CN202133599U - Particle size detection device - Google Patents

Abstract

The utility model discloses a particle size detection device which comprises a highly stable multi-wavelength light source, an input optical fiber, a first high-precision electronically controlled optical adjustment table, a CCD (charge coupled device) imaging system, a second high-precision electronically controlled optical adjustment table, an output optical fiber, a high-precision detector, and a computer. The utility model utilizes the computer to control the high-precision optical adjustment tables, the space between the end face of the input optical fiber connected with the light source and the end face of the output optical fiber connected with a power-factor indicator is controlled and locked within 100um, to-be-detected solution is dripped to form a sample cell structure, then multi-wavelength scanning is used to improve conventional MIE theory through combining with a Lorentz model, and a multivariate genetic algorithm is proposed to realize real-time on-line detection. The device provided by the utility model has small volume, is convenient to move, dispenses with advanced refracting power measurement to the to-be-detected solution and external independent high-power laser light source, eliminates reflection and scattering of the sample cell, and can realize on-line particle detection.

Description

A kind of grain graininess pick-up unit

Technical field

The utility model relates to the particle in emulsion and the suspension, particularly a kind of pick-up unit of the grain graininess based on optical fiber coupling.

Background technology

In recent years, grain graininess detects in the application of agricultural, chemical industry, biomedical aspect more and more widely.Comprise sieve method, sedimentation, microscopic method, electro-induction method, ultrasonic method, light scattering method etc.Wherein sieve method is surveyed through colorimetric or the method for utilizing the structure of special pore size distribution to screen the relative granularity particle.Full light scattering method is a kind of absolute method of measurement that need not demarcate, on measuring principle still is measurement mechanism, all is superior to other method of testings.Particle sizing based on full light scattering method has received application widely and research in recent years, and many scholar experts have proposed than effective calculation and relevant apparatus.But ubiquity three aspect problems in the existing method and apparatus:

Though one, conventional light scattering method testing tool need be by the instruments such as filter screen of special pore size distribution; But consider collimation; Usually adopt LASER Light Source, through sample cell of various optical device structures, the signal that obtains is sent into computing machine carry out secondary treating then; When especially adopting the multi-wavelength scanning method, light source, optical device and detector etc. all there is certain requirement;

Two, the definite problem of particle refractive index to be measured; Confirm the refractive index of particle to be measured generally can't know particle to be measured and gas, liquid refractive index according to the test formula needs of scattering theory, therefore adopt multi-wavelength to measure; When relating to dispersion relation; Brought basic problem to detection,, then can bring measuring error and uncertainty if take the method estimated;

Three, the problems such as reflection and scattering that have the transparent pool wall of sample cell, the intensity of its emission and scattering will directly have influence on ground unrest, if tested grain graininess is less, concentration is lower, will cause scattered signal faint and be submerged in the noise.

Confirm and sample cell reflection, the scattering problem of described liquid to be measured or gas refractive index to be measured are still the focus that research is paid close attention to so far, not propose good solution.

Summary of the invention

The purpose of the utility model is in order to solve the deficiency of above-mentioned prior art; An a kind of liquid particles automatic detection device based on optical fiber coupling is provided, and device complexity in the full light scattering method particle sizing is high, refractive index to be measured is confirmed and reflection, the scattering problem of sample cell to solve.

The basic thought of the utility model is: utilize computer control high-precision optical adjustment platform; The two end face of the input optical fibre and the output optical fibre that is connected power meter that connect light source is controlled and is locked in the spacing of one 100 μ m; Splash into solution to be measured and form sample pool structure; Utilize multi-wavelength scanning then, theoretical in conjunction with Lorentz model refinement traditional M IE; Propose the multivariate genetic algorithm and realize real-time online detection.

The technical solution of the utility model is following:

A kind of grain testing apparatus; Characteristics are levied and are that its formation comprises: high stability multi wave length illuminating source, input optical fibre, the automatically controlled optics adjustment of first high precision platform, CCD imaging system, the automatically controlled optics adjustment of second high precision platform, output optical fibre, detected with high accuracy device and computing machine; One end of described input optical fibre links to each other with the output terminal of described high stability multi wave length illuminating source; The input end of the described detected with high accuracy device of one termination of described output optical fibre links to each other; The other end of described input optical fibre is fixed on the automatically controlled optics adjustment of described first high precision platform and extends outside the appearance; Be called free end; The other end of described output optical fibre is fixed on the automatically controlled optics adjustment of described second high precision platform and extends outside the appearance; Be called free end, the free end of described input optical fibre and the free end of output optical fibre in opposite directions relatively and be positioned at the camera watch region of described CCD imaging system, described computing machine links to each other with the detected with high accuracy device with described high stability multi wave length illuminating source, the automatically controlled optics adjustment of first high precision platform, CCD imaging system, the automatically controlled optics adjustment of second high precision platform respectively.

The wavelength of described high stability multi wave length illuminating source output and power are by described computer control.

The free-ended relative position of the free end of described input optical fibre and output optical fibre carries out the precision adjustment by automatically controlled optics adjustment platform of described first high precision of described computer control and the automatically controlled optics adjustment of second high precision platform under the monitoring of described CCD imaging system.

The signal that described computing machine receives the input of the described detected with high accuracy device line data of going forward side by side is handled.

A kind of method of utilizing described grain testing apparatus to carry out particle detection, its characteristics are levied and are that this method comprises the following steps:

(1) utensil cleans:

For avoiding artificial pollution, the utensil that will use for experiment must could use through strict cleaning process in advance;

(2) device preheating and initial adjustment:

In order to make light stability, described high stability multi wave length illuminating source and detected with high accuracy device must pass through preheating and reach optimum performance, and be 30 minutes preheating time, measure temperature during experiment and are controlled at 20 ± 5 ℃; Precision through automatically controlled optics adjustment platform of described first high precision of computer control and the automatically controlled optics adjustment of second high precision platform moves; Under the monitoring of described CCD imaging system, carry out the precision adjustment, make the free end of described input optical fibre and the same optical axis of free end of output optical fibre;

(3) calibration solution is measured:

1. calibration solution adopts distilled pure water;

2. the free end face with described input optical fibre and output optical fibre cleans up on the free-ended end face of input optical fibre with isopropyl alcohol respectively; Splash into an amount of pure water with glue head dropper; By surface tension effects; Pure water liquid pearl can be attached on the end face of input optical fibre, again the stepper motor through the automatically controlled optics of computer control second high precision adjustment platform make two free end faces near, also touch until the output optical fibre end face with the liquid pearl; The spacing of regulating two end faces is 100 μ m and locking, forms the adjustable sample cell of pond spacing L;

3. successively export from wavelength 1000nm through the described high stability multi wave length illuminating source of computer control and begin; Be spaced apart 100nm; Until 1300nm amounts to the light of 4 wavelength, described detected with high accuracy device writes down 4 I0 of corresponding optical power value, send computing machine to deposit array in;

(4) liquid to be measured is measured:

1. add particle to be measured in the solvent and place low speed rotation on the sol evenning machine, solution to be measured is fully stirred;

2. blow the calibration solution between the fiber end face off; The drop of drawing solution to be measured then is between two fiber end faces; This moment the stepper motor through the automatically controlled optics of computer control second high precision adjustment platform make both ends of the surface near; Also touch with the liquid pearl until the output optical fibre end face, the spacing of regulating two end faces is 100 μ m and locking, forms the adjustable sample cell of pond spacing L;

3. successively export from wavelength 1000nm through the described high stability multi wave length illuminating source of computer control and begin; Be spaced apart 100nm; Until 1300nm amounts to the light of 4 wavelength, described detected with high accuracy device writes down 4 I of corresponding optical power value, send computing machine to deposit array in;

(5) data processing:

4 I that step (3) is obtained ₀And 4 I obtaining of step (4), and 4 wavelength parameter, bring formula into

$\mathrm{\δ}=\underset{i=1}{\overset{M-1}{\mathrm{\Σ}}}{[\frac{\ln {(I/I_0)}_{{\mathrm{\λ}}_i}}{\ln {(I/I_0)}_{{\mathrm{\λ}}_{i+1}}}-\frac{\underset{j=1}{\overset{N}{\mathrm{\Σ}}}\left[f\left(D_j\right)D_j^2{K}_{\mathrm{ext}}({\mathrm{\λ}}_i,m_i,D_j)\right]}{\underset{j=1}{\overset{N}{\mathrm{\Σ}}}\left[f\left(D_j\right)D_j^2{K}_{\mathrm{ext}}({\mathrm{\λ}}_{i+1},m_{i+1},D_j)\right]}]}^2$

${\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">m</figure-callout>}}^2=\frac{1}{2}\sqrt{{(1+\frac{A^2(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}{{(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}^2+{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">C</figure-callout>}}^2{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2})}^2+{\left(\frac{A^{2}C\stackrel{\&OverBar;}{\mathrm{\&omega;}}}{{(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}^2+{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">C</figure-callout>}}^2{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2}\right)}^2}+\frac{1}{2}(1+\frac{A^2(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}{{(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}^2+{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">C</figure-callout>}}^2{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2})$

Obtain A, B, C, k and

Wherein:

B=10-16 * ω ₀, C=10-16 * γ and

D _jFor getting from 0.1 to 20 successively, every at a distance from 0.1 all value, the number of wavelengths 4 that the M representative is tested, N representative diameter number of partitions: (20-0.1)/and 0.1+1=200, calculate mean grain size and distribution thus.

The technique effect of the utility model is following:

Particle automatic testing method and device based on the optical fiber coupling that the utility model provides have improved the deficiency that existing full light scattering method carries out particle detection from mechanism.The remarkable result of the utility model shows with conventional method and compares, solved it and pre-estimated the error that refractive index to be measured is brought, and well solved the problem of the reflection and the scattering of sample cell; Utilize continuous spectrum to carry out multi-wavelength scanning; Adopt genetic algorithm to carry out optimization computation, finally accomplish calculating such as particle test and size distribution, significantly improved operability and accuracy; Improved the accuracy of online particle test effectively; Reduced the error that bring in the experience sum of errors common sample pond estimated, had the conveniently moving of carrying simultaneously, adjustment is quick, the precision advantages of higher.Help realizing real-time online particle detection fast and accurately.

The utility model grain testing apparatus have volume little, be convenient to move; It is little influenced by vibration interference, need not in advance liquid to be measured to be carried out detecting refractive index, does not need external independent high power laser light source; And there are not the reflection and the scattering of traditional sample pool structure pool wall; The control accuracy of sample cell is high, and the measurement mechanism main body is an optical fiber, changes cheaply easy.The deficiency of existing optical scatter method of testing can be effectively solved, the real-time online particle detection can be realized.

Description of drawings

Fig. 1 is the structural representation of the utility model grain testing apparatus.

Fig. 2 is fiber end face enlarged drawing and sample cell synoptic diagram.

Fig. 3 is concrete utensil cleaning step.

Embodiment

Below in conjunction with embodiment and accompanying drawing the utility model is described further, but should limit the protection domain of the utility model with this.

One, the utility model ultimate principle:

One particle size distribution function is the polydispersion granular system of N (D), and multiple different mean grain size define method can be arranged.In light total scattering method, define the average extinction coefficient K of polydispersion granular system according to the principle of equivalent extinction coefficient _mWith mean diameter D ₃₂As follows:

$K_m=\frac{{\&Integral;\mathrm{<figure-callout\; id="2"\; label="K\; ext\; ND"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">K</figure-callout>}}_{\mathrm{ext}}{\mathrm{ND}}^{}{}^2\mathrm{dD}}{\&Integral;{\mathrm{<figure-callout\; id="2"\; label="ND"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">ND</figure-callout>}}^2\mathrm{dD}}---\left(1\right)$

$D_{32}=\frac{\&Integral;N\left(D\right){\mathrm{<figure-callout\; id="3"\; label="D"\; filenames="DEST\_PATH\_HDA0000098820330000011.png"\; state="\{\{state\}\}">D</figure-callout>}}^3\mathrm{dD}}{\&Integral;{N\left(D\right)\mathrm{<figure-callout\; id="2"\; label="D"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">D</figure-callout>}}^2\mathrm{dD}}---\left(2\right)$

$\ln (I/I_0)=-\frac{\mathrm{\&pi;}}{4}L\underset{a}{\overset{b}{\&Integral;}}N\left(D\right)D^2{K}_{\mathrm{ext}}\mathrm{dD}---\left(3\right)$

Wherein: I ₀Be respectively incident intensity with I and reach through the projection light intensity behind the sample, L is the thickness of sample measurement district (sample cell), and D is an actual particle size, D ₃₂Be called Sauter mean diameter.

Can know K by formula (1) _mNumerical value should be relevant with distribution of particles function N (D), but according to existing research, in the scope of particle diameter parameter α (α=D π/λ (lambda1-wavelength))＜4, the K that obtains by different particle size distribution function N (D) _mWith press D ₃₂The extinction coefficient K that obtains _ExtNumerically differ very little, promptly little with the shape correlation of particle size distribution function.Can use by D for this reason ₃₂The K that tries to achieve _ExtReplace K _mActual computation shows that this scope can expand α=30 to.Formula (1) and formula (2) substitution formula (3) can be got:

$\ln (I/I_0)=-\frac{\mathrm{\&pi;}}{4}\mathrm{LN}{D}_{32}^2{K}_m(\mathrm{\&lambda;},n,D_{32})---\left(4\right)$

Wherein: N is the total number of particles of granular system, and n is a refractive index.So just change into the measurement of a polydispersion granular system and be equivalent to have single diameter D ₃₂The measurement of monodisperse particles system, and provide the mean grain size D of this granular system ₃₂Formula (4) is carried out numerical solution can be changed into:

$\ln \left(\frac{I}{I_0}\right)=-\frac{\mathrm{\&pi;}}{4}\mathrm{\&Delta;DLN}\underset{j=1}{\overset{M_2}{\mathrm{\&Sigma;}}}\left[D_j^2{K}_{\mathrm{ext}}\left(D_j\right)f\left(D_j\right)\right]j=\mathrm{1,2},L,M_2---\left(5\right)$

In the formula: M ₂Be the stepping number of particle diameter, f (D _j) and K _Ext(D _j) be respectively j particle diameter stepping [D _j, D _J+1] particle diameter volume frequency distribution function and the mean value of extinction coefficient, Δ D be the particle diameter stepping at interval.

Introduce the multi-wavelength method of testing, then formula (5) becomes:

$\left[\begin{array}{c}\ln {(I/I_0)}_{{\mathrm{\&lambda;}}_1}\\ M\\ \ln {(I/I_0)}_{{\mathrm{\&lambda;}}_{{\mathrm{<figure-callout\; id="1"\; label="M"\; filenames="DEST\_PATH\_HDA0000098820330000011.png"\; state="\{\{state\}\}">M</figure-callout>}}_1}}\end{array}\right]={\left[A_{\mathrm{ij}}\right]}_{{\mathrm{<figure-callout\; id="1"\; label="M"\; filenames="DEST\_PATH\_HDA0000098820330000011.png"\; state="\{\{state\}\}">M</figure-callout>}}_1\&times;{\mathrm{<figure-callout\; id="2"\; label="M"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">M</figure-callout>}}_2}\&CenterDot;\left[\begin{array}{c}f\left(D_1\right)\\ M\\ M\\ f\left(D_{M_2}\right)\end{array}\right]---\left(6\right)$

$A_{\mathrm{ij}}=-\frac{\mathrm{\&pi;}}{4}\mathrm{\&Delta;}{\mathrm{DLND}}_j^2{K}_{\mathrm{ext}}({\mathrm{\&lambda;}}_i,m_i,D_j)i=\mathrm{1,2},L,{\mathrm{<figure-callout\; id="1"\; label="M"\; filenames="DEST\_PATH\_HDA0000098820330000011.png"\; state="\{\{state\}\}">M</figure-callout>}}_1;j=\mathrm{1,2},L,M_2---\left(7\right)$

Wherein: M ₁Be the number of wavelengths of measuring, M ₂Interval number for the particle division.(6) (7) two formula simultaneous can be got:

$\frac{\ln {(I/I_0)}_{{\mathrm{\&lambda;}}_i}}{\ln {(I/I_0)}_{{\mathrm{\&lambda;}}_{i+1}}}=\frac{\underset{j=1}{\overset{M_2}{\mathrm{\&Sigma;}}}\left[f\left(D_j\right)D_j^2{K}_{\mathrm{ext}}({\mathrm{\&lambda;}}_i,m_i,D_j)\right]}{\underset{j=1}{\overset{M_2}{\mathrm{\&Sigma;}}}\left[f\left(D_j\right)D_j^2{K}_{\mathrm{ext}}({\mathrm{\&lambda;}}_{i+1},m_{i+1},D_j)\right]}i=\mathrm{1,2},L,M_1-1---\left(8\right)$

Introducing has the Lorentz dispersion equation of pervasive meaning:

${\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">m</figure-callout>}}^2=\frac{\sqrt{{(1+{\mathrm{\&chi;}}_r)}^2+{\mathrm{\&chi;}}_i^2}+(1+{\mathrm{\&chi;}}_r)}{2}$

${\mathrm{\&chi;}}_r=\frac{N_e{\mathrm{<figure-callout\; id="2"\; label="q"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">q</figure-callout>}}^2}{{\mathrm{\&epsiv;}}_{0}m}\&CenterDot;\frac{{{\mathrm{\&omega;}}_0}^2-{\mathrm{\&omega;}}^2}{{({{\mathrm{\&omega;}}_0}^2-{\mathrm{\&omega;}}^2)}^2+{\mathrm{\&gamma;}}^2{\mathrm{\&omega;}}^2}---\left(9\right)$

${\mathrm{\&chi;}}_i=\frac{N_e{\mathrm{<figure-callout\; id="2"\; label="q"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">q</figure-callout>}}^2}{{\mathrm{\&epsiv;}}_{0}m}\&CenterDot;\frac{\mathrm{\&gamma;\&omega;}}{{({{\mathrm{\&omega;}}_0}^2-{\mathrm{\&omega;}}^2)}^2+{\mathrm{\&gamma;}}^2{\mathrm{\&omega;}}^2}$

In the formula: N _eBe the resonance subnumber in the granule medium unit volume, q is an electron charge, ε ₀Be permittivity of vacuum, m is an electron mass, and γ is a ratio of damping, ω ₀Be natural frequency, ω=2 π c/ λ are light wave circular frequency, χ _rAnd χ _iBe respectively the real part and the imaginary part of electric polarization coefficient.

Make three unknown numbers:

B=10 ^-16* ω ₀, C=10 ^-16* γ and

Then (9) formula can be rewritten into:

${\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">m</figure-callout>}}^2=\frac{1}{2}\sqrt{{(1+\frac{A^2(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}{{(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}^2+{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">C</figure-callout>}}^2{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2})}^2+{\left(\frac{A^{2}C\stackrel{\&OverBar;}{\mathrm{\&omega;}}}{{(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}^2+{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">C</figure-callout>}}^2{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2}\right)}^2}+\frac{1}{2}(1+\frac{A^2(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}{{(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}^2+{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">C</figure-callout>}}^2{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2})---\left(10\right)$

Consider the multi-wavelength test, adopt the multivariate genetic algorithm for solving, evaluation function is defined as:

$\mathrm{\&delta;}=\underset{i=1}{\overset{M-1}{\mathrm{\&Sigma;}}}{[\frac{\ln {(I/I_0)}_{{\mathrm{\&lambda;}}_i}}{\ln {(I/I_0)}_{{\mathrm{\&lambda;}}_{i+1}}}-\frac{\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}\left[f\left(D_j\right)D_j^2{K}_{\mathrm{ext}}({\mathrm{\&lambda;}}_i,m_i,D_j)\right]}{\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}\left[f\left(D_j\right)D_j^2{K}_{\mathrm{ext}}({\mathrm{\&lambda;}}_{i+1},m_{i+1},D_j)\right]}]}^2---\left(11\right)$

Wherein: K _ExtBeing extinction coefficient, is the function of wavelength, refractive index and particle diameter; F (D) is a particle diameter volume distributed median frequency function, and its expression formula is: $\frac{\mathrm{DV}\left(D\right)}{\mathrm{DD}}=f\left(D\right)=\left(\frac{k}{\stackrel{\&OverBar;}{D}}\right)\&CenterDot;{\left(\frac{k}{\stackrel{\&OverBar;}{D}}\right)}^{k-1}\&CenterDot;\mathrm{Exp}[-{\left(\frac{D_j}{\stackrel{\&OverBar;}{D}}\right)}^k],$

It is the function of zero dimension distribution parameter k and characteristic dimension parameter

.

Therefore five unknown numbers being arranged here is k,

A, B, C, utilizes multi-wavelength and combines the multivariate genetic algorithm can calculate particle grain size distribution.

The same continuation found the solution with identical method mitscherlich's law, can obtain corresponding mean grain size formula:

$D_{32}=\frac{\underset{0}{\overset{\&infin;}{\text{\&Integral;}}}{D}^3\left(\frac{k}{\stackrel{\&OverBar;}{D}}\right)\&CenterDot;{\left(\frac{D}{\stackrel{\&OverBar;}{D}}\right)}^{k-1}\&CenterDot;\exp [-{\left(\frac{D}{\stackrel{\&OverBar;}{D}}\right)}^k]\mathrm{dD}}{\underset{0}{\overset{\&infin;}{\&Integral;}}{D}^2\left(\frac{k}{\stackrel{\&OverBar;}{D}}\right){\left(\frac{D}{\stackrel{\&OverBar;}{D}}\right)}^{k-1}\&CenterDot;\exp [-{\left(\frac{D}{\stackrel{\&OverBar;}{D}}\right)}^k]\mathrm{dD}}---\left(12\right)$

Two, emulated data:

Improve the genetic algorithms use binary coding, roulette is selected, 2 crossover operators, and discrete variation, crossover probability are 0.8, the variation probability is 0.7/ chromosome length, initial population size 100,400 generations of evolutionary generation.With three parameter substitutions of A, B, C (11) formula that genetic algorithm is obtained, obtain refractive index.Analog result is as shown in table 1, than adopting average refractive index method inverting parameter to be measured, considered the caused variations in refractive index of effect of dispersion after, inversion result and setting value are more approaching.When reality is tested; Because factors such as instrument error, environment temperature can cause certain measuring errors; In order to verify the reliability of this method, on the delustring measured value of the first, the five wavelength at NO.7 place, added 2% random noise respectively to represent measuring error; The result shows that inversion error is lower than 1%, and in fact 2% random noise is very big.This shows that the relative medium refraction index with it of mean grain size that adopts the method count particles is fully feasible.

The genetic algorithm evolutionary generation changed for 200 generations into, and other parameter is provided with constant.Analog result is as shown in table 2, can find out that inversion result and setting value are quite approaching, on the delustring measured value of first, the 3rd wavelength at NO.4 place, introduces 2% random noise, and inversion error is lower than 10%, and is more more accurate than average refractive index method.Can be finally inversed by particle grain size distribution more accurately after can finding out the situation of considering variations in refractive index, though after having added 2% random noise and theoretical distribution still can coincide relatively goodly.

The inversion result of table 1 mean grain size and refractive index

The inversion result that R-R distributes under table 2 non-standalone mode

Three, the liquid particles pick-up unit that is coupled based on optical fiber

Grain testing apparatus is as shown in Figure 1, and Fig. 2 has then shown the enlarged image of two fiber end faces, and the structural equivalents in the circle is in the sample cell of conventional particles tester.Visible by figure; The formation of the utility model grain testing apparatus comprises: high stability multi wave length illuminating source 1, input optical fibre 2, the automatically controlled optics adjustment of first high precision platform 3, CCD imaging system 4, the automatically controlled optics adjustment of second high precision platform 5, output optical fibre 6, detected with high accuracy device 7 and computing machine 8; One end of described input optical fibre 2 links to each other with the output terminal of described high stability multi wave length illuminating source 1; The input end of the described detected with high accuracy device 7 of one termination of described output optical fibre 6 links to each other; The other end of described input optical fibre 2 is fixed on the automatically controlled optics adjustment of described first high precision platform 3 and extends outside the appearance; Be called free end; The other end of described output optical fibre 6 is fixed on the automatically controlled optics adjustment of described second high precision platform 5 and extends outside the appearance; Be called free end, the free end of the free end of described input optical fibre 2 and output optical fibre 6 in opposite directions relatively and be positioned at the camera watch region of described CCD imaging system 4, described computing machine 8 links to each other with detected with high accuracy device 7 with described high stability multi wave length illuminating source 1, the automatically controlled optics adjustment of first high precision platform 3, CCD imaging system 4, the automatically controlled optics adjustment of second high precision platform 5 respectively.

The wavelength of described high stability multi wave length illuminating source 1 output and power are by described computing machine 8 controls.

The free-ended relative position of described input optical fibre 2 and output optical fibre 6 carries out the precision adjustment by automatically controlled optics adjustment platform 3 of described first high precision of described computing machine 8 controls and the automatically controlled optics adjustment of second high precision platform 5 under the monitoring of described CCD imaging system 4.

The signal that described computing machine 8 the receives described detected with high accuracy device 7 inputs line data of going forward side by side is handled.

Two optical fiber are installed in respectively on two the automatically controlled optics adjustment of high precision platforms.The automatically controlled optics of high precision adjustment platform is controlled the position of three directions and lockable position by computing machine.The CCD imaging system monitoring optical fiber endface position and the drop situation that side-looking are arranged simultaneously and overlook both direction.The high stability multi wave length illuminating source is in order to providing the light signal of stable a plurality of wavelength, and the detected with high accuracy device is surveyed the light intensity signal of different wave length, and sends into computing machine and accomplish data processing.High stability light source, photo-detector, the automatically controlled adjustment platform of high precision all pass through gpib interface and are connected with computing machine; Realize adjustment platform control of position, locking through software programming; The conversion of optical source wavelength, the data acquisition of photo-detector, and final data is handled.The algorithm of system uses MATLAB to carry out simulating, verifying, and finally uses VB to carry out program composition.Two high-precision opticals that computerized control are adjusted platforms, and the plane lapping FC/PC end face that can make two optical fiber is to each other apart from being controlled at 40-100 μ m; When splashing into calibration solution and liquid to be measured, because surface tension effects makes drop between two fiber facet, form connection, constituted theoretic sample pool structure, promptly be equivalent to the sample cell of conventional particles checkout equipment, like Fig. 2.

The implementation process of utility model and result thereof:

1, the parameter study of sample under different spacing:

The nominal diameter that confirmatory experiment uses is respectively the U.S. Du Ke company polystyrene latex ball standard particle of 1.00m (sample 1) and 2.00m (sample 2), and base liquid is a pure water, and the concentration of two kinds of samples is respectively 0.5% and 1.0%.Particle proportion is about 1, after 30 minutes magnetic stirs, can guarantee even suspension, measures fast then.The optical fiber that constitutes sample cell is that commercially available optical communication is with band FC/PC wire jumper and by the single mode silica fibre of plane lapping.Having selected 4 centre wavelengths for use, is respectively 1000nm, 1100nm, 1200nm and 1300nm.Each sample repeated test 6 times, the extinction value I/I that surveys ₀Average and classify as shown in table 3, the table 4.

Table 3 sample 1 test value.

Table 4 sample 2 test values

2, carry out algorithm process:

Utilize genetic algorithm to handle, calculate the particle mean grain size D32 of two kinds of samples under the spacing of different ponds respectively, shown in the table 5.

Table 5 particle diameter calculated value

Pond spacing (μ m)	Mean grain size (sample 1)	Mean grain size (sample 2)
			10	2.477	4.614
20	1.645	2.951

80	0.974	1.972
			100	0.981	1.985
180	0.689	1.546
			300	0.564	1.278

3, interpretation of result:

Actual measurement situation by shown in table 3, the table 4 can know, when the sample cell spacing different, promptly measuring distance not simultaneously, the result difference property that records is very big.Under situation with a kind of sample (the maintenance sample concentration is constant), extinction value I/I ₀Increase with measuring distance reduces.In the measuring distance scope of 10 μ m to 20 μ m; Extinction value is very high, and this is because the end face of incident optical and outgoing optical fiber leaves very closely, makes most of scattered light also can be received by outgoing optical fiber; I is that the light intensity sum by a part of scattered light and transmitted light constitutes, and makes I become bigger.Increase along with measuring distance; The end face of incident optical and outgoing optical fiber gradually away from; What outgoing optical fiber can receive also reduces through the scattered light behind the particulate samples gradually, and when both ends of the surface were enough far away, scattered light all can't be received by outgoing optical fiber basically; Outgoing optical fiber receptible nearly all be pure transmitted light, this moment measurement comparatively accurate.Along with both ends of the surface further away from each other, a part of transmitted light can't be received by outgoing optical fiber, like this, make I become less, extinction value I/I ₀Become very low.Can find out that by table 5 as preceding said, when sample cell spacing during at 80 μ m and 100 μ m, the sample particle diameter and the normal diameter that record are comparatively identical.A large amount of experiment proofs of this paper can both comparatively be met the measurement result of its nominal diameter for two kinds of standard particles of experiment usefulness in the scope of 80 μ m to 100 μ m.When distance is excessive or too small, with having bigger measuring error.

In order further to verify the accuracy of refractive index disposal route described herein, sample thief pond spacing L=100 μ m surveys institute grain diameter and adopts conventional refractive index technique of estimation relatively, adopts the valuation of conventional refractive index estimation method refractive index to get 1.57, ignores chromatic dispersion.Corresponding Suo Taier mean grain size D ₃₂, mean bias and repeatability lists in table 6, can find out, the mean bias of this paper method and repeatability is all less than 5%, and is superior to conventional refractive index estimation method.

Table 6 particle diameter test result

The basic operation flow process specifies:

(1) utensil cleans

For avoiding artificial pollution; The utensil that will use for experiment must could use through strict cleaning process in advance; The cleaning synoptic diagram is as shown in Figure 3, and the utensil that can use in the experiment comprises large beaker, cucurbit, blender jar, spirit lamp, drop bottle, glue head dropper, dry ball, cotton rod etc.The utensil that wherein need clean is a large beaker, blender jar and cucurbit, and drop bottle, glue head dropper, cleaning process is all accomplished in ultrasonic cleaning machine.Existing is example with the large beaker, and concrete utensil cleaning step is as shown in Figure 3 as follows.

(2) instrument preheating

In order to make light stability, laser instrument and power meter must pass through preheating and reach optimum performance, and be 30 minutes preheating time, measure temperature during experiment and are controlled at 20 ± 5 ℃.

(3) calibration solution is measured

In the experiment, calibration solution adopts distilled pure water, in order to eliminate the influence of calibration solution, does not add particle to be measured earlier and measures.One end of input optical fibre is connected to high stability multi-wavelength light source output terminal, and the other end is connected to an end of the automatically controlled optics adjustment of high precision platform support; Simultaneously an end of output optical fibre is connected to the input end of detected with high accuracy device, the other end is connected to the other end of the automatically controlled optics adjustment of high precision platform support; Two fiber end faces are cleaned up with isopropyl alcohol respectively.On the end face of input optical fibre; Splash into an amount of pure water with glue head dropper, by surface tension effects, pure water liquid pearl can be attached on the end face of input optical fibre; This moment the stepper motor through the automatically controlled optics of computer control high precision adjustment platform make both ends of the surface near; Also touch until the output optical fibre end face, regulate the spacing and the locking of two end faces, form the adjustable sample cell of pond spacing L with the liquid pearl.Pass through spacing computer controlled built in 100 μ m.Then through the computer control multi wave length illuminating source; Light source is successively exported from wavelength 1000nm begun, be spaced apart 100nm, until 1300nm amounts to the light of 4 wavelength; Corresponding detection also along with receiving from wavelength 1000nm to 1300nm, is spaced apart the optical power value of the different wave length of 100nm.These 4 values are 4 I0 corresponding to different wave length in the formula, obtain data by computing machine through the GPIB communication interface and deposit array in, to treat the computing in later stage.

(4) liquid configuration to be measured

Liquid to be measured is placed on the sol evenning machine, adds and stir particle (cleaning) low speed rotation, liquid to be measured is fully stirred.Blow the calibration solution between the fiber end face off; Draw drop to be measured then between two fiber end faces; Same as (3) calibration solution is measured described step and is surveyed; By computer program control, obtain the optical power value under different wave length, i.e. 4 I in the formula corresponding to different wave length through the GPIB communication interface;

(5) data processing

4 I that 4 I0 that will obtain through step (3) and step (4) obtain; And 4 wavelength parameter; Bring in the formula (11), and convolution (10) can be obtained A, B, C, k and

calculates mean grain size and distribution thus.

Because a lot of parameters and function are implicit functions in the formula (11); Therefore in computing machine through software programming a series of equations; In order to pass through 4 I0 and 4 I; And 4 λ, can calculate A, B, C, k and

for convenience's sake: all formula is following by the iteration sequence identification:

${\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">m</figure-callout>}}^2=\frac{1}{2}\sqrt{{(1+\frac{A^2(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}{{(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}^2+{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">C</figure-callout>}}^2{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2})}^2+{\left(\frac{A^{2}C\stackrel{\&OverBar;}{\mathrm{\&omega;}}}{{(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}^2+{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">C</figure-callout>}}^2{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2}\right)}^2}+\frac{1}{2}(1+\frac{A^2(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}{{(B^2-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2)}^2+{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">C</figure-callout>}}^2{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}^2})---\left(13\right)$

$\mathrm{\&delta;}=\underset{i=1}{\overset{M-1}{\mathrm{\&Sigma;}}}{[\frac{\ln {(I/I_0)}_{{\mathrm{\&lambda;}}_i}}{\ln {(I/I_0)}_{{\mathrm{\&lambda;}}_{i+1}}}-\frac{\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}\left[f\left(D_j\right)D_j^2{K}_{\mathrm{ext}}({\mathrm{\&lambda;}}_i,m_i,D_j)\right]}{\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}\left[f\left(D_j\right)D_j^2{K}_{\mathrm{ext}}({\mathrm{\&lambda;}}_{i+1},m_{i+1},D_j)\right]}]}^2---\left(14\right)$

Wherein: Dj is for getting from 0.1 to 20 successively, whenever at a distance from 0.1 all value.

The number of wavelengths (4) of M representative test

N representative diameter number of partitions: (20-0.1)/0.1+1=200

$f\left(D_j\right)=\left(\frac{k}{\stackrel{\&OverBar;}{D}}\right)\&CenterDot;{\left(\frac{k}{\stackrel{\&OverBar;}{D}}\right)}^{k-1}\&CenterDot;\exp [-{\left(\frac{D_j}{\stackrel{\&OverBar;}{D}}\right)}^k]---\left(15\right)$

Can see and also have a Kext function here; This function is an implicit function, and just a function of functions that is formed by a lot of combination of function is listed the correlated expression formula below; All these expression formulas all program, and realize automatic computing through computer program.

$K_{\mathrm{ext}}=\frac{2}{{\mathrm{\&alpha;}}^2}\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}(2n+1)\mathrm{Re}(a_n+b_n)---\left(16\right)$

Wherein:

this shows extinction ratio direct and an and two functional dependences of bn.

Then write out the expression formula of an and bn below successively, these two expression formulas neither directly can be listed as the function of writing certainly, still need further recursive resolve.

$a_n=\frac{{\mathrm{\&psi;}}_n\left(a\right){\mathrm{\&psi;}}_n^{\&prime;}\left(\mathrm{ma}\right)-m{{\mathrm{\&psi;}}^{\&prime;}}_n\left(a\right){\mathrm{\&psi;}}_n\left(\mathrm{ma}\right)}{{\mathrm{\&zeta;}}_n\left(a\right){\mathrm{\&psi;}}_n^{\&prime;}\left(\mathrm{ma}\right)-m{{\mathrm{\&zeta;}}^{\&prime;}}_n\left(a\right){\mathrm{\&psi;}}_n\left(\mathrm{ma}\right)}$

$b_n=\frac{{\mathrm{m\&psi;}}_n\left(a\right){\mathrm{\&psi;}}_n^{\&prime;}\left(\mathrm{ma}\right)-{{\mathrm{\&psi;}}^{\&prime;}}_n\left(a\right){\mathrm{\&psi;}}_n\left(\mathrm{ma}\right)}{m{\mathrm{\&zeta;}}_n\left(a\right){\mathrm{\&psi;}}_n^{\&prime;}\left(\mathrm{ma}\right)-{{\mathrm{\&zeta;}}^{\&prime;}}_n\left(a\right){\mathrm{\&psi;}}_n\left(\mathrm{ma}\right)}---\left(17\right)$

Here still need further recurrence:

${\mathrm{\&psi;}}_n\left(z\right)={\left(\frac{\mathrm{z\&pi;}}{2}\right)}^{\frac{1}{2}}{J}_{n+\frac{1}{2}}\left(z\right)$

${\mathrm{\&zeta;}}_n\left(z\right)={\left(\frac{\mathrm{z\&pi;}}{2}\right)}^{\frac{1}{2}}{H}_{n+\frac{1}{2}}\left(z\right)---\left(18\right)$

In the formula: Z can be α and m α in the following formula, and m is a refractive index here, just the m in the formula (13).Wherein Jn+1/2 (z) and Hn+1/2 (z) are semi-integer order Bessel function and second type of Hankel function.Can find out by following formula, as long as derive n (z) and n (z).Recursion formula just can be obtained the value of an and bn.And Bessel's function and Hankel function all satisfy following recursion formula:

$Y_{n+1}\left(z\right)=\frac{2n}{z}{Y}_n\left(z\right)-Y_{n-1}\left(z\right)$

$Y_n^{\&prime;}\left(z\right)=\frac{1}{2}[Y_{n-1}\left(z\right)-Y_{n+1}\left(z\right)]---\left(19\right)$

(18) are taken back in (19) can get:

${\mathrm{\&psi;}}_n\left(z\right)=\frac{2n-1}{z}{\mathrm{\&psi;}}_{n-1}\left(z\right)-{\mathrm{\&psi;}}_{n-2}\left(z\right)$

${\mathrm{\&zeta;}}_n\left(z\right)=\frac{2n-1}{z}{\mathrm{\&zeta;}}_{n-1}\left(z\right)-{\mathrm{\&zeta;}}_{n-2}\left(z\right)$

${{\mathrm{\&psi;}}^{\&prime;}}_n\left(z\right)={\mathrm{\&psi;}}_{n-1}\left(z\right)-\frac{\mathrm{<figure-callout\; id="2"\; label="n"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">n</figure-callout>}}{2}{\mathrm{\&psi;}}_n\left(z\right)$

${{\mathrm{\&zeta;}}^{\&prime;}}_n\left(z\right)={\mathrm{\&zeta;}}_{n-1}\left(z\right)-\frac{\mathrm{<figure-callout\; id="2"\; label="n"\; filenames="DEST\_PATH\_HDA0000098820330000012.png"\; state="\{\{state\}\}">n</figure-callout>}}{2}{\mathrm{\&zeta;}}_n\left(z\right)---\left(20\right)$

In conjunction with starting condition:

ψ ₀(z)＝sin?z

${\mathrm{\&psi;}}_1\left(z\right)=\frac{1}{z}\sin z-\cos z$

ζ ₀(z)＝sin?z+i?cos?z

${\mathrm{\&zeta;}}_1\left(z\right)=\frac{1}{z}(\sin z+i\cos z)-(\cos z-i\sin z)={\mathrm{\&psi;}}_1\left(z\right)+i(\frac{\cos z}{z}+\sin z)---\left(21\right)$

This shows; Through Computerized Editing a comprehensive function that comprises all formula between formula (13)-(21) handle subfunction; This operating function program only need provide 4 wavelength, 4 I0 and 4 I, can obtain five parameters to be asked, and then obtains required mean grain size and distribution.

Claims (4)

1. grain graininess pick-up unit; Be characterised in that its formation comprises: high stability multi wave length illuminating source (1), input optical fibre (2), the automatically controlled optics adjustment of first high precision platform (3), CCD imaging system (4), the automatically controlled optics adjustment of second high precision platform (5), output optical fibre (6), detected with high accuracy device (7) and computing machine (8); One end of described input optical fibre (2) links to each other with the output terminal of described high stability multi wave length illuminating source (1); One end of described output optical fibre (6) links to each other with the input end of described detected with high accuracy device (7); The other end of described input optical fibre (2) is fixed on the automatically controlled optics adjustment platform of described first high precision (3) and extends outside the appearance; Be called free end; The other end of described output optical fibre (6) is fixed on the automatically controlled optics adjustment platform of described second high precision (5) and extends outside the appearance; Be called free end; The free end of the free end of described input optical fibre (2) and output optical fibre (6) in opposite directions relatively and be positioned at the camera watch region of described CCD imaging system (4), described computing machine (8) links to each other with detected with high accuracy device (7) with described high stability multi wave length illuminating source (1), the automatically controlled optics adjustment of first high precision platform (3), CCD imaging system (4), the automatically controlled optics adjustment platform of second high precision (5) respectively.

2. grain graininess pick-up unit according to claim 1 is characterized in that the wavelength and the power of described high stability multi wave length illuminating source (1) output is controlled by described computing machine (8).

3. grain graininess pick-up unit according to claim 1; The free-ended relative position that it is characterized in that described input optical fibre (2) and output optical fibre (6) carries out the precision adjustment by automatically controlled optics adjustment platform (3) of described first high precision of described computing machine (8) control and the automatically controlled optics adjustment of second high precision platform (5) under the monitoring of described CCD imaging system (4).

4. grain graininess pick-up unit according to claim 1 is characterized in that signal that described computing machine (8) receives described detected with high accuracy device (7) the input line data of going forward side by side handles.

Images