CN104374752B - Rapid detection method for nutrient elements of crops based on collinear laser-induced breakdown spectroscopy - Google Patents
Rapid detection method for nutrient elements of crops based on collinear laser-induced breakdown spectroscopy Download PDFInfo
- Publication number
- CN104374752B CN104374752B CN201410653810.7A CN201410653810A CN104374752B CN 104374752 B CN104374752 B CN 104374752B CN 201410653810 A CN201410653810 A CN 201410653810A CN 104374752 B CN104374752 B CN 104374752B
- Authority
- CN
- China
- Prior art keywords
- sample
- crop
- laser
- spectral line
- libs
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/71—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light thermally excited
- G01N21/718—Laser microanalysis, i.e. with formation of sample plasma
Abstract
The invention discloses a rapid detection method for the nutrient elements of crops based on collinear laser-induced breakdown spectroscopy. The rapid detection method comprises the following steps: preparing the fresh leaves of a crop into a laminated sample, raising the optical paths of collinear double pulse laser and then focusing the collinear double pulse laser directly above a sample table for striking the surface of the sample, thereby generating plasmas; acquiring characteristic spectral lines generated by plasma cooling; selecting four spectral lines without self-absorption and self-reversal and highest in intensity out of the characteristic spectral lines according to nutrient elements to be detected; establishing a plurality of multiple regression models with sample reference values as outputs Y' and the intensities of the four spectral lines as inputs, selecting four optimal models from all the models, and establishing a comprehensive model with the prediction results of the four models as inputs and the sample reference values as outputs Y'; for the sample of a to-be-detected crop, obtaining the intensities of four spectral lines of the sample to be detected and inputting the intensities into the comprehensive model, and then calculating the content of the nutrient elements. The rapid detection method for the nutrient elements of crops based on collinear laser-induced breakdown spectroscopy realizes rapid detection on the nutrient elements of crops, and has the characteristics of simplicity in operations, low cost and the like.
Description
Technical field
The present invention relates to crop alimentary Element detection technology, particularly relate to a kind of based on conllinear induced with laser
The crop alimentary element method for quick of breakdown spectral.
Background technology
Crop alimentary element (nitrogen, phosphorus, potassium etc.) is closely related with the upgrowth situation of crop and yield.
Nitrogen, phosphorus, potassium are the big main nutrient elements of crop three.Nitrogen is the constituent of pigment, can promote plant
Photosynthesis, increases crop yield；Phosphorus is the component of crop multiple compounds, to crop quality
There is material impact；Potassium can promote crop photosynthesis, strengthens crop anti-adversity.Obtain crop alimentary
Element information is conducive to understanding the upgrowth situation of crop, realizes the most quantitatively according to the nutritional deficiency situation of crop
Fertilising.At present, the detection method of crop alimentary element mainly have triumphant formula nitriding, Du Mashi combustion method,
Colorimetry, flame combustion process etc..But, the operation of these methods is complicated, cost is high, it usually needs disappear
Consume a large amount of chemical reagent and cause environmental pollution, it is impossible to meeting field-crop nutritional information quick obtaining
Requirement.
Laser induced breakdown spectrograph (Laser-Induced Breakdown Spectroscopy, LIBS)
Utilize plasma ablation that pulse laser produces and excite the material in sample (normally solid),
And the spectrum exciting Plasma-Atomic to be launched is obtained by spectrogrph, identify in sample with this
Elementary composition composition, and then the identification of material, classification, qualitative and quantitative analysis can be carried out.Should
Technology without carrying out pretreatment (or simple pretreatment) to sample, it is possible to quickly realizes damage long-range, micro-
Sample element (especially metallic element) is detected, can be used for solid, gas, fluid sample
Detection.In recent years, LIBS quickly grows, and is widely used in multiple field,
Such as heavy metal analysis and explosive detection etc. in biomedicine, archaeology, environmental monitoring, water.
Such as, the patent documentation of Publication No. CN103257126A discloses a kind of based on induced with laser
The Fast Determination of Pesticide Residue method of breakdown spectral technology, the LIBS spectrum of collecting sample, use smooth
Sequence nonopiate amplitude-frequency method carries out pretreatment to LIBS spectrum；Utilize subwindow to reset to analyze and competition
Property self adaptation weight weighting algorithm extract LIBS elemental characteristic spectral line and the group feature of pesticide residues respectively
Spectral line, and be separately input in pesticide residues species detection model and pesticide residues content detection model,
First carry out pesticide residues species detection, then carry out pesticide residues content detection.
Application LIBS energy quick obtaining crop alimentary information, is advantageously implemented agriculture
Industry precision, intelligentize and informatization.But, owing to LIBS amount of analysis is the least,
Matrix effect has significantly impact to analysis result.Therefore crop sample is typically required and make
Sample tabletting ensures the repeatability of detection.
Summary of the invention
The invention discloses a kind of crop alimentary element rapid detection system and method, it is achieved that based on swashing
Crop tabletting nutrient is quickly detected by photoinduction breakdown spectral, can meet Different Crop laser simultaneously
The features such as the detection of spectrum, has system fading margin convenient, low cost, and detection is quick, micro-damage.
The concrete technical scheme of the present invention is as follows:
A kind of crop alimentary element method for quick based on conllinear LIBS, including
Step:
1) fresh crop leaf is carried out pretreatment, and be fabricated to sample tabletting；
2) after coaxial double-pulse laser light path being raised, by focusing on impact sample table directly over sample stage
Face, produces plasma；
3), when laser energy tends towards stability, gather the characteristic spectral line that plasma cooling produces, and lead to
Cross sample stage and change laser impact sample position, obtain the characteristic spectral line of sample diverse location；
4) change multiple samples of Different Crop, repeat step 1)～step 3), obtain different sample
Characteristic spectral line；
5) according to nutrient to be measured, choose from described characteristic spectral line do not exist self-absorption and from
Reversion and the highest four spectral lines of intensity, the intensity of selected spectral line is respectively I1、I2、I3、I4；
6) using sample reference value as output Y ', with described the intensity of spectral line I1、I2、I3、I4As
Input, sets up multiple multivariate regression models, chooses wherein four optimal modelses, and with these four models
The Y that predicts the outcome1、Y2、Y3、Y4As input, set up using sample reference value Y ' as output and combine
Matched moulds type；Aggregative model equation is: Y=mY1+nY2+jY3+kY4+ l, m, n, j, k are respectively should
Y in aggregative model1、Y2、Y3、Y4Corresponding regression coefficient, l is the constant of this aggregative model；
7) for crop sample to be detected, step 1 is repeated)～step 5), obtain four the intensity of spectral line
I1、I2、I3、I4Aggregative model described in input, calculates nutritive element content.
In step 1) in, fresh crop leaf is positioned in vacuum drying oven and is dried until constant weight, so
Rear taking-up blade carries out freeze grinding, makes grinding rear blade diameier be less than 50 μm, then is put in pressure
Sheet machine carries out tabletting, is fabricated to described sample tabletting.
Fresh crop leaf usually contains higher moisture content, as rape leaf moisture content can reach
70%～80%.When laser impacts fresh crop leaf, the moisture in crop leaf can absorb laser energy
Amount, and produce the interference signal of water.It is thus typically necessary to fresh crop leaf is placed vacuum drying oven
In be dried and remove the interference of moisture in fresh crop leaf.For ensureing the homogeneity of sample to be detected, right
Dried blade carries out freeze grinding.Dried blade powder should be fully ground, it is ensured that powder is straight
Footpath is less than 50 μm.Abrasive flour granule is crossed senior general and is impacted the testing result of nutrient, makes
Detection repeatability declines.Additionally, the powder after grinding is carried out tabletting be avoided that powder is by laser arteries and veins
Punching effect produces to splash and destroys sample surface planarization, is provided that a uniform schedule of samples of opposed flattened
Face, it is ensured that the accuracy of detection and repeatability.
The described light path system of climbing includes the first reflecting mirror, the second reflecting mirror being sequentially arranged along light path
With the 3rd reflecting mirror, described second reflecting mirror is positioned at the surface of the first reflecting mirror, the 3rd reflecting mirror position
Surface in sample stage.
The light path system of climbing is mainly used in lifting light path, and will be converted into along vertical by laser in the horizontal direction
Direction is propagated.The application light path system of climbing can be prevented effectively from rising laser positions and cause laser unstable
Factor.Laser impacts effectively exciting of sample surfaces, beneficially plasma directly over sample,
Ensure that plasma is uniformly and symmetrically distributed.Wherein, described the first reflecting mirror, the second reflecting mirror and
Three reflecting mirrors are Nd:YAG double-frequency laser reflecting mirror.
Preferably, described sample stage includes having the displacement platform of Three Degree Of Freedom, be movably arranged on displacement
Lifter plate on platform and the object stage being slidably fitted on displacement platform, described sample tabletting is placed on loading
On platform；Described lifter plate is provided with transparent constraint window, impacts sample after laser light constraint window；
The lower section of described lifter plate is provided with restraining plate, and this restraining plate is placed in the surface of sample tabletting, restraining plate
On constraint chamber is distributed, this constraint chamber is for retraining the plasma of sample excitation.
Sample stage limits enhanced line intensity by space, can laterally and longitudinally carry out by plasma
Constraint, and according to different sample requirement regulation constraint space size adjustment the intensity of spectral line, the intensity of spectral line
Enhancing scope is 2-10 times.Constraint window is mainly used in plasma and longitudinally retrains, and right
The characteristic spectral line that incident laser produces with plasma has preferable absorbance.Restraining plate is mainly used in
Laterally retraining of plasma, and make characteristic spectral line carry out constraint propagation, improve spectral line and collect
Efficiency and the intensity of spectral line.
Wherein, described lifter plate is provided with light-transmitting opening, is coated with light-passing board at this light-transmitting opening, described
Light-transmitting opening and the constraint window described in light-passing board composition.Light-passing board is poly (methyl methacrylate) plate, and the selection of material is
N-BK7, thickness is 1～5mm, for ensureing that laser effectively excites the effective collection with characteristic spectral line,
Light-passing board should be greater than 90% to the transmitance of optical maser wavelength and characteristic spectral line, simultaneously for preventing laser to having
Machine glass plate causes damage, and its energy threshold should be greater than 10J/cm2.Line enhancement effect suffers restraints
Window is from the impact of sample distance.Owing to properties of samples is different with element spectral line intensity to be detected,
The constraint window of the present invention can move in vertical direction, regulates the intensity of spectral line as required.
Restraining plate is the aluminium sheet of chromium plating, and thickness is 1～3mm.Constraint intracavity at aluminium sheet carries out chromium plating,
Make inner chamber have higher reflectance, make characteristic spectral line constraint propagation, improve spectral line collection efficiency.About
Bundle chamber uses conical structure, due to the structure that it is up-small and down-big, and can be more preferably compared with the structure of column type
Obtain plasma to retrain.It addition, when plasma is from cone constraint chamber injection, due to
Its space constraint strengthens, and its electron density and movement velocity all can be strengthened, the most advantageously in
The enhancing of spectral line signal.
In step 6) in, described multivariate regression models includes that partial least square method, multiple linear return
Return, principal component regression, gradually linear regression, ridge regression, Logistic return, least square-support
Vector machine, artificial neural network, extreme learning machine, Gaussian process return.
The present invention uses multiple multivariate regression models to set up the forecast model of crop alimentary element.Described
Multivariate regression models had both included that linear regression model (LRM) also included nonlinear regression model (NLRM).Each different returning
Model is returned all to have different pluses and minuses and applicable object.As main composition return, multiple linear returns, partially
The linear regression methods such as little square law be more suitable for spectral value and the crop alimentary element linearity preferable time.
When there is each component and interacting or there is noise of instrument or baseline drift, these linear correction methods
Just preferable model can not be obtained.At this moment, artificial neural network, least squares support vector machines etc.
Non-linear regression method is more suitable for the foundation of optimum prediction model.Owing to crop is that a kind of matrix is complicated
Sample, there is bigger difference in its character of different crops.Additionally, in agricultural production, nutrition unit
The detection of element is easily subject to the impact of environmental factors.Use multiple regression model described above the most preferably
Four kinds of optimal modelses can be effectively ensured Different Crop sample and Different Nutrient Elements optimum prediction effect.
Meanwhile, in described multivariate regression models, choose coefficient of determination R2Big and root-mean-square error
Four little for RMSE optimal modelses.
Coefficient of determination R2It it is all the evaluation index of evaluation model effect with root-mean-square error RMSE.Certainly
Determine coefficient and be mainly used to characterize the effect of data matching statistical model.When the coefficient of determination is closer to 1,
Then modelling effect is the best；Root-mean-square error is used for the deviation weighing predictive value with actual value, with accuracy
Effect similar, when root-mean-square error is the least, be closer to 0, show that the predictive ability of model is the strongest.
In this method, it is ranked up according to coefficient of determination size, and prioritizing selection root-mean-square error is less
Regression model, selects four models of optimum from above-mentioned set up multiple multivariate regression models.?
Step 3) in, use baseline translation and spectral normalization that described characteristic spectral line is carried out pretreatment,
Remove background noise by baseline translation, use spectral normalization to come correcting energy fluctuation and matrix school
Just.
The invention have the advantages that:
(1) parameter of energy quickly conditioning instrumentation, it is achieved the optimization of optimum detection parameter.
(2) achieve the quick detection of crop alimentary element, there is the spies such as simple to operate, low cost
Point.Effectively overcome the feature that traditional detection method detection time length, operation are complicated, it is to avoid chemistry
The reagent pollution to environment.
(3) crop sample is through dry, grinding, tabletting, can be effectively improved the repeatability of detection,
Avoid that sample is uneven brings detection error.
(4) sample stage limits enhanced line intensity by space, can laterally and indulge by plasma
To retraining, improve spectral line collection efficiency and the intensity of spectral line；It is avoided that and excites granule again to cover
Sample surfaces.
Accompanying drawing explanation
Fig. 1 is the system construction drawing that crop alimentary element quickly detects；
Fig. 2 is the method flow diagram that crop alimentary element quickly detects；
Fig. 3 is the structure chart of sample stage；
Fig. 4 is the top view of sample stage in Fig. 3.
Detailed description of the invention
As it is shown in figure 1, the device that in the present invention, crop alimentary element method for quick is suitable for includes
Double-impulse solid laser 1, delay time generator 2, energy attenuator 3, laser energy monitors system in real time
System 4, light path climbs system 5, detector 6, spectrogrph 7, sample stage 8, condenser lens 9, light
Fine collection system 10 and sample surface imaging system 11.
Delay time generator 2 controls Q-switch triggered time of dipulse solid generators 1 and xenon lamp leaves
Open the triggered time.Under delay time generator 2 controls, Double-impulse solid laser 1 produces coaxial 532nm
(1064nm) double-pulse laser.Laser energy is decayed by laser through energy attenuator 3.Logical
Cross and rotate half-wave plate 31 and control from beam splitter 32 energy out.After decay, laser is respectively through first
Reflecting mirror 52 and the second reflecting mirror 51, then change light path through the 3rd reflecting mirror 53, travel downward.
Laser focuses on impact sample surfaces through condenser lens 9, produces plasma.Plasma cooling is produced
Raw characteristic spectral line is collected through optical fiber collection system 10, through spectrogrph 7 light splitting, finally by detecting
Photomultiplier tube on device 6 is converted into the signal of telecommunication, is connected on computer 12 by USB, passes through
The model database established, calculates crop alimentary constituent content.
Light path climbs system 5 by the first reflecting mirror the 52, second reflecting mirror 51 and the 3rd reflecting mirror 53
Composition, preferably Nd:YAG double-frequency laser reflecting mirror, the reflectance at its 532/1064nm is high
In 98%.
Optical fiber collection system 10 is made up of light collector and optical fiber.Sample surface imaging system 11 includes
CCD camera 114, imaging lens the 113, the 3rd beam splitter 112, LED light source for illuminating 111.
Laser energy real-time monitoring system includes thermoelectric pulse probe the 42, second beam splitter 41, USB
Connecting line, computer 12.Energy attenuator 3 is made up of half-wave plate 31 and the first beam splitter 32.Visit
Surveying device 6 is ICCD detector, and spectrogrph 7 is echelle spectrometer.
Delay time generator 2 controls Double-impulse solid laser 1 and adjusts Q triggered time, xenon lamp to open triggering
Time and the gate duration of ICCD detector.Delay time generator arranges Double-impulse solid laser and adjusts Q
The triggered time of signal is 150 ± 20 μ s after xenon lamp signal triggers.The control opening time of detector
It is 1.5 μ s after the second road laser Q-switching switch triggering time.
The repetition rate of Double-impulse solid laser 1 is 1-10Hz, and energy is 80-200mJ
(@532nm), 80-300mJ (@1064nm), the interval time between two bundle laser is 500ns.
Double-impulse solid laser 1 has 4 kinds of working methods: 1) 1064nm coaxially exports, and 2)
532nm coaxially exports, and 3) 1064 different axle outputs, 4) output of 1064nm, 532nm different axle.Double
Pulsed solid stale laser 1 can be by 2 pulse 1064nm solid state lasers, frequency-doubling crystal and conjunction bundle
Light path realizes.
Condenser lens 9 is installed on to be had in the Lens mounting carrier of regulation degree of freedom along optical axis direction, is used for
Regulation sample and the distance of lens, thus control the laser parameter impacted on sample.Condenser lens selects
With N-BK7, plating V-arrangement film, its reflectance in 532,1064nm place is less than 0.25%.
As shown in Figure 3 and Figure 4, sample stage 8 includes gear rack elevating bar 901, gear slider 902,
Right angle keyset 903, plasma confinement window 904, cylindrical guide 905, plasma confinement plate
906, V-type slide block 907, combines electric displacement platform 909, pull bar 910, retrains chamber 911.Combination electromigration
Platform 909 uses the displacement platform of Three Degree Of Freedom.Gear rack elevating bar 901 is vertically installed at combination electric displacement platform 909
On, gear slider 902 engages with gear rack elevating bar 901, can be lower mobile in the vertical direction.Right angle
Keyset 903 is secured by bolts on gear slider 902, and plane is provided with the light-transmitting opening of rectangle, thoroughly
Guang Kou edge is provided with support level, is provided with light-passing board in light-transmitting opening, forms plasma confinement window
904.Cylindrical guide 905 is fixed on combination electric displacement platform 909, and plasma confinement plate 906 can edge
Cylindrical guide 905 slides up and down.Cylindrical guide 905 and gear rack elevating bar 901 all pass through bolt-connection
On combination electric displacement platform 909.Combination electric displacement platform 909 has V-groove, and V-type slide block 907 is at pull bar
Can horizontally slip along V-groove under 910 effects, pull bar 910 is threadably secured in V-type slide block 907
In.Sample 908 is placed on V-type slide block 907 (being equivalent to object stage), and laser hits down from above
Beat, by plasma confinement window 904, and through plasma confinement plate 906, impact to sample
Product surface, excites plasma, cooling to send characteristic spectral line and collected by optical fiber collection system 10.
The material of plasma confinement window 904 is N-BK7, and thickness is 1-5mm, and transmitance is big
In 90%, energy threshold is more than 10J/cm2.Plasma confinement window 904 be mainly used in equity from
Daughter longitudinally retrains, and the characteristic spectral line producing incident laser and plasma has preferably
Absorbance.N-BK7 is a kind of common optical glass, it is possible to through 350nm-2000nm wave band
Light, the absorbance of its laser is more than 90%, and energy threshold is more than 10J/cm2.Plasma confinement window
Mouth 904 vertically can move under gear slider 902 acts on.Line enhancement effect is subject to
Constraint window is from the impact of sample distance.Due to properties of samples and element spectral line intensity to be detected not
With, the plasma confinement window 904 of the present invention can move in vertical direction, adjusts as required
Joint the intensity of spectral line.
In the present embodiment, the material of plasma confinement plate 906 is the aluminium sheet of chromium plating, and thickness is
1mm, is evenly distributed with cone constraint chamber 911 in the middle of it, upper conical surface diameter and lower conical surface diameter are respectively
It is 2mm and 3mm.The spacing in constraint chamber 911 should plan displacement phase one with combination electric displacement platform 909
Cause.Plasma confinement plate 906 is mainly used in laterally retraining of plasma, and makes feature
Spectral line carries out constraint propagation, improves spectral line collection efficiency and the intensity of spectral line, and the intensity of spectral line strengthens scope and is
2-10 times.In the constraint chamber 911 of aluminium sheet, carry out chromium plating, make inner chamber have higher reflectance, make
Characteristic spectral line constraint propagation, improves spectral line collection efficiency.Constraint chamber uses conical structure, due to it
Up-small and down-big structure, more preferable can obtain plasma compared with the structure of column type and retrain.It addition,
When plasma from cone constraint chamber injection time, owing to its space constraint strengthens, its electron density with
Movement velocity all can be strengthened, the most advantageously in the enhancing of spectral line signal.Plasma confinement
With combination electric displacement platform 909, the spacing in the constraint chamber of plate 906 plans that displacement is consistent, can be suitably used for swashing
The working method of photoinduction breakdown spectral Surface scan.Using plasma restraining plate 906 of the present invention covers
Sample surfaces, the most smooth sample surfaces improves repeatability, it is to avoid other is to be measured to excite particle contamination
Region.When testing sample is the surface irregularity samples such as fresh blade, its region to be measured and lens away from
From there are differences, and then affect the laser parameter of laser arrival sample.The inspection of LIBS
Surveying stability closely bound up with the laser parameter in region to be detected, the most smooth sample surfaces is conducive to solid
Determine laser parameter and improve the repeatability of detection.Additionally, the up-small and down-big circle of plasma confinement plate 906
Pyramidal structure is conducive to farthest avoiding being excited particle contamination by a upper detection region, it is ensured that institute
Detection object is the element in region to be detected.
V-type slide block 907 can move along V-groove under pull bar 910 acts on, and uses this V-type slide block
Guide rail structure, it is to avoid raise the troublesome operation such as plasma confinement window and plasma confinement plate.Plasma
Connection between body restraining plate 906 and cylindrical guide 905 for being close to connect, plasma under gravity
Body restraining plate 906 can not freely glide.
Control combination electric displacement platform 909 makes the pyramid type constraint chamber 911 of plasma confinement plate 906 with upper
The position of Fang Jiguang is corresponding, and the work step-length arranging combination electric displacement platform 909 is cone constraint chamber phase
Neighborhood distance or multiple.Coaxial double-pulse laser passes down from above through light path system, line focus lens
Broadcast, through plasma confinement window 904, and through cone constraint chamber impact sample.Plasma
Its laser transmittance of body constraint window 904 is more than 90%, and energy threshold is more than 10J/cm2.Deng from
Daughter cooling sends characteristic spectral line and is collected by the optical fiber collection system 10 of top.Each position can basis
Actual requirement selects the number of times of required impact, by combination electric displacement platform 909 after completing a position
Move and carry out multiple positions spectra collection.When without plasma confinement window 904 and plasma confinement
During plate 906, the plasma inspired is free excited state；When the plasma excited is carried out
After restriction, the density of plasma increases, and plasma temperature raises, and enhances the spectrum of excited atom
Line strength.
As shown in Figure 2, it is achieved crop alimentary element method for quick comprises the following steps:
(1) fresh crop leaf is carried out pretreatment.Fresh blade is positioned in vacuum drying oven dry
Dry until constant weight.Take out blade and carry out freeze grinding, grind rear blade diameier and should be less than 50 μm；
Weigh 0.25g blade powder in tablet machine, carry out tabletting.Mould for tabletting press is 15mm, uses pressure
Power is 15t, keeps pressure 3min.
(2) according to the different operating parameter selected, systematic parameter is regulated.Changed by energy attenuator
Become and reach sample surface laser energy.The half-wave plate of rotating energy attenuator changes laser polarization direction,
Different polarization direction is different by the energy of beam splitter.Regulate the first condenser lens along optical axis direction position
Move, control lens to sample surface distance, control sample surface laser parameter (regulation Lens mounting carrier
Make lens arrive the distance focal length less than lens of sample surface along optical axis direction displacement, and make spectral line noise
Than reaching maximum).Arrange the delay time generator sequencing contro tune Q triggered time, xenon lamp is opened when triggering
Between and the gate duration of ICCD detector.When the signal to noise ratio gathering spectral line is higher than threshold value, start to make
Thing sample spectra acquisition, otherwise repeats step (2), until meeting requirement.
(3) atomic spectrum of crop sample is gathered.When the laser that energy real-time detecting system is gathered
When energy tends towards stability, start to gather the atomic spectrum of crop sample.Change laser by sample stage to hit
Beat sample position, it is to avoid repeat impact.Each sample tabletting gathers 10 positions, and each position is adopted
Collect 10 spectrum, totally 100 spectrum.100 atomic spectrums are averaged and represent a sample.
(4) atomic spectrum collected is carried out pretreatment, (remove background to make an uproar including baseline translation
Sound), spectral normalization (correcting energy fluctuation and matrix correction) etc..According to atom information data base
Select the spectral line of surveyed element, selected characteristic spectral line should be and there is not self-absorption and self reversal and intensity
Four strong spectral lines, the intensity of selected spectral line is set to I1、I2、I3、I4。
(5) by reference method such as atomic absorption spectrum, the nutrition of the multiple sample of Different Crop is obtained
Element reference value, is set to Y '.Using sample reference value Y ' as output, repeat step (1) to (4)
Obtain different shape sample line strength I1、I2、I3、I4As input, set up respectively partial least square method,
Multiple linear regression, principal component regression, gradually linear regression, ridge regression, Logistic return,
A young waiter in a wineshop or an inn takes advantage of-support vector machine, artificial neural network, extreme learning machine, Gaussian process regression model,
Therefrom choose 4 optimum prediction model Y1、Y2、Y3、Y4.As a example by multiple linear regression model,
Then model equation is: Y4=aI1+bI2+cI3+dI4+ e, a, b, c, d are respectively this multiple linear and return
Return I in equation1、I2、I3、I4Corresponding regression coefficient, e is the constant of this multiple linear regression equations.
Predict the outcome Y with four forecast models1、Y2、Y3、Y4As input, with sample reference value Y ' work
Setting up aggregative model for output, model equation is: Y=mY1+nY2+jY3+kY4+ l, m, n, j, k
It is respectively Y in this aggregative model1、Y2、Y3、Y4Corresponding regression coefficient, l is this aggregative model
Constant.The model database of various crop is set up for the most pre-with the aggregative model of final quadratic fit
Survey crop alimentary constituent content.
(6) obtain crop sample to be detected, repeat 4 spectral lines of step (1) to (4) acquisition strong
Degree I1、I2、I3、I4As input, calculate nutrient with the model database that step (5) is set up
Content.
It is as follows that what said method realized is embodied as case:
With N, P, K nutritive element content in application LIBS detection rape leaf,
Described system and method is illustrated elaboration.
1, the fruit and vegerable standard substance of seven kinds of known N, P, K concentration are used to set up model database here,
Known nutrient reference value Y ' is as shown in the table.
| Numbering | Nitrogen (%) | Phosphorus (%) | Potassium (%) |
| GBW10014 | 2.8 | 0.46 | 1.55 |
| GBW10019 | 0.31 | 0.066 | 0.77 |
| GBW10020 | 2.47 | 0.125 | 0.77 |
| GBW10021 | 2.79 | 0.38 | 2.26 |
| GBW10022 | 3.22 | 0.466 | 1.14 |
| GBW10023 | 5.0 | 0.585 | 3.36 |
| GBW10025 | 10.6 | 1.17 | 1.41 |
2, weigh 0.25g sample in tablet machine, carry out tabletting.Mould for tabletting press is 15mm, uses
Pressure is 15t, keeps pressure 3min.
3, according to the different operating parameter selected, systematic parameter is regulated.Changed by energy attenuator
Reach sample surface laser energy.The half-wave plate of rotating energy attenuator changes laser polarization direction, no
Different by the energy of beam splitter with polarization direction.Regulate the first condenser lens along optical axis direction displacement,
Control lens, to sample surface distance, control sample surface laser parameter, make spectral line signal to noise ratio reach
Greatly.Arranging delay time generator sequencing contro adjusts Q triggered time, xenon lamp to open triggered time and ICCD
The gate duration of detector.When the signal to noise ratio gathering spectral line is higher than threshold value, start crop sample spectrum
Gathering, otherwise repeating step 3, until meeting requirement.
4, the atomic spectrum of crop sample is gathered.When the laser energy that energy real-time detecting system is gathered
When amount tends towards stability, start to gather the atomic spectrum of crop sample.Change laser by sample stage to impact
Sample position, it is to avoid repeat impact.Each sample tabletting gathers 10 positions, each station acquisition
Article 10, spectrum, totally 100 spectrum.100 atomic spectrums are averaged and represent a sample.
5, the atomic spectrum collected is carried out pretreatment, including baseline translation (removal background noise),
Spectral normalization (correcting energy fluctuation and matrix correction) etc..Institute is selected according to atom information data base
Surveying the spectral line of element, selected characteristic spectral line should be and there are not self-absorption and the strongest four of self reversal intensity
Spectral line, the intensity of selected spectral line is set to I1、I2、I3、I4。
6, using sample reference value Y ' as output, repeat step 2 to 5 and obtain various criterion product sample
The intensity of spectral line I1、I2、I3、I4As input, set up partial least square method respectively, multiple linear returns
Return, principal component regression, gradually linear regression, ridge regression, Logistic return, least square-support
The models such as vector machine, artificial neural network, extreme learning machine, Gaussian process recurrence, therefrom choose 4
Individual optimum prediction model Y1、Y2、Y3、Y4(it is ranked up according to coefficient of determination size, and preferentially
4 that select root-mean-square error less is optimum prediction model Y1、Y2、Y3、Y4).Pre-with four
Survey model prediction result Y1、Y2、Y3、Y4As input, build using sample reference value Y ' as output
Vertical aggregative model, the coefficient of determination R of N, P, K element in this example2It is respectively 0.91,0.95,0.90.
7, obtain rape leaf sample, fresh blade is positioned in vacuum drying oven and is dried until constant weight.
Take out blade and carry out freeze grinding, grind rear blade diameier and should be less than 50 μm；Repeat step 2
4 the intensity of spectral line I are obtained to step 51、I2、I3、I4As input, the mould set up with step 6
Type data base calculates nutritive element content.
For verifying the feasibility of the method, when using nitrogen, phosphorus, potassium content to be respectively 4.85%, 0.46%,
The rape leaf of 2.01% is as sample to be detected, it was predicted that nitrogen, phosphorus, potassium content are respectively 4.72%,
0.53%, 2.23%, the method is close with the nutritive element content measured with standard method.
Claims (10)
1. a crop alimentary element method for quick based on conllinear LIBS,
It is characterized in that, including step:
1) fresh crop leaf is carried out pretreatment, and be fabricated to sample tabletting；
2) after coaxial double-pulse laser light path being raised, by focusing on impact sample table directly over sample stage
Face, produces plasma；
3), when laser energy tends towards stability, gather the characteristic spectral line that plasma cooling produces, and lead to
Cross sample stage and change laser impact sample position, obtain the characteristic spectral line of sample diverse location；
4) change multiple samples of Different Crop, repeat step 1)～step 3), obtain different sample
Characteristic spectral line；
5) according to nutrient to be measured, choose from described characteristic spectral line do not exist self-absorption and from
Reversion and the highest four spectral lines of intensity, the intensity of selected spectral line is respectively I1、I2、I3、I4；
6) using sample reference value as output Y ', with described the intensity of spectral line I1、I2、I3、I4As
Input, sets up multiple multivariate regression models, chooses wherein four optimal modelses, and with these four models
The Y that predicts the outcome1、Y2、Y3、Y4As input, set up using sample reference value Y ' as output and combine
Matched moulds type；Aggregative model equation is: Y=mY1+nY2+jY3+kY4+ l, m, n, j, k are respectively should
Y in aggregative model1、Y2、Y3、Y4Corresponding regression coefficient, l is the constant of this aggregative model；
7) for crop sample to be detected, step 1 is repeated)～step 5), obtain four the intensity of spectral line
I1、I2、I3、I4Aggregative model described in input, calculates nutritive element content.
2. crop alimentary element based on conllinear LIBS as claimed in claim 1
Method for quick, it is characterised in that in step 1) in, fresh crop leaf is positioned over vacuum
Baking oven is dried until constant weight, then takes out blade and carry out freeze grinding, make grinding rear blade powder straight
Footpath is less than 50 μm, then is put in tablet machine and carries out tabletting, is fabricated to described sample tabletting.
3. crop alimentary element based on conllinear LIBS as claimed in claim 1
Method for quick, it is characterised in that use light path to climb system by coaxial double-pulse laser light path liter
Height, the described light path system of climbing includes the first reflecting mirror, the second reflecting mirror being sequentially arranged along light path
With the 3rd reflecting mirror, described second reflecting mirror is positioned at the surface of the first reflecting mirror, the 3rd reflecting mirror position
Surface in sample stage.
4. crop alimentary element based on conllinear LIBS as claimed in claim 3
Method for quick, it is characterised in that described the first reflecting mirror, the second reflecting mirror and the 3rd reflection
Mirror is Nd:YAG double-frequency laser reflecting mirror.
5. crop alimentary element based on conllinear LIBS as claimed in claim 1
Method for quick, it is characterised in that described sample stage include having Three Degree Of Freedom displacement platform,
The lifter plate being movably arranged on displacement platform and the object stage being slidably fitted on displacement platform, described sample
Tabletting is placed on object stage；
Described lifter plate is provided with transparent constraint window, impacts sample after laser light constraint window；
The lower section of described lifter plate is provided with restraining plate, and this restraining plate is placed in the surface of sample tabletting, about
Constraint chamber is distributed on bundle plate, and this constraint chamber is for retraining the plasma of sample excitation.
6. crop alimentary element based on conllinear LIBS as claimed in claim 5
Method for quick, it is characterised in that described lifter plate is provided with light-transmitting opening, covers at this light-transmitting opening
It is stamped light-passing board, described light-transmitting opening and the constraint window described in light-passing board composition.
7. crop alimentary element based on conllinear LIBS as claimed in claim 5
Method for quick, it is characterised in that described restraining plate is the aluminium sheet of chromium plating, constraint chamber is cone.
8. crop alimentary element based on conllinear LIBS as claimed in claim 1
Method for quick, it is characterised in that described multivariate regression models includes partial least square method, many
Unit's linear regression, principal component regression, gradually linear regression, ridge regression, Logistic return, minimum
Two take advantage of-support vector machine, artificial neural network, extreme learning machine, Gaussian process return.
9. crop alimentary element based on conllinear LIBS as claimed in claim 8
Method for quick, it is characterised in that in multiple multivariate regression models, chooses coefficient of determination R2
Big and that root-mean-square error RMSE is little four optimal modelses.
10. crop alimentary element based on conllinear LIBS as claimed in claim 1
Method for quick, it is characterised in that in step 3) in, use baseline translation and spectral normalization
Described characteristic spectral line is carried out pretreatment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410653810.7A CN104374752B (en) | 2014-11-17 | 2014-11-17 | Rapid detection method for nutrient elements of crops based on collinear laser-induced breakdown spectroscopy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410653810.7A CN104374752B (en) | 2014-11-17 | 2014-11-17 | Rapid detection method for nutrient elements of crops based on collinear laser-induced breakdown spectroscopy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104374752A CN104374752A (en) | 2015-02-25 |
| CN104374752B true CN104374752B (en) | 2017-01-11 |
Family
ID=52553799
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410653810.7A Active CN104374752B (en) | 2014-11-17 | 2014-11-17 | Rapid detection method for nutrient elements of crops based on collinear laser-induced breakdown spectroscopy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104374752B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105510305A (en) * | 2015-12-01 | 2016-04-20 | 浙江大学 | Method for detecting content of aluminum in potato starch |
| CN105403532A (en) * | 2015-12-01 | 2016-03-16 | 浙江大学 | Method for quickly identifying walnut powder production places |
| CN105717074A (en) * | 2016-01-22 | 2016-06-29 | 浙江大学 | Method for identifying matcha powder and green tea powder by aid of laser-induced breakdown spectra on basis of characteristic wavelengths |
| CN106501236A (en) * | 2016-09-30 | 2017-03-15 | 浙江大学 | Copper content detection method in rice based on single-pulse laser induced breakdown spectroscopy |
| CN106546575A (en) * | 2016-10-20 | 2017-03-29 | 浙江大学 | Copper content detection method in a kind of rice based on LIBS |
| CN106841174B (en) * | 2017-01-24 | 2019-08-23 | 浙江大学 | Blade heavy metal content detection method based on Mg element ratios correction moisture content |
| CN106770195B (en) * | 2017-01-24 | 2019-11-29 | 浙江大学 | Blade heavy metal content detection method based on CN element ratios correction moisture content |
| CN106841173B (en) * | 2017-01-24 | 2019-09-10 | 浙江大学 | Blade heavy metal content detection method based on K element ratiometric correction moisture content |
| CN107843843A (en) * | 2017-09-30 | 2018-03-27 | 江苏理工学院 | A kind of on-vehicle battery SOC on-line prediction methods based on big data and extreme learning machine |
| CN108267429A (en) * | 2017-11-30 | 2018-07-10 | 南京信息工程大学 | The quantitative analysis method of full iron and basicity in a kind of sinter |
| CN113340874B (en) * | 2020-03-02 | 2023-07-18 | 中国科学院沈阳自动化研究所 | Quantitative analysis method based on combination ridge regression and recursive feature elimination |
| CN111398252A (en) * | 2020-03-17 | 2020-07-10 | 浙江大学 | Method and system for quantitatively detecting cadmium element in rice leaves |
| CN111398254B (en) * | 2020-03-26 | 2021-09-24 | 浙江大学 | Rice leaf heavy metal copper element quantitative detection method and system |
| CN112629670B (en) * | 2020-12-28 | 2021-09-24 | 上海交通大学 | High-frequency dynamic flame temperature measuring method and system |
| CN112947635B (en) * | 2021-01-15 | 2022-04-29 | 江南大学 | Multi-model wheat seedling growth cabin optimal parameter prediction method based on Kalman filter |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4690558A (en) * | 1984-12-18 | 1987-09-01 | Kawasaki Steel Corporation | Method of laser emission spectroscopical analysis and apparatus therefor |
| US5847825A (en) * | 1996-09-25 | 1998-12-08 | Board Of Regents University Of Nebraska Lincoln | Apparatus and method for detection and concentration measurement of trace metals using laser induced breakdown spectroscopy |
| FR2938066B1 (en) * | 2008-11-06 | 2010-12-17 | Centre Nat Rech Scient | SYSTEM AND METHOD FOR QUANTITATIVE ANALYSIS OF THE ELEMENTARY COMPOSITION OF LASER-INDUCED PLASMA SPECTROSCOPY MATERIAL (LIBS) |
| CN102183494B (en) * | 2011-02-18 | 2013-01-16 | 清华大学 | Method for improving element measurement precision by using plasma space restricting effect |
| CN102262075B (en) * | 2011-07-26 | 2013-10-30 | 清华大学 | Method for measuring elemental concentration through laser-induced breakdown spectroscopy based on spectrophotometry |
| CN102262076B (en) * | 2011-07-26 | 2013-05-22 | 清华大学 | Method for laser-induced breakdown spectroscopy element concentration determination based on spectral line combination |
| CN102998974A (en) * | 2012-11-28 | 2013-03-27 | 上海交通大学 | Multi-model generalized predictive control system and performance evaluation method thereof |
| CN103488874B (en) * | 2013-09-01 | 2016-09-14 | 西北大学 | A kind of combination LIBS sorting technique to ferrous materials |
-
2014
- 2014-11-17 CN CN201410653810.7A patent/CN104374752B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN104374752A (en) | 2015-02-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104374752B (en) | Rapid detection method for nutrient elements of crops based on collinear laser-induced breakdown spectroscopy | |
| CN104374753B (en) | Crop heavy metal based on double-pulse laser induced breakdown spectrum and method for detecting microelements | |
| Corsi et al. | A fast and accurate method for the determination of precious alloys caratage by laser induced plasma spectroscopy | |
| CN105008898B (en) | The system and method for LIBS and IR absorption spectrometry researchs for combination | |
| McGlone et al. | Internal quality assessment of mandarin fruit by vis/NIR spectroscopy | |
| US20230062339A1 (en) | Methods for Spectrally Resolving Fluorophores of a Sample and Systems for Same | |
| CN101995395B (en) | Method for online detecting material by laser induction multiple spectrum united fingerprint network | |
| CN104374763B (en) | Adjustable reheating double pulse laser-induced breakdown spectroscopy device | |
| CN104374751B (en) | Device based on conllinear LIBS crop nutrient quick detection | |
| CN104406942B (en) | A kind of multi-mode laser induced breakdown spectroscopy device | |
| CN104483292B (en) | A kind of method that use multiline ratio method improves laser microprobe analysis accuracy | |
| CN102507512A (en) | On-line in situ detecting method for infrared-ultraviolet double pulse laser induced breakdown spectroscopy | |
| CN104778349B (en) | One kind is used for rice table soil nitrogen application Classified Protection | |
| CN204214783U (en) | One is adjustable heats double-pulse laser induced breakdown spectral device again | |
| CN108535200B (en) | Detection device and method for heavy metal cadmium in leaves of leaf vegetables | |
| Yu et al. | Response surface methodology for optimizing LIBS testing parameters: A case to conduct the elemental contents analysis in soil | |
| CN204214779U (en) | Based on the device that conllinear Laser-induced Breakdown Spectroscopy crop alimentary element detects fast | |
| CN104374760B (en) | Atomic fluorescence spectrophotometry crop heavy metal based on laser ablation plume and trace element method for quick | |
| CN111829961A (en) | Food quality detection method and system based on portable imaging spectrometer | |
| Huang et al. | Online identification and classification of Gannan navel oranges with Cu contamination by LIBS with IGA-optimized SVM | |
| CN113504185A (en) | Multispectral agricultural product heavy metal content detection modeling system based on LIBS calibration | |
| CN109696407A (en) | A kind of coco bran matrix available nitrogen spectral method of detection based on characteristic wavelength | |
| CN109444093B (en) | Synchronous imaging detection method and system | |
| TANG et al. | The rapid detection method of chlorophyll content in rapeseed based on hyperspectral technology | |
| CN117147526B (en) | Large-area soil pH value rapid detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |