calibration models
Evaluating：Calibration set standard deviation, cross-verification collection standard deviation and forecast set standard deviation and reference quantitative calibration models school
Positive collection predicted value and measured value related coefficient, select calibration set standard deviation, cross-verification collection standard deviation and forecast set standard deviation
Minimum, the modeling wave band 1300nm-1700nm of quantitative calibration models calibration set predicted value and measured value related coefficient maximum and
Two bands of 1800nm-2400nm model wave band for four kinds of qualities of brown coloured silk cotton；Then best modeled wave band is selected not
Become, only change Pretreated spectra mode, establish the quantitative calibration models under different Pretreated spectra modes, it is each by comparing
The evaluating of quantitative calibration models under Pretreated spectra mode, it is similary to select calibration set standard deviation, cross-verification collection standard deviation
It is minimum with forecast set standard deviation, the Savitzky- of quantitative calibration models calibration set predicted value and measured value related coefficient maximum
Golay is smooth, and Savitzky-Golay derivative spectrums pretreatment mode models Pretreated spectra side for four kinds of qualities of brown coloured silk cotton
Formula, it is finally smooth with two region modeling wave bands of 1300nm-1700nm and 1800nm-2400nm and light Savitzky-Golay,
Savitzky-Golay derivatives compose pretreatment mode, establish quantitative calibration models, and the model is best in quality for four kinds of brown coloured silk cotton
Quantitative calibration models；The method screening that a kind of fixed pretreatment item and other pretreatment items are respectively combined every time is most preferably located in advance
Reason method by being tested with processing method combined sorting, finally determines that Savitzky-Golay is smooth, Savitzky-Golay
Derivative, both preprocess methods are combined as the final preprocess method of interior quality calibration model in color cotton four；Finally use
At 1300nm-1700nm and at 1800-2400nm, two bands and Savitzky-Golay are smooth, Savitzky-
Golay derivative spectrum preprocess methods establish the quantitative calibration models of brown coloured silk four kinds of qualities of cotton, which is total four kinds of color
The optimal quantitative calibration models of the index of quality；
Principle is evaluated using calibration model, respectively to four kinds of crucial interior qualities of brown coloured silk cotton：Mic value, intensity are stretched
It is equal that long rate and the long quantitative calibration models progress of upper half respectively obtain mic value, intensity, elongation and upper half after further optimizing
The optimal quantitative calibration models of long brown coloured silk cotton interior quality index, brown coloured silk cotton mic value, intensity, elongation and upper half
The sum of the prediction result figure of long interior quality quantitative calibration models and cross-verification square-error (PRESS) figure, are shown in Fig. 2-figure
9, wherein Fig. 2, Fig. 4, Fig. 6, Fig. 8 is (PRESS) figure, is the selection gist of quantitative calibration models main cause subnumber, minimum in figure
The corresponding main cause subnumber of PRESS values is the optimal main cause subnumber of quantitative calibration models；Fig. 3, Fig. 5, Fig. 7, Fig. 9 are brown coloured silk cotton
The prediction result figure of quality quantitative calibration models in four；
From Fig. 3, Fig. 5, Fig. 7, Fig. 9, brown coloured silk cotton upper half length, mic value, elongation, in regularity four
There are linear relationship between the predicted value and measured value of the near-infrared quantitative calibration models of the index of quality, quantitative model is corrected
Calibration set standard deviation (SEC), verification collection standard deviation (SEP), K=fold cross-verification collection standard deviation (SECV) value is smaller, together
When the mutual gap of these parameter values it is also smaller, three parameter values are relatively close together, and accord with K=fold cross-verification collection standard deviations
(SECV), verification collection standard deviation (SEP) value >=SEC*1.5, the measured value related coefficient (RC) of four kinds of interior quality calibration models
Value is all higher than 0.92, illustrates that these four interior quality calibration models of brown coloured silk cotton have preferably predictive ability, what model obtained
Correlation between predicted value and measured value is very good, brown coloured silk cotton sample product near infrared spectrum and brown coloured silk cotton interior quality it
Between there are linear correlation, reached preferable prediction result, show this method in four kinds of quantitative analysis brown coloured silk cotton in product
Matter index is feasible；
The long interior quality quantitative correction mould of brown coloured silk cotton mic value, intensity, elongation and upper half that the present invention establishes
Type estimated performance evaluation result is shown in Table 2：
2 brown coloured silk cotton PLS quantitative calibration models estimated performance evaluation results of table
Brown coloured silk cotton mic value, intensity, elongation and the long interior quality ginseng near-infrared quantitative calibration models SEP of upper half
It disclosure satisfy that the quantitative analysis error requirements of color cotton quality, the quantitative calibration models of foundation are applied to Xinjiang brown coloured silk cotton mark
In grand value, intensity, elongation and long the quick of interior quality of upper half, nondestruction quantitative analysis；
Evaluation to the estimated performance of model：It is optimal to import brown coloured silk cotton for the model evaluation software carried using instrument
Quantitative calibration models, the Prediction editted, near infrared spectrometer analysis system software will just provide less than one minute
The prediction result of quality and predicted value (are detected with measured value with Wu Site fibrometer systems in the four of all forecast sets imported
Obtained measured value) between absolute deviation；Quantitative calibration models are to brown coloured silk cotton mic value, intensity, elongation and upper half
The prediction result of these four interior qualities is grown respectively at the survey measured with four kinds of interior qualities with Wu Site fibrometer systems
Definite value is compared, and statistical result is listed in table 3- tables 6 respectively：
The long calibration model prediction result of the brown cotton upper half of table 3
The brown cotton mic value calibration model prediction result of table 4
The brown cotton elongation calibration model prediction result of table 5
6 brown cotton regularity calibration model prediction results
It can be seen that by the data of table 3- tables 6：Color cotton mic value, intensity, elongation and upper half grow four kinds of interior qualities
Index corrects quantitative model predicted value with it with compared with large capacity Wu Site fibrometer system measured values with near-infrared, these
The prediction error of index illustrates to utilize established near-infrared in the range of the admissible analytical error of these index of quality
Quantitative model is corrected, it is feasible to carry out quick nondestructive quantitative analysis to color four kinds of interior qualities of cotton.
Conclusion
Nothing is carried out to color cotton mic value, intensity, elongation and the long interior quality of upper half using near-infrared spectrum technique
Damage rapid detection method with the fibrometer system detection method applied at present compared with have it is easy to operate, it is at low cost (staff training with
Instrument and environment and energy consumption etc.), speed is fast, efficient, not harsh to experimental situation requirement, is particularly suitable for the color cotton sample of high-volume
The Rapid identification of product, must be permanent unlike fibrometer system to experimental situation requirement because near infrared spectroscopy instrument is small
Near infrared spectroscopy instrument is taken in actual cotton purchase and manufacture field, can solved in these fields by the condition of constant temperature and humidity
The superiority that can not be realized using fibrometer system device at present.Color cotton fiber Quality Detection classification aspect has been well solved to exist
The technical barrier that purchase and manufacture field are run into.At present because the requirement of fibrometer system device experimental situation is high, in purchase link
Partly long, the indexs such as elongation still pull tape measurement with hand, and sample measures testing staff one by one, during this also
Estimate and test, regularity, other indexs such as mic value.This method need to pass through certain professional skill to the more demanding of operating personnel
The personnel of training can implement this work.Hand pulls the usual checkout procedure of tape measurement and takes longer, and labor intensive is larger, can not
Meet the requirement that bag is examined.
The method of the invention solves the key that color cotton runs into terms of purchasing with quality identification in processing link at present
Technical barrier is worth promoting and applying in color cotton quality identification.
Claims (1)
1. a kind of lossless rapid detection method of Xinjiang coloured silk cotton interior quality near infrared spectrum, it is characterised in that follow these steps into
Row：
A, the measure of sample parameters value：The brown coloured silk cotton that impurity removing is cleaned out, is numbered, right respectively with fibrometer system
The interior quality that mic value, intensity, elongation and the upper half of each brown coloured silk cotton sample product are long is analyzed, and obtains brown
The detection data of color cotton interior quality, i.e. quality determination value；
B, sample sets are classified：The sample of the brown coloured silk cotton interior quality measured value obtained in step a is divided into Calibration, is tested
Sample sets and Prediction are demonstrate,proved, Calibration and verification sample collection are used for establishing quantitative calibration models, and Prediction is used
Evaluate the accuracy of the model and validity, wherein Calibration accounts for the 60% of sample total amount, verification sample collection is 20%, in advance
Sample collection is 20%；
C, the foundation of sample near infrared spectrum scanning and sample sets：The brown obtained respectively to step b near infrared spectroscopy instrument
Calibration, verification sample collection and the Prediction of color cotton carry out near infrared spectrum scanning, respectively obtain brown coloured silk cotton sample
Product Calibration, verification sample collection and Prediction near infrared spectrum scanning figure；
D, brown coloured silk cotton each sample collection is edited：By the sample sets obtained in step c since Calibration, by all correction samples
The collection of illustrative plates of product collection infrared diaphanoscopy, according to Calibration serial number order, imported into batches near infrared spectrometer " model and
Sample sets management " section, establishes brown coloured silk cotton Calibration in this section, and to the Calibration into edlin：It opens
Calibration, added on Calibration section the mic value of brown coloured silk cotton sample product in step a, intensity, elongation and
The title and its corresponding quality determination value of the long interior quality of upper half, the correcting sample editted are concentrated, and each sample includes
Atlas of near infrared spectra and corresponding four kinds of quality determination values；
It establishes brown coloured silk cotton verification sample collection and Prediction is carried out as the step of Calibration, all samples collection is built
After standing well, in " sample sets management " section, it is seen that established brown coloured silk cotton Calibration, verification sample collection and pre- sample
Collection；
E, quantitative calibration models are established：In near infrared spectrometer " model management " section, by established correcting sample in step d
Collection and verification sample collection are imported into model foundation software, with Partial Least Squares Regression quantitative approach, select 1300nm-
1700nm is modeling wave band with two bands of 1800nm-2400nm, in the modeling parameters and Pretreated spectra of system default
Under the conditions of, quantitative calibration models are established, modeling software is quantitative using being calculated automatically under the modeling configuration condition for importing and selecting
Calibration model, and the prediction result figure of quantitative calibration model and corresponding model parameter, obtain in " model management " section
The quantitative calibration models of brown coloured silk four kinds of qualities of cotton；
F, quantitative calibration models are optimized：Using modeling software editting function, the four kinds of qualities of brown coloured silk cotton obtained in step e are determined
Amount calibration model is modified and is optimized, and model optimization process is：Only change modeling wave band every time, other modeling configurations and spectrum
Pretreatment mode remains unchanged, and establishes the quantitative calibration models of different-waveband, by comparing each modeling wave band quantitative correction mould
The evaluating of type：Calibration set standard deviation, cross-verification collection standard deviation and forecast set standard deviation and reference quantitative correction mould
Type calibration set predicted value and measured value related coefficient, select calibration set standard deviation, cross-verification collection standard deviation and forecast set standard
Difference is minimum, the modeling wave band 1300nm-1700nm of quantitative calibration models calibration set predicted value and measured value related coefficient maximum
With 1800nm-2400nm wave band is modeled for four kinds of qualities of brown coloured silk cotton；Then select this two modeling wave bands constant, only change light
Pretreatment mode is composed, the quantitative calibration models under different Pretreated spectra modes are established, by comparing each Pretreated spectra side
The evaluating of quantitative calibration models under formula, it is similary to select calibration set standard deviation, cross-verification collection standard deviation and forecast set standard
Poor minimum, quantitative calibration models calibration set predicted value and the Pretreated spectra mode of measured value related coefficient related coefficient maximum are
The optimal spectrum pretreatment mode of four kinds of qualities of brown coloured silk cotton finally selects 1300nm-1700nm and 1800nm-2400nm to model
Wave band and Savitzky-Golay are smooth, and Savitzky-Golay derivative spectrum pretreatment modes establish quantitative calibration models, should
Model is four kinds of quantitative calibration models best in quality of brown coloured silk cotton；
G, brown coloured silk cotton Prediction 0.1g-1.5g in step b is placed near infrared spectrometer sample disc preheated in advance
In, through near infrared spectrum scanning, the optimal quantitative calibration models obtained in steps for importing f are answered to get to brown coloured silk cotton sample condition
The prediction result of interior quality.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711267876.2A CN108061724A (en) | 2017-12-05 | 2017-12-05 | The lossless rapid detection method of Xinjiang coloured silk cotton interior quality near infrared spectrum |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711267876.2A CN108061724A (en) | 2017-12-05 | 2017-12-05 | The lossless rapid detection method of Xinjiang coloured silk cotton interior quality near infrared spectrum |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108061724A true CN108061724A (en) | 2018-05-22 |
Family
ID=62136205
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711267876.2A Pending CN108061724A (en) | 2017-12-05 | 2017-12-05 | The lossless rapid detection method of Xinjiang coloured silk cotton interior quality near infrared spectrum |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108061724A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112098357A (en) * | 2020-08-21 | 2020-12-18 | 南京农业大学 | Strawberry sensory quality grade evaluation method based on near infrared spectrum |
| CN112649391A (en) * | 2020-12-03 | 2021-04-13 | 浙江大学 | Fiber quality grade online detection system and application thereof |
-
2017
- 2017-12-05 CN CN201711267876.2A patent/CN108061724A/en active Pending
Non-Patent Citations (4)
| Title |
|---|
| JOSEPH G. MONTALVO等: "Studies to measure cotton fibre length,strength, micronaire and colour by vis/NIRreflectance spectroscopy.Part II: Principal components regression", 《NEAR INFRARED SPECTROSC》 * |
| YONGLIANG LIU等: "Application of near infrared spectroscopy in cotton", 《INFORMATION PROCESSING IN AGRICULTURE》 * |
| 孙武勇等: "近红外光谱技术在棉纤维品质检测和产地溯源中的研究进展", 《中国农学通报》 * |
| 江苏: "近红外技术在烟用辅料质量控制中的应用研究", 《中国优秀硕士学位论文全文数据库 工程科技I辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112098357A (en) * | 2020-08-21 | 2020-12-18 | 南京农业大学 | Strawberry sensory quality grade evaluation method based on near infrared spectrum |
| CN112098357B (en) * | 2020-08-21 | 2021-12-10 | 南京农业大学 | Strawberry sensory quality grade evaluation method based on near infrared spectrum |
| CN112649391A (en) * | 2020-12-03 | 2021-04-13 | 浙江大学 | Fiber quality grade online detection system and application thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100590417C (en) | Plant blade physiologic index non-destructive test method | |
| CN101308086B (en) | Fruit internal quality on-line checking apparatus based on near infrared spectra technology | |
| CN101769866B (en) | Device for detecting milk components and method thereof | |
| CN102879353B (en) | The method of content of protein components near infrared detection peanut | |
| CN103645155B (en) | The fast non-destructive detection method of fresh mutton tenderness | |
| CN101413885A (en) | Near-infrared spectrum method for rapidly quantifying honey quality | |
| JPH03504769A (en) | Method and apparatus for spectroscopic comparison of compositions | |
| CN104390927B (en) | The quick determination method of ash content in coal sample | |
| CN107091815A (en) | A kind of method for identifying rosewood | |
| CN105548062B (en) | The quick nondestructive synchronization detecting method of fresh beef appetizer multi objective | |
| CN109211829A (en) | A method of moisture content in the near infrared spectroscopy measurement rice based on SiPLS | |
| CN106950192A (en) | A kind of method of Contents of Main Components quick detection in vegetable protein beverage based on near-infrared spectral analysis technology | |
| CN107860743A (en) | Utilize the method and its application of the model of reflective near infrared fibre-optical probe structure fast prediction oil property | |
| CN108061724A (en) | The lossless rapid detection method of Xinjiang coloured silk cotton interior quality near infrared spectrum | |
| CN108827907A (en) | It is a kind of based near infrared spectrum to the rapid assay methods of color cotton coloration | |
| CN104849234A (en) | Assay method for analyzing contents of principal components of imidacloprid based on near-infrared spectrum | |
| CN108169168A (en) | Test and analyze rice grain protein content mathematical model and construction method and application | |
| CN109100315B (en) | Wavelength selection method based on noise-signal ratio | |
| CN110231306A (en) | A kind of method of lossless, the quick odd sub- seed protein content of measurement | |
| CN110231302A (en) | A kind of method of the odd sub- seed crude fat content of quick measurement | |
| CN102519903B (en) | Method for measuring whiteness value of Agaricus bisporus by using near infrared spectrum | |
| CN106198423B (en) | A method of ham sausage grade is identified based on visible-near-infrared spectrum analytical technology | |
| CN110057756A (en) | A kind of infrared photoacoustic spectra detection method of wheat flour quality | |
| CN106053380A (en) | Method for quickly analyzing proportions of tree species of mixed pulp wood by aid of near infrared spectrum technologies | |
| CN101387497A (en) | Method for measuring plant fiber material fibre morphology by near-infrared spectrum technology |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination |