JP2008111810A - Comparison method of light intensity distribution data of diffracted/scattered light, and particle size distribution measuring device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a comparison method capable of clarifying a difference numerically by enlarging the difference to the utmost in numerical values of indexes showing the difference, when comparing each intensity distribution data of diffracted/scattered light having a small difference. <P>SOLUTION: This method includes a process for irradiating laser light to a dispersed particle group, a process for acquiring light intensity distribution data having each detected light quantity data as a component by detecting diffracted/scattered light from the dispersed particle group by a plurality of photodetection elements, and a process for comparing each light intensity distribution data acquired respectively relative to two different dispersed particle groups. The comparison process includes a process for treating each light intensity distribution data as a vector having each detected light quantity data as components, and calculating an intersectional angle index related to an intersectional angle formed by the two light intensity distribution data emphatically by operating a weight matrix thereto; and a process for determining the degree of a difference between the two light intensity distribution data based on the intersectional angle index determined in the calculation process. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a method for comparing the light intensity distribution data of diffracted / scattered light emitted from a dispersed particle group irradiated with laser light, and the particle size distribution for measuring the particle size distribution of the particle group using the laser diffraction / scattering method. The present invention relates to a measurement apparatus, and more particularly to a technique for enabling quantitative comparison of light intensity distribution data of diffracted / scattered light.

  In the particle size distribution measurement by the conventional laser diffraction / scattering method, diffracted / scattered light from dispersed particle groups (particle groups dispersed in a medium such as liquid, gas, solid) irradiated with laser light as measurement light. Is used to detect spatial light intensity distribution data that uses detected light quantity data of each light detecting element as a vector component and converts it into particle size distribution data using a calculation algorithm related to the refractive index of the particles. It has become. This particle size distribution measurement method using the laser diffraction / scattering method has a very wide measurable particle size range, a short measurement time, and excellent reproducibility. In various fields such as food products and pharmaceuticals, it is used for evaluation of newly developed products in the research stage and product quality control.

  However, in the case of conventional particle size distribution measurement, the refractive index of the particles is set to an appropriate value when selecting a calculation algorithm for converting light intensity distribution data (vector) into particle size distribution data (select calculation conditions). There is a need to. Since it is difficult to directly measure the refractive index of particles, the values obtained from scientific dictionaries are substituted. However, if there is an error between the refractive index of actual particles, accurate particle size distribution data cannot be obtained. .

  Whether or not the set refractive index (calculation conditions) is appropriate can be checked by converting the particle size distribution data back to light intensity distribution data again using an inverse conversion algorithm, and by checking the degree of coincidence of both light intensity distribution data before and after conversion. (For example, refer to Patent Document 1). The higher the coincidence between the two light intensity distribution data, the more appropriate the refractive index. Therefore, it is conceivable to select and set a refractive index that matches well. However, as described above, since an accurate comparison result between the light intensity distribution data is not easy, it is not easy to set an appropriate refractive index after all.

In order to solve this problem, the light intensity distribution data is treated as a vector having each detected light quantity data as a component, and based on an “intersection angle index” (for example, cosine of both vectors) related to the intersection angle formed by the two light intensity distribution data. Invented a method of quantifying and judging the degree of difference between these two light intensity distribution data (see Patent Document 2), and made it possible to quantitatively compare the light intensity distribution data of diffracted / scattered light. .
That is, when the graph pattern matches (the light intensity distribution data matches) by capturing the difference between the graph patterns (distribution patterns) of the two light intensity distribution data as the intersection angle formed by the vectors, the intersection angle of the vectors is 0. The degree (cosine value is 1) can be expressed quantitatively as 90 degrees (cosine value is 0) if the graph patterns do not match at all.
Japanese Patent Application Laid-Open No. 07-325026 Japanese Patent No. 3633169

According to the method disclosed in Patent Document 2 described above, the value of the crossing angle index that reflects the magnitude of the difference is obtained according to the magnitude of the difference between the two light intensity distribution data to be compared. Therefore, when the difference between the two intensity distribution data is visually large (see FIG. 2 of Patent Document 2), the numerical value of the crossing angle index greatly varies, so that the difference is expressed greatly.
However, when this method is applied to the comparison of the light intensity distributions of the ones that do not have a large difference in appearance (see FIG. 3 of Patent Document 2), a large difference is obtained as a numerical value as an index representing the difference between the distribution data. It was not easy to make a comparison.

  Therefore, even when comparing light intensity distributions that do not have a large difference in appearance, it is devised to produce as large a difference as possible in the numerical value of the index representing the difference between the distribution data. It is an object of the present invention to provide a comparison method for numerically clarifying a difference in data, and to provide a particle size distribution measuring apparatus having such a function.

  In order to solve the above-mentioned problems, a method for comparing the light intensity distribution data of diffracted / scattered light according to the present invention includes irradiating a dispersed particle group with laser light, and diffracting / scattered light from the dispersed particle group. Light intensity distribution data having each detected light quantity data as a component is obtained by detection with a light detecting element. When comparing the obtained light intensity distribution data for two different dispersed particle groups, each light intensity distribution data is treated as a vector having each detected light quantity data as a component, and each vector contains the light intensity distribution data. An appropriate weight matrix capable of emphasizing the portion where the feature appears is applied, and then an intersection angle index related to the intersection angle formed by these two light intensity distribution data is calculated. Then, the degree of difference between the two light intensity distribution data is determined based on the calculated intersection angle index.

  That is, when calculating the crossing angle index used when comparing light intensity distribution data (vectors), the light intensity vector is not subjected to calculation as it is, but is multiplied by an appropriate weight matrix before being used for calculation. We inferred that it would be possible to make a larger difference between the values of the crossing angle index corresponding to the high and low degrees.

In order to examine the validity of this inference, the four light intensity distribution data 1 to 4 that are slightly different in appearance as shown in FIG. 2 were checked. The light intensity distribution data 1 and 1-4 were compared. As the intersection angle index, the cosine of the intersection angle (cos θ) was used. The calculation results when the weight matrix (here, the weight matrix (1a) of the inverse differential function described later) is not multiplied and when it is multiplied are as follows.
Note that, as shown in FIG. 2, a plurality of light intensity distribution data that are slightly different in appearance are generated even when actually comparing different samples, but the refractive index of a particle size distribution measuring apparatus described later is used. This occurs in the case of comparison of light intensity distribution data in an algorithm used for selection.

<When weight matrix is not multiplied>
1: Combination of light intensity distribution data: 1, 1, cos θ: 1
2: Combination of light intensity distribution data: 1, 2, cos θ: 0.997727
3: Combination of light intensity distribution data: 1, 3, cos θ: 0.993703
4: Combination of light intensity distribution data: 1, 4, cos θ: 0.990132
<When weight matrix is multiplied>
1: Combination of light intensity distribution data: 1, 1, cos θ: 1
2: Combination of light intensity distribution data: 1, 2, cos θ: 0.9976631
3: Combination of light intensity distribution data: 1, 3, cos θ: 0.993165
4: Combination of light intensity distribution data: 1, 4, cos θ: 0.988670

  When the weight matrix is multiplied, a larger difference can be generated in the cosine of the crossing angle than when the weight matrix is not multiplied, and the difference in the degree of coincidence can be clarified. This is graphed in FIG.

  By the way, what kind of shape is suitable as the weight matrix is determined from the light intensity distribution of FIG. 2 in this case. As can be seen from FIG. 2, the difference between the four light intensity distributions is the left part and the right part centering on sensor element number 59 in which the light intensity distribution shows a peak. Therefore, a weight matrix that emphasizes the difference in this portion is considered suitable. Therefore, in the above example, as described above, the weight matrix of the inverse differential function represented by the following equation (1a) is adopted.

In equation (1a), i is the sensor element number, i _p is the sensor element number (= 59) at which the light intensity peaks, and σ is a parameter (standard deviation) that determines the spread of the distribution, and is expressed by the following equation. In this example, σ = 3. FIG. 4 shows the equation (1) as a graph.

  As the weight matrix, a differential function or an integral function represented by the following expressions (1b) and (1c) may be used instead of this depending on the light intensity data to be compared. In other words, an optimum one may be used in accordance with the light intensity distribution data (actually, calculation is performed using each weight matrix, and the optimum one may be selected later).

Therefore, according to the comparison method of the light intensity distribution data of the diffracted / scattered light, the crossing angle index corresponding to the crossing angle of the vectors of the two light intensity distribution data is calculated, and the numerical value of the obtained crossing angle index is quantitative. You can see the comparison result. For example, when the crossing angle index is cosθ of the crossing angle of both vectors, cosθ = 1 if the two light intensity distribution data completely match, and cosθ = 0 if the two light intensity distribution data do not match at all. (Cos θ is never negative), and the larger cos θ, the higher the coincidence of both light intensity distribution data.
Naturally, if the difference between the two light intensity distributions to be compared is small, the difference between the cosines of the crossing angles is also small. Therefore, if the calculation is performed after multiplying by the weight matrix, the difference between the numerical values of the cosines of the intersection angles calculated from the two closer light intensity distributions becomes larger.

  Further, the particle size distribution apparatus of the present invention, which is made from another viewpoint, detects light irradiating means for irradiating a dispersed particle group with laser light, and diffracted / scattered light from the dispersed particle group by a plurality of light detecting elements. The light detection means for obtaining light intensity distribution data having each detected light quantity data as a component, and the actually measured light intensity distribution data using a plurality of types of coefficient matrices related to the refractive index of the particles constituting the dispersed particle group. The particle size distribution obtaining means for converting into the particle size distribution data corresponding to each coefficient matrix by the calculation algorithm and the particle size distribution data for each coefficient matrix are converted into each coefficient matrix by the inverse operation algorithm using the coefficient matrix. A light intensity distribution reverse finding means for inversely converting into corresponding light intensity distribution data, and comparing the actually measured light intensity distribution data with the inversely converted light intensity distribution data for each coefficient matrix; From the plurality of inversely converted light intensity distribution data, search for the inversely converted light intensity distribution data having a high degree of coincidence with the actually measured light intensity distribution data, and to the inversely converted light intensity distribution data. Light intensity distribution comparison means for selecting a corresponding coefficient matrix as an optimal calculation condition, and particle size distribution determination for determining the particle size distribution data obtained using the selected coefficient matrix as an appropriate particle size distribution of the dispersed particle group In the particle size distribution measuring apparatus, the light intensity distribution comparison means uses the measured light intensity distribution data and the inversely converted light intensity distribution data as vectors having respective detected light amount data as components. The intersection angle index related to the intersection angle between the measured light intensity distribution data and the inversely converted light intensity distribution data for each coefficient matrix By comparing the intersection angle index calculation means for calculating the measured light intensity distribution data and the intersection angle index corresponding to each calculated coefficient matrix, the inverse of the measured light intensity distribution data having a high degree of coincidence. Selection means is provided for searching for converted light intensity distribution data and selecting a coefficient matrix corresponding to the inverse converted light intensity distribution data as an optimum calculation condition.

  And also in this particle size distribution apparatus, it is preferable to use any of the above-mentioned (1a), (1b) and (1c) as the weight matrix.

  According to the particle size distribution measuring apparatus of the present invention, when obtaining the particle size distribution data of the particle group, the detected light amount data of each light detecting element that detects the diffracted / scattered light from the dispersed particle group irradiated with the laser beam is used as the vector component. Is converted into particle size distribution data by an arithmetic algorithm using a coefficient matrix related to the refractive index of the particles. Here, since the refractive index of the particles has not been determined, a plurality of types of refractive indexes are appropriately set, and calculation is performed using a plurality of types of coefficient matrices (calculation conditions) related to the respective refractive indexes. Therefore, the particle size distribution data is obtained for each coefficient matrix.

  Subsequently, each particle size distribution data is inversely converted into light intensity distribution data by an inverse operation algorithm. If you find the light intensity distribution data for each inversely converted coefficient matrix that has the closest value to the measured light intensity distribution data, the coefficient matrix used to determine the inversely converted light intensity distribution data (Calculation conditions) was appropriate. Therefore, the light intensity distribution data inversely converted from the actually measured light intensity distribution data (the light intensity distribution data converted in reverse is often “light intensity distribution data with no apparent difference”). Are treated as vectors, and an intersection angle index of both vectors is obtained for each coefficient matrix. By comparing these intersection angle indexes, the light intensity distribution data having the highest degree of coincidence is searched for the actually measured light intensity distribution data, and the optimum coefficient matrix is determined. The particle size distribution data calculated using this coefficient matrix gives valid particle size distribution data of the dispersed particle group to be measured.

In this series of operations, in the vicinity of the optimum refractive index, there is a high possibility that the value difference between the light intensity distribution data corresponding to each inversely transformed coefficient matrix will be small, but the weight matrix is applied. By doing so, this difference can be enlarged, and it becomes easy to arrive at the optimum coefficient matrix.

  According to the comparison method of the present invention, the comparison result of the intensity distribution data is quantitatively shown in the form of an intersection angle index corresponding to the intersection angle of both vectors when the two light intensity distribution data are handled as vectors. Comparison of light intensity distribution data can be performed easily and accurately.

  According to the particle size distribution measuring apparatus of the present invention, a plurality of light intensity distribution data actually measured for the particle group to be measured and a plurality of coefficients obtained by inverse conversion for each coefficient matrix related to an appropriately set refractive index. When selecting the optimal coefficient matrix based on the degree of coincidence with the light intensity distribution data, the intersection angle index between the measured light intensity distribution data and the light intensity distribution data obtained by inverse conversion is used to determine the Since the degree of coincidence can be determined quantitatively, the coefficient matrix can be easily and accurately selected in the execution process of the calculation algorithm for particle size distribution measurement.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of the particle size distribution measuring apparatus of the embodiment.

  In the particle size distribution measuring apparatus of the embodiment, as shown in FIG. 1, a laser light source 4 that irradiates a parallel particle beam via a collimator 3 to a dispersed particle group 2 in a sample cell 1 made of a transparent material; Optical sensors 5a to 5c arranged in space so as to detect diffracted / scattered light from the dispersed particle group 2 are provided. The optical sensor 5a is a ring detector type forward scattering / diffracted light detection sensor, and the detection surface is divided into a ring shape or a semi-ring shape corresponding to the diffraction / scattered light image formed in a ring shape by the condenser lens 6. Each divided section becomes one photodetecting element. The optical sensor 5b is a side scattered light detection sensor, and the optical sensor 5c is a backscattered light detection sensor. Further, in the embodiment apparatus, there is provided a dispersion tank 8 provided with a stirrer 7 for stirring the liquid medium and the particle group to disperse the particle group in the liquid medium. The sample cell 1 and the dispersion tank 8 are connected by a flow path provided with a pump 9, and the mixture of the liquid medium and the particle group is circulated between the sample cell 1 and the dispersion tank 8.

  The apparatus according to this embodiment includes a preamplifier 11 that amplifies the detection signals of the optical sensors 5a to 5c and an A / D conversion unit 12 that converts the amplified detection signal into a digital signal. The output signal from the A / D conversion unit 12 is stored in the light intensity distribution memory 13 as light intensity distribution data using each detected light amount data as a component. Further, the CPU 14 converts the actually measured light intensity distribution data into particle size distribution data by a calculation algorithm using a coefficient matrix related to the refractive index of the particles, or inverse conversion algorithm for the particle size distribution data obtained by the calculation. In addition to performing the inverse conversion to the light intensity distribution data, the calculation of the intersection angle index corresponding to the intersection angle of the two light intensity distribution data is also performed. That is, the CPU 14 corresponds to a particle size distribution obtaining unit, a light intensity distribution inverse obtaining unit, a light intensity distribution comparing unit (including an emphasized intersection angle index calculating unit and a selecting unit), and a particle size distribution determining unit in the present invention.

  In addition, the apparatus according to the embodiment includes a coefficient holding unit 15 that holds a coefficient matrix and a weight matrix (weight matrix) calculated by the CPU 14 based on the refractive index input from the operation unit 16 and used in an arithmetic algorithm. Further, the embodiment apparatus includes, as the output unit 17, a video display device such as a TV monitor or a liquid crystal panel or a printing device such as a printer.

Next, a description will be given of a process for obtaining a calculation algorithm and an intersection angle index for conversion between light intensity distribution data and particle size distribution data, which is executed in the embodiment apparatus.
As shown in FIG. 1, when the dispersed particle group 2 is irradiated with laser light, a light intensity distribution pattern of diffracted / scattered light is spatially generated. This light intensity distribution pattern changes depending on the size of the particles. Since particles of different sizes are mixed in an actual sample, the light intensity distribution pattern is a combination of diffracted and scattered light from each particle. As a result, the light intensity distribution data (vector) s is m pieces. The detected light amount data of the light detection element is expressed as s (vector) represented by the following equation (2), where (incident light amount) is a vector component (element) s _i (i = 1, 2,..., M). .

On the other hand, the particle size distribution data (vector) q is a particle size range [x _j , x _{j + 1} ] (j = 1, 2) obtained by dividing the particle size range (maximum particle size X ₁ , minimum particle size X _{n + 1} ) to be measured into n parts. ,..., N) are expressed by the following formula (3) in which the particle amount data of each particle diameter section is a vector component (element) q _j (j = 1, 2,..., N). This is expressed as q (vector). When the particle size distribution is the frequency distribution%, normalization is performed so that (q ₁ + q ₂ +... + Q _j +... + Q _n ) = 1 (100%).

The light intensity distribution data (vector) s and the particle size distribution data (vector) q are in a relationship represented by the following expression (4) through a coefficient matrix (matrix) A related to the refractive index of the particles. Therefore, the inverse operation algorithm for inversely transforming the particle size distribution data (vector) q into the light intensity distribution data (vector) s is obtained by multiplying the coefficient matrix A and the particle size distribution data (vector) q as shown in the equation (4). It becomes.

The coefficient matrix A is a matrix for converting the particle size distribution data (vector) q into light intensity distribution data (vector) s as shown by the following equation (4). The physical meaning of the component (element) a _ij (i = 1, 2,..., M, j = 1, 2,..., N) in the particle diameter interval [x _j , x _{j + 1} ] This is the amount of light incident on the i-th photodetecting element of light diffracted / scattered by the particle group of the unit particle amount to which it belongs. The numerical value of a _ij can be theoretically calculated in advance. For this, the Fraunhofer diffraction theory is applied when the particle diameter is sufficiently larger than the wavelength of the laser beam irradiated. On the other hand, the Mie scattering theory is applied in a sub-micron region where the particle diameter is approximately the same as the wavelength of the laser beam irradiated. The Fraunhofer diffraction theory is considered to be an effective approximation when it is an approximation in a specific case of the Mie scattering theory, that is, in the forward minute angle scattering, when the particle diameter is sufficiently larger than the wavelength of the laser beam irradiated.

In order to obtain the element a _ij of the coefficient matrix A according to the above theory, the refractive index of the particle and the medium for dispersing the particle is input (set) from the operation unit 16. In this case, the set refractive index is generally expressed as a complex number. In the embodiment apparatus, the CPU 14 calculates a coefficient matrix A corresponding to the refractive index input from the operation unit 16. That is, when the refractive index is changed, a different coefficient matrix A is obtained (the coefficient matrix A as a calculation condition is changed).

  On the other hand, the solution of the particle size distribution data (vector) q obtained by the least square method based on the above equation (4) is as shown in the following equation (5). This equation (5) is an arithmetic algorithm for converting the light intensity distribution data (vector) s into the particle size distribution data (vector) q. Of course, the calculation algorithm and the inverse calculation algorithm are just examples, and various variations are possible.

Here, ^AT is a transposed matrix of A, and () ^-1 indicates an inverse matrix.
Subsequently, a process of calculating an intersection angle index between two light intensity distribution data (vectors) will be described. In this embodiment, the intersection angle index is the cosine of the intersection angle after being multiplied by the weight matrix.

  As shown in the following equations (6) and (7), one of the two light intensity distribution data (vectors) is r and the other is s. Then, the cosine (cos θ) of the intersection angle θ between the light intensity distribution data (vector) (Wr) and (Ws) multiplied by the weight matrix W expressed by the expression (8) is expressed by the following expression (9). . The cos θ in the following equation (9) does not depend on the magnitude of the vector, that is, it is a preferable form in which the dispersion concentration of the particle group does not affect the result.

  In equation (9), (Wr, Ws) is the inner product of (vector) Wr and (vector) Ws. | Wr | and | Ws | are the sizes of Wr and Ws, respectively. That is, | Wr | = √ ((Wr, Wr)), | Ws | = √ ((Ws, Ws)). However, (Wr, Wr) is the inner product of Wr and Wr, and (Ws, Ws) is the inner product of Ws and Ws.

  Next, the measurement operation by the particle size distribution measuring apparatus of the embodiment having the above-described configuration will be described with reference to the flowchart shown in FIG.

  Step S1: Particle groups to be measured are dispersed in a solvent and sent to the sample cell 1, the sample cell 1 is irradiated with laser light, and light intensity distribution data (vector) of diffracted / scattered light by the optical sensors 5a to 5c. Measure r. The measured light intensity distribution data is stored in the light intensity distribution memory 13.

Step S2: A plurality of types of refractive indexes are appropriately set via the operation unit 16 as the refractive indexes of the particles. And CPU14 calculates | requires the coefficient matrix A used for the calculation algorithm which converts light intensity distribution data into particle distribution data first using the 1st refractive index (specifically relative refractive index with a liquid medium). Further, the CPU 14 uses the coefficient matrix A to calculate the light intensity distribution data (vector) measured in step S1 by calculation using an arithmetic algorithm in which (vector) r is substituted for (vector) s in equation (5). r is converted into particle distribution data (vector) q.

Step S3: The CPU 14 inversely converts the particle distribution data (vector) q obtained in step S2 into light intensity distribution data (vector) s by the inverse conversion algorithm expressed by the equation (4).

Step S4: The actually measured light intensity distribution data (vector) r stored in the light intensity distribution memory 13 and the inversely converted light intensity distribution data (vector) s obtained in step 3 are treated as vectors. The cosine (cos θ) as a vector crossing angle index is obtained and stored by the equation (9) incorporating a weight matrix.

Step S5: Repeat the calculation of cos θ for all the set refractive indexes.

Step S6: Switch to another refractive index, return to Step S2, and repeat the process up to Step S5.

Step S7: From the calculated cos θ, the one closest to “1” is selected.

Step S8: On the other hand, when the value of the cosine (cos θ) as the intersection angle index is closest to “1”, that is, when it can be considered that the two light intensity distribution data are the best match, then Are output to the output unit 17 as appropriate. Further, it is determined that the coefficient matrix at that time (in other words, the refractive index) is appropriate as a calculation condition, and the coefficient matrix is held in the coefficient holding unit 15. Hereinafter, when measuring particle distribution data of a group of particles made of the same material, the particle distribution data can be obtained by the calculation algorithm of equation (5) using the coefficient matrix held in the coefficient holding unit 15. it can.

The specific measurement result obtained using the Example apparatus mentioned above is shown below.
Here, as a measurement target, for example, a polystyrene latex particle group having a particle diameter of about 1 μm was used as a sample, and water was used as a liquid medium. The light intensity distribution data at this time is shown in FIG.

  After the measurement of the light intensity distribution data, the refractive index of the particles was set, for example, in the following seven ways, and the refractive index of water was all set to (1.33-0.00i). In each case, the particle size distribution data (vectors) q (1) to q (8) with respect to the light intensity distribution data was obtained by the calculation algorithm of the equation (5). Further, the light intensity distribution data s (1) to s (8) is inversely converted by the inverse calculation algorithm of the equation (4) for each obtained particle size distribution data (vector) q (1) to q (8). . These are shown together in FIG. From this graph, it can be seen that there is a difference in light intensity between the front and rear portions (especially the rear portion in this embodiment) around sensor element number 59. Therefore, as in the above example, the same weight matrix (that is, the formula (1a)) can be adopted. The cosine (cos θ) of the intersection angle θ between the measured light intensity distribution data (vector) r and each of the inversely converted light intensity distribution data (vectors) s (1) to s (8) is calculated according to equation (9). Asked. The results are as follows. For comparison, the cosine of the intersection angle when the weight matrix is not multiplied is also shown. A graph of this is shown in FIG.

Refractive index Unweighted intersection cosine Weighted intersection cosine setting 1 1.40-0.00i 0.995745 0.991471
Setting 2 1.45-0.00i 0.996917 0.993811
Setting 3 1.50-0.00i 0.996903 0.993768
Setting 4 1.55-0.00i 0.998266 0.996547
Setting 5 1.60-0.00i 0.996629 0.993213
Setting 6 1.65-0.00i 0.995967 0.991862
Setting 7 1.70-0.00i 0.994711 0.989153
Setting 8 1.80-0.00i 0.992230 0.984441

  It can be seen that the difference between the appropriate value (maximum intersection cosine value) and the inappropriate value is larger in the case of weighting than in the case of no weighting.

(Other embodiments)
Next, another embodiment of the present invention will be described. In the embodiment described above, the method of determining the degree of difference between the two data using the intersection angle index of the two light intensity distribution data is a plurality of coefficient matrices corresponding to the light intensity distribution data obtained by inverse conversion (that is, , A plurality of refractive indices) is applied to the process of selecting the optimum coefficient matrix (that is, the optimum refractive index). However, the comparison method of the light intensity distribution data of the diffracted / scattered light according to the present invention is, for example, when determining the quality of the overall physical properties such as the particle size distribution of the sample particle group, the surface state of the particle and the shape of the particle. It can also be applied to. As described above, the light intensity distribution data of diffracted / scattered light itself contains comprehensive information on the physical properties such as the surface state and particle shape of the sample particle group to be measured, in addition to the particle size distribution information. Therefore, the light intensity distribution data of the sample particle group is compared with the light intensity distribution data of the non-defective particle group obtained in advance to determine the degree of difference between the two, thereby determining the quality of the sample particle group. be able to.

In such an apparatus for determining the quality of the sample particle group, a reference light intensity distribution memory 18 as shown in FIG. 1 is provided. In the reference light intensity distribution memory 18, reference light intensity distribution data (in this example, light intensity distribution data of a good particle group) is stored in advance. Hereinafter, the operation of the apparatus for determining the quality of the sample particle group will be described with reference to the flowchart shown in FIG.

First, the light intensity distribution data of the sample particle group is measured (step S11), and a cosine (cos θ) as an intersection angle index between the light intensity distribution data and the reference light intensity distribution data is calculated (step S12). For calculating the cosine, equation (9) incorporating a weight matrix is used. It is determined whether or not the obtained intersection angle index cos θ is larger than a predetermined value (step S13). This predetermined value is the extent to which the light intensity distribution data of the sample particle group matches the light intensity distribution data of the reference non-defective particle group to determine that the sample particle group is non-defective The value is determined experimentally in advance. When the crossing angle index cos θ is larger than the predetermined value, the sample particle group has a particle size distribution and the overall physical properties such as the surface state and particle shape thereof are close to the non-defective particle group. It determines with a non-defective product (step S14). On the other hand, when the intersection angle index cos θ is smaller than the predetermined value, the overall physical properties of the sample particle group are considerably separated from those of the non-defective particle group. S15). According to the embodiment apparatus as described above, accurate quality control can be performed in the manufacturing process of medicines and foods that handle powder.

The present invention is not limited to the above embodiment, and can be modified as follows.
(1) The method for comparing the light intensity distribution data of diffracted / scattered light according to the present invention may be applied to various uses, such as sample particle group ranking, in addition to particle size distribution measurement and particle group non-defective determination. Can do.

(2) In the above embodiment, the crossing angle index corresponding to the crossing angle of the two light intensity distribution data is the cosine of the crossing angle, but the crossing angle index may be the sine of the crossing angle (sin θ) or the crossing angle itself. Good. In the case of sine, both light intensity distribution data are completely coincident when “0”, and both light intensity distribution data are completely unmatched when “1”. In the case of the intersection angle itself, both light intensity distribution data are completely coincident at “0 °”, and both light intensity distribution data are completely unmatched at “90 °”.

  The present invention can be used in a particle size distribution measuring apparatus.

It is a block diagram which shows the whole structure of the particle size distribution measuring apparatus of an Example. It is an example of the light intensity distribution with a slight difference. It is the cosine of the intersection angle of the light intensity distribution with a slight difference. It is an element of the weight matrix. It is light intensity distribution data of a polystyrene latex particle group having a particle diameter of about 1 μm. It is the light intensity distribution data obtained by inverse conversion from the particle size distribution obtained by changing the refractive index in various ways from the light intensity distribution data of the polystyrene latex particle group having a particle diameter of about 1 μm. It is a graph which shows the cosine of the crossing angle based on the light intensity distribution data of FIG. It is a flowchart of the particle size distribution measurement by an Example apparatus.

Explanation of symbols

2 ... Dispersed particle group 4 ... Laser light sources 5a to 5c ... Optical sensor 13 ... Light intensity distribution memory 14 ... CPU
15 ... Coefficient holding unit 17 ... Reference light intensity distribution memory

Claims (4)

  The process of irradiating the dispersed particle group with laser light is different from the process of detecting diffracted / scattered light from the dispersed particle group with a plurality of light detection elements and obtaining light intensity distribution data using each detected light quantity data as a component. And comparing the light intensity distribution data obtained for each of the two dispersed particle groups, the comparison process treats each light intensity distribution data as a vector having each detected light quantity data as a component, A process of calculating an intersection angle index related to the intersection angle formed by these two light intensity distribution data by emphasizing the weight matrix by applying a weight matrix, and the two light intensity distribution data based on the intersection angle index obtained in the calculation process. A method for comparing light intensity distribution data of diffracted / scattered light, comprising a step of determining a degree of difference. 2. The diffracted / scattered light according to claim 1, wherein a weight matrix in which any one of the following formulas (1a) to (1c) is a diagonal component and a value other than the diagonal component is 0 is used. Comparison method of light intensity distribution data.
  Light irradiating means for irradiating the dispersed particle group with laser light, and light detection for detecting diffracted / scattered light from the dispersed particle group with a plurality of light detecting elements and obtaining light intensity distribution data using each detected light quantity data as a component Means and the measured light intensity distribution data are converted into particle size distribution data corresponding to each coefficient matrix by an arithmetic algorithm using a plurality of types of coefficient matrices related to the refractive index of the particles constituting the dispersed particle group. Light intensity distribution inverse calculation that reversely converts the particle size distribution data for each coefficient matrix into light intensity distribution data corresponding to each coefficient matrix by means of an inverse operation algorithm using the coefficient matrix. Each of the measured light intensity distribution data and the inversely converted light intensity distribution data for each coefficient matrix, and from among the plurality of inversely converted light intensity distribution data, The light intensity distribution is searched for the inverse converted light intensity distribution data having a high degree of coincidence with the actually measured light intensity distribution data, and the coefficient matrix corresponding to the inverse converted light intensity distribution data is selected as the optimum calculation condition. In the particle size distribution measuring apparatus, comprising: a comparison unit; and a particle size distribution determination unit that determines the particle size distribution data obtained by using the selected coefficient matrix as an appropriate particle size distribution of the dispersed particle group. The comparison unit treats the actually measured light intensity distribution data and the inversely converted light intensity distribution data as vectors having respective detected light amount data as components, and is inversely converted to the actually measured light intensity distribution data. Emphasis is calculated by emphasizing the intersection angle index related to the intersection angle formed by the light intensity distribution data for each coefficient matrix by applying a weight matrix. By comparing the angle index calculation means and the intersection angle index corresponding to each of the calculated coefficient matrices, the light intensity distribution data of the inverse conversion having a high degree of coincidence with the actually measured light intensity distribution data is searched for, A particle size distribution measuring apparatus comprising selection means for selecting a coefficient matrix corresponding to the inversely converted light intensity distribution data as an optimum calculation condition. 4. The particle size distribution measuring apparatus according to claim 3, wherein a weight matrix in which any one of the following formulas (1a) to (1c) is a diagonal component and a value other than the diagonal component is 0 is used. .