RU2778816C1 - Method and apparatus for measuring the fraction of fine particles - Google Patents

Abstract

FIELD: measuring technology.

SUBSTANCE: method for measuring the fraction of fine particles is proposed, which includes: step S1 of measuring the distance between the distance measuring device and pieces of material; step S2 of calculating the characteristic value of the feature from the distance data obtained at step S1; and step S3 of converting the characteristic value of the feature calculated at step S2 to the value of the fraction of fine particles. The characteristic value of the feature calculated at step S2 represents the distance fluctuation calculated from the distance data obtained at step S1. A higher proportion of fine particles on the pieces of material means a greater microscopic distance fluctuation caused by microscopic irregularity on the surface of the pieces of material in the direction of height in three-dimensional form.

EFFECT: fraction of fine particles on pieces of material can be measured in real time with high accuracy.

10 cl, 9 dwg

Description

Scope of technical application

The present invention relates to a fine particle proportion measurement solution for measuring the proportion of fine particles adhering to the surface of lumps of material, such as ore and coke, which are raw materials for blast furnaces.

State of the art

In manufacturing processes that use raw materials such as minerals, the particle size distribution of the raw materials affects the course of the manufacturing processes. Thus, a preliminary measurement of the particle size distribution of the raw materials is necessary. Particularly in the operation of a blast furnace, it is important to know the particle size distribution of the raw materials, such as ore and coke, to ensure gas penetration into the furnace. It is also necessary to closely monitor the operation of the blast furnace so that the proportion of fine particles (eg, fine particles with a particle size of 5 mm or less) adhere to larger particles (pieces) of raw materials.

In order to know the particle size distribution of the raw materials during normal operation of the blast furnace, the particle sizes were measured by periodic sampling and screening of the raw materials. However, this method takes time to perform the analyses. Patent Document 1 discloses a technical solution in which raw material samples are automatically taken to increase the frequency of analysis. However, if sampling is performed too frequently, the workflow can be slowed down. Also, when sampling performed for the purpose of verification, they face the problem of representativeness.

In contrast to the conventional solution mentioned above, in which particle sizes were measured by sampling and sifting raw materials, solutions have been proposed in which the particle sizes of raw materials during transportation were measured in real time using, for example, a camera.

For example, Patent Document 2 discloses a solution for determining the particle size of a bulk raw material conveyed on a conveyor. The technical solution includes obtaining an image of the bulk raw material on a conveyor to create image data, determining the brightness distribution from the data, and determining the particle size of the bulk raw material using the maximum height of the peaks of the brightness distribution curve.

Patent Document 3 discloses a solution in which the water content of a material to be charged into a blast furnace is determined from spectroscopic information obtained from near-infrared reflected light included in the light reflected from the material to be charged. In this technical solution, the proportion of fine particles of the material to be loaded is determined in real time based on the relationship between the water content of the material to be loaded and the proportion of fine particles adhering to the material to be loaded.

Patent Document 4 discloses a technical solution in which the average brightness of image data obtained by capturing an image of pieces of material is determined as a feature characteristic value. In this technical solution, the proportion of small particles adhering to the surface of the pieces of material is determined by the characteristic value of the feature described above.

List of documents cited

Patent documents (PD)

PD 1: Unexamined Japanese Patent Application No. 2005-134301

PD 2: Unexamined Japanese Patent Application No. 2000-329683

PD 3: Unexamined Japanese Patent Application No. 2015-124436

PD 4: International Publication No. 2018/101287

Brief description of the invention

Technical problem

When using the conventional technical solutions described above, the following problems are encountered.

Due to the limitations of the resolution of the camera, the solution disclosed in Patent Document 2 cannot fully ensure the measurement accuracy of bulk raw material containing fine particles adhering to it.

With the technical solution disclosed in Patent Document 3, it is not possible to achieve a sufficient level of measurement accuracy because the relationship between the water content of the material to be loaded and the proportion of fine particles in the material to be loaded is not necessarily high.

Like the solution disclosed in Patent Document 3, it is not possible to achieve a sufficient level of measurement accuracy with the solution disclosed in Patent Document 4 because the relationship between the image brightness achieved by the camera and the proportion of fine particles is not necessarily high.

Accordingly, the purpose of the present invention is to solve the problems that arise when using the conventional technical solutions described above. The present invention provides a fine particle proportion measurement method and a fine particle proportion measurement apparatus, which can measure the proportion of fine particles adhering to the surface of material pieces in real time with high accuracy.

Solution

The inventors of the present invention have made various studies to develop a technical solution for measuring the proportion of fine particles, which can solve the problems described above. The inventors of the present invention have found that the proportion of fine particles adhering to the surface of pieces of material can be measured in real time with high accuracy by means of the following technical solution. The technical solution includes: measuring the distance to pieces of material (the object of measurement is the proportion of small particles) using a rangefinder; determining, from the resulting distance data, a feature characteristic value that correlates with the proportion of fine particles; and converting the characteristic feature value to a fine particle fraction value.

The present invention, based on the discoveries described above, is summarized as follows.

1. The method for measuring the proportion of fine particles includes: step S1 measuring the distance x between the device for measuring the distance and the pieces of material; a step S2 of calculating a feature characteristic value from the distance data acquired in step S1; and a step (S3) of converting the feature characteristic value calculated in step S2 into a fine particle proportion value.

2. According to the method for measuring the proportion of fine particles according to claim 1, the feature characteristic value calculated in step S2 represents the fluctuation of the distance calculated from the distance data obtained in step S1.

3. According to the method for measuring the proportion of fine particles according to claim 2, the characteristic feature value calculated in step S2 is based on the standard deviations in the standard deviation matrix obtained by applying the standard deviation filter to the distance data obtained in step S1.

4. According to the method for measuring the proportion of fine particles according to claim 3, the feature characteristic value calculated in step S2 is the modal standard deviation in the standard deviation matrix obtained by applying the standard deviation filter to the distance data obtained in step S1.

5. According to the method of measuring the proportion of fine particles according to 3 or 4, the filtering range of the standard deviation filter is 10 × 10 pixels or less.

6. According to the method for measuring the proportion of fine particles according to any one of paragraphs. 1-5, in step S1, the distance to the material pieces x is measured with a distance measuring device mounted above the pieces of material x.

7. Apparatus for measuring the proportion of fine particles contains: a device 1 for measuring distance, configured to measure the distance to the pieces of material x; and a computing device 2 containing computing means 2a and converting means 2b. Computing means 2a are configured to calculate the characteristic value of the feature from the distance data obtained by the distance measuring device 1; and the converting means 2b are configured to convert the feature characteristic value calculated by the computing means 2a into a fine particle fraction value.

8. In the apparatus according to claim 7 for measuring the proportion of fine particles, the characteristic characteristic value representing the distance fluctuation is calculated by the computing means 2a from the distance data obtained by the distance measuring device 1.

9. In the apparatus of claim 8 for measuring the proportion of fine particles, the calculation means 2a apply a standard deviation filter on the distance data obtained by the distance measuring device 1, and calculate the characteristic value of the feature based on the standard deviations of the resulting matrix standard deviations.

10. In the apparatus of claim 9 for measuring the proportion of fine particles, the calculation means 2a apply a standard deviation filter to the distance data obtained by the distance measuring device 1, and calculate the modal standard deviation from the resulting standard deviation matrix as characteristic value of the feature.

11. In the apparatus of claim 9 or 10 for measuring the proportion of fine particles, the filtering range of the standard deviation filter is 10 × 10 pixels or less.

12. In the apparatus according to any one of paragraphs. 7-11 for measuring the proportion of fine particles, the distance measuring device 1 is installed above the pieces of material, and with it, the distance to the pieces of material x is measured.

Beneficial Effects of the Invention

Through the use of the present invention, the proportion of fine particles adhering to the surface of the pieces of material can be measured in real time with high accuracy. In this way, it is possible, for example, to know exactly the fines fraction of the coke fed to the blast furnace (ie the fraction of coke fines) before charging the coke into the blast furnace. This contributes to the stable operation of the blast furnace.

Brief description of the drawings

In FIG. 1 is an explanatory diagram of an application of the present invention for measuring the proportion of fines in coke before it is loaded into a blast furnace;

in fig. 2 shows an exemplary 3D coke shape image based on the measured distance obtained by the distance measuring device in the embodiment shown in FIG. one;

in fig. 3 is an explanatory diagram of an exemplary computational solution used in computing a characteristic value of a feature from distance data obtained by a distance measurement device, where a standard deviation filter is applied to the distance data to determine a matrix of standard deviations of the distance data according to the present invention;

in fig. 4 is a plot of the distribution of standard deviations in a matrix of standard deviations determined by the calculations presented in FIG. 3;

in fig. 5 shows a black and white two-tone image obtained with a threshold value of 0.5, which is the standard deviation in the output matrix of 800 × 1000 standard deviations, when applying a standard deviation filter with a 3 × 3 matrix filtering range (3 × 3 pixels), on data about distance of the image shown in Fig. 2;

in fig. 6 is a schematic diagram of an overlay image of coke particles (lumps);

in fig. 7 is a graph of the relationship between the characteristic value of a feature determined according to the present invention and the known proportion of fine particles, where the standard deviation in the standard deviation matrix is used as the characteristic value of the feature;

in fig. 8 is a graph showing the relationship between a characteristic value of a feature determined according to the present invention and a known proportion of fine particles, where the modal standard deviation in the standard deviation matrix is used as the characteristic value of the feature;

in fig. 9 is a block diagram of an embodiment of the present invention.

Description of Embodiments

The method for measuring the proportion of fine particles according to the present invention is a method for measuring the proportion of fine particles on the pieces of material x, based on the amount of fine particles adhering to the surface of the pieces of material x. The method for measuring the proportion of fine particles includes: step (S1) measuring the distance to the pieces of material x using a distance measuring device; step (S2) calculating a feature characteristic value from the distance data obtained in step (S1); and a step (S3) of converting the feature characteristic value calculated in step (S2) into a fine particle proportion value.

The apparatus for measuring the proportion of fine particles according to the present invention is used to carry out the method for measuring the proportion of fine particles described above. Apparatus for measuring the proportion of fine particles contains: a device 1 for measuring distance, configured to measure the distance to the pieces of material x; and a computing device 2 containing computing means 2a and converting means 2b. The computing means 2a calculates the characteristic value of the feature from the distance data based on the distance obtained by the distance measuring device 1. The conversion means 2b converts the characteristic value of the feature calculated by the computing means 2a to the value of the proportion of fine particles.

Examples of pieces of material x against which to measure the proportion of fine particles according to the present invention include, but are not limited to, pieces of raw materials (eg, ore and coke) used in metal smelting processes, such as iron making processes.

Here, the proportion of fine particles is defined as the proportion of the mass of fine particles with a predetermined size or less of the total mass of the pieces of material x containing fine particles adhering to their surface.

An embodiment of the present invention is further described using an example in which the pieces of material x (fine fraction measurement object) are pieces of coke before charging into a blast furnace.

In FIG. 1 shows an embodiment of the present invention and specifically an example in which the present invention is applied to measure the proportion of fine particles in coke before it is loaded into a blast furnace. In the drawing, position number 3 indicates a feeder, position number 4 indicates a sieve, position number 5 indicates a conveyor and position number _xc indicates coke, which is the object of measurement of the proportion of fine particles. Also, numerals 1 and 2 denote a distance measuring device and a computer device, respectively, which constitute the fine particle fraction measuring apparatus of the present invention. The computer device 2 comprises: a computing means 2a by which a feature characteristic value is calculated from the distance data obtained by the distance measuring device 1; and converting means 2b by which the feature characteristic value calculated by the computing means 2a is converted to a fine particle proportion value.

The coke x _c loaded into the blast furnace is stored in the feeder 3. The coke x _c discharged from the feeder 3 is screened using a sieve 4 to remove fine dust, transferred to the conveyor 5 and transported on the conveyor 5 to the blast furnace (or to the feeder above the oven). The coke x _c conveyed by the conveyor 5 contains coke particles (chunks) on the sieve 4 and adhering fine particles (coke fines) adhering to the coke particles that do not pass through the sieve 4.

In this embodiment, the coke <constituent parts> x _c transported by the conveyor 5 is measured as follows to determine the proportion of fine particles, mainly consisting of adhering fine particles described above.

First, with the distance measuring device 1 mounted above the conveyor 5, the distance to the coke x _c on the conveyor 5 is measured, and data determining the three-dimensional shape of the coke x _c is taken (step S1). For example, a two-dimensional laser range finder can be used as the device 1 for measuring distance.

When using the laser rangefinder described above, the distance to the coke x _c is measured line by line by scanning the coke x _{c with} a laser beam across the width of the conveyor. Each line corresponds to the width of the laser beam, which irradiates the coke x _c across the width of the conveyor. Since the coke x _c is transported and moved on the conveyor 5, the three-dimensional shape of the coke x _c is obtained by linear measurement of the distance to the coke x _c by the laser range finder in regular measurement cycles, and then sequentially connect the measured values in a line. The technical solution described above is one in which the three-dimensional shape of the object to be measured is captured by the so-called light section method. To implement the technical solution, a conventional laser rangefinder and conventional data processing tools can be used.

The laser range finder is preferably one that covers a measurement area corresponding to the width of the conveyor and is suitable for measuring the entire (or entire surface) of the coke x _c transported by the conveyor 5. Shorter measurement cycles, in particular 1 kHz or more, are preferred. . In this embodiment, the measurement cycles are set to 4 kHz.

The distance measuring device 1 need not be a 2D laser range finder. For example, a distance measuring device based on the stereoscopic principle using two cameras can be used.

In FIG. 2 shows an exemplary image of the 3D shape of the coke x _c (ie the image of the coke x _c on the conveyor 5, top view) based on the distance obtained by the 2D laser rangefinder. In the grayscale image shown in FIG. 2, the lighter areas are taller and closer to the rangefinder. The image contains 800 pixels in width (along the width of the conveyor) and 1000 pixels in length (in the direction of movement of the conveyor). Each image pixel corresponds to 2 mm in width and 4 mm in length. The distance is determined for each pixel.

This means that the image shown in Fig. 2 is obtained by connecting 1000 lines of distance data (hereinafter also referred to as "distance data") across the width of the pipeline, where each line contains 800 pixels. Thus, distance data consisting of 800 pixels × 1000 pixels is obtained. The resolution in the height direction is 5 µm.

In conventional image processing performed using 3D shape data, the 3D shape data of coke including irregularities is generally subjected to a processing signal that separates the coke into individual particles. The number of coke particles produced by the impact causing particle separation is counted for each size range to create a histogram and calculate the particle size distribution.

For example, when the distance measuring device 1 has a minimum resolution of 4 mm × 2 mm in length and width, and a resolution of 5 μm in the height direction, because the coke charged into the blast furnace usually has a particle size of 35 mm or more, the resolutions are sufficient to measure the size of coke particles (lumps) by means of the processing signal. However, adhering fine particles (coke fines) adhering to the surface of the coke particles (lumps) often contain particles having a size of 1 mm or less. If such adhering fine particles are spherical, then the resolution of the distance measuring device 1 is insufficient in length and width, although it is sufficient in the height direction. Accordingly, unlike the case of counting coke particles (lumps), it is difficult to count the number of adhering fine particles and to determine the proportion of fine particles.

If the distance measuring device is brought as close as possible to the coke to narrow the width of the measurement range of the distance measuring device and increase the length and width resolution, it is possible to capture the adhering fine particles on the surface of the coke and measure the proportion of fine particles. However, when the distance measuring device is brought closer to the coke, adhering fine particles on only a portion of the coke may be covered. Thus, it is not possible to determine the proportion of fines in the total coke transported on the conveyor (ie the proportion of fines across the width of the conveyor). Installing a plurality of distance measuring devices may be possible. However, this makes data processing complex, adds installation constraints, and increases installation cost.

According to the present invention, where the entire width of the conveyor is in the measurement area, the proportion of fine particles can be measured with high accuracy even when the resolution in length and width is not high enough. This means that the present invention is made to measure the coke distance x _c determined from the resulting distance data as a characteristic value of a feature representing microscopic distance fluctuations, and convert the characteristic value of the feature to a fine particle fraction value so that fraction of fine particles adhering to the coke surface x _c .

According to the present invention, a characteristic value of a feature representing a microscopic distance fluctuation is calculated from the distance data (according to the three-dimensional coke shape data) measured by the distance measuring device 1 as described above (step S2). The resolution of the distance data obtained by the distance measuring device 1 is not high enough in length and width to measure the proportion of fine particles, but is high enough in the height direction even when the entire width of the conveyor is in the measurement area.

The inventors of the present invention have found that if the number of fine particles adhering to the surface of the coke increases (i.e., if the proportion of fine particles increases), then the microscopic irregularity of the coke surface in the height direction in the three-dimensional form increases (i.e., microscopic fluctuations in the distance coke increase in the direction of height). Accordingly, according to the present invention, the microscopic distance fluctuation is used as the characteristic value of the feature, and the proportion of fine particles is determined from the characteristic value of the feature.

In this embodiment, the microscopic distance fluctuation is considered as a local height fluctuation at a given location, a standard deviation filter (i.e., a filter for calculating standard deviations) having a small range for filtering the distance data is applied, and a characteristic feature value is obtained from the resulting standard deviations. deviations. The more microscopic irregularities in the surface of the coke resulting from a high proportion of small particles, the more standard deviations are obtained as a result of applying the standard deviation filter. The characteristic value of a feature is thus obtained from the standard deviations.

First, the captured distance data is divided into pixels. For example, the distance data shown in FIG. 2 are divided by 800 × 1000 pixels. Next, an 800×1000 matrix is created by setting the appropriate distance for each pixel. Then, a commonly used standard deviation filter having a filtering range of a 3 × 3 matrix (corresponding to 3 × 3 pixels) is superimposed on the distance data of an 800 × 1000 matrix to determine the standard deviation for each 3 × 3 matrix. deviations.

In FIG. 3 illustrates an exemplary calculation methodology (calculation image) involving applying a standard deviation filter to distance data. In FIG. 3 position number 6 denotes the standard deviation filter; position number 7 denotes the 4 × 4 matrix of output standard deviations. In this example, a 6 standard deviation filter with a 3 x 3 matrix filtering range is overlaid on the distance data as a 4 x 4 matrix (corresponding to 4 x 4 pixels). In FIG. 3(A) shows an overlay image of a 6 standard deviation filter as a 3×3 matrix on a 4×4 matrix of input data. In FIG. 3(B) shows an image of storing the computed values in a 4×4 matrix of output standard deviations. The output standard deviation matrix shown in FIG. 3(B) is created from the input data matrix shown in FIG. 3(A). It should be noted that the numbers shown in Fig. 3(A) and FIG. 3(B) are examples provided for the purpose of explanation.

The procedure is described in detail below.

First, a 4×4 output matrix 7 is created and 0s (zeros) are stored in the entire output matrix 7. As shown in FIG. 3(A), the standard deviation of all 9 pixels in the target filtering range is calculated and the calculated standard deviation is output to the position in the output matrix 7 corresponding to the center position in the filtering range. The 6 standard deviation filter is then moved to the next range where the same process is repeated. In the example in FIG. 3, the filter calculation process described above is performed a total of four times for one 4×4 matrix of input data.

In the case of the distance data shown in FIG. 2, these calculations are superimposed on the distance data of an 800×1000 matrix such that the standard deviation matrix for the 800×1000 distance data becomes the output. The characteristic value of a feature is determined using an 800 x 1000 standard deviation matrix.

In order to capture microscopic fluctuations, the standard deviation filter 6 should preferably have a small filtering range (number of pixels). The filtering range is preferably 10×10 pixels or less (or a total of 100 pixels or less), and more preferably 5×5 pixels or less (or a total of 25 pixels or less). The filtering range is particularly preferably 3×3 pixels (or a total of 9 pixels) as described above, and most preferably 2×2 pixels. This is required because with a small filtration range it is possible to reduce the influence of, for example, the shape of the coke or the inclination of the measured surface of the coke relative to the distance measuring device.

In FIG. 4 is a plot of the distribution of standard deviations in a matrix of standard deviations obtained by applying the procedure described above to the distance data shown in FIG. 2. The top of the standard deviations, indicated in the drawing by position number 8, represents the modal standard deviation. In this example, the standard deviations in the standard deviation matrix are divided into a sufficient number of classes (50,000 classes in this example (at intervals of 0.00002)), and the most frequently occurring standard deviation in the classes is defined as the modal standard deviation. It should be noted that this eliminates the frequency of the standard deviation equal to 0 (zero).

In FIG. Figure 5 shows a black and white duplex image obtained with a threshold value of 0.5, which is the standard deviation in the output matrix of 800 × 1000 standard deviations when applying a standard deviation filter with a 3 × 3 matrix filtering range (3 × 3 pixels) on distance data at fig. 2. The white part of the two-tone image corresponds to a standard deviation of 0.5 or more, and the black part of the two-tone image corresponds to a standard deviation of less than 0.5. When comparing the images in Fig. 5 and FIG. 2 discloses that a plurality of matrix elements having a standard deviation of 0.5 or more are located on steps or boundaries between superimposed coke pieces (indicated by numeral 9 in FIG. 5).

In FIG. 6 shows a schematic diagram of the overlay of particles (pieces) of coke. In the drawing, the position number 10 indicates the imposition of pieces of coke. Position number 11 indicates a standard deviation filter applied to the surface of the coke, and position number 12 indicates a standard deviation filter applied to the step or boundary between the superimposed pieces of coke. As described above, in the standard deviation matrix obtained from the distance data, there are many values calculated by applying the standard deviation filter to steps or boundaries between superimposed coke pieces, and the standard deviations are very large.

According to the present invention, standard deviations representing microscopic fluctuations caused by the presence of small particles adhering to the surface of the coke should be included in the characteristic value of the feature. Accordingly, it is preferable to exclude the standard deviations caused by the presence of steps or boundaries between the superimposed pieces of coke. The mean standard deviation in a standard deviation matrix (eg, an 800 x 1000 standard deviation matrix) includes discontinuities caused by the presence of steps or boundaries between superimposed coke chunks. On the other hand, modal standard deviation 8 in the matrix of standard deviations shown in FIG. 4 is very little affected by the presence of steps or boundaries between superimposed coke chunks if the coke particles (chunks) are large enough. This means that although the mean standard deviation in the standard deviation matrix can be used as the characteristic value of the feature, it is preferable that the modal standard deviation 8 in the standard deviation matrix shown in FIG. 4 was used as the characteristic value of the feature.

In this embodiment, a method in which a standard deviation filter is used is used to calculate a feature characteristic value representing a microscopic distance fluctuation. However, the method is not limited to this. For example, a technique for using the high frequency components of a 2D Fourier transform can be applied by considering 800×1000 matrices as an image.

Next, the characteristic feature value obtained as described above is converted to the value of the proportion of fine particles (step S3). This conversion is performed using a coefficient (relational expression) predetermined based on the relationship between the characteristic value of the feature and the known proportion of fine particles.

In FIG. 7 and 8 (each) show the relationship between the characteristic value of the feature and the known proportion of fine particles. The mean standard deviation in the matrix of standard deviations is used as the characteristic value of the feature in FIG. 7, and the modal standard deviation in the standard deviation matrix is used as the characteristic value of the feature in FIG. 8. In FIG. 7 and 8 show the results obtained by using coke lumps as the material pieces described above. The coke pieces have a particle size of 35 mm or more, and contain small particles of coke with a particle size of 1 mm or less adhering to their surface. In FIG. 7 and 8, the vertical axis represents the proportion of small particles (mass %) of coke of 1 mm or less adhering to the coke lumps, and the horizontal axis represents the characteristic value of the feature representing the microscopic fluctuations calculated from the distance data obtained by measuring the lumps. coke containing small coke particles of 1 mm or less adhering to them.

The coke pieces with a particle size of 35 mm or more used in the tests were prepared by sieving the coke using a sieve with a mesh size of 35 mm. One hundred or more pieces of coke were used. The proportion of fine coke particles of 1 mm or less adhering to the surface of the coke pieces was calculated as follows. First, the coke lumps were dried at 120°C to 200°C for 4 hours or more to a constant weight, and then sieved using a sieve with a mesh size of 1 mm. The proportion of fine particles described above was calculated as a fraction (or percentage) of the difference in the weight of the pieces of coke before and after screening, referred to the weight before screening. When using this method, the fact that adhering fine particles fall off in the dry state was taken into account.

In FIG. 7 and 8 show the net relationship between the characteristic value of the feature and the proportion of fine particles. In the case of determining the mean value of the standard deviation in the matrix of standard deviations shown in FIG. 7, however, the correlation coefficient is R = 0.60 and the measurement error (rms error) is σ = 0.23, whereas in the case of the modal standard deviation in the standard deviation matrix shown in FIG. 8, the correlation coefficient is R = 0.70, and the measurement error σ = 0.20. The modal standard deviation in the standard deviation matrix thus shows the best relationship. Thus, when there are multiple steps and boundaries between superimposed coke chunks, it is preferable to use the modal standard deviation in the standard deviation matrix. The slope "a" and y-gap "b" obtained by linear recording of the characteristic value of the feature and the known proportion of fine particles are a = 4.89 and b = -2.16 in FIG. 7 and a=35.6 and b=-1.6 in FIG. 8. Although linear registration is performed here, multiple linear registration or non-linear registration can be performed.

According to the present invention, a coefficient (relational expression) is pre-determined based on the relationship between a characteristic value of a feature (preferably modal standard deviation in a standard deviation matrix) and a known proportion of fine particles, such as those shown in FIG. 7 and 8. The feature characteristic value (preferably the modal standard deviation in the standard deviation matrix) is then converted to a fine particle fraction value using a factor. The proportion of fine particles in the coke is thus measured with a high measurement accuracy (for example, with a measurement error σ of 0.3 or less).

In FIG. 9 is a block diagram of the embodiment of the present invention described above. In this embodiment, as described above, the ratio (relational expression) of converting the characteristic value of the feature to the value of the proportion of fine particles is predetermined based on the relationship between the characteristic value of the feature and the known proportion of fine particles.

The distance measuring device 1 measures the distance to the pieces of material x (fine particle proportion measurement object) such as coke x _c conveyed on the conveyor to capture the distance data (3D shape data) (step S1). Next, by means of computing means 2a of the computing device 2, a standard deviation filter is applied to the distance data to determine standard deviations; calculating the modal standard deviation in the standard deviation matrix; and using the modal standard deviation as the characteristic value of the feature (step S2). Next, the conversion means 2b of the computer device 2 converts the characteristic value of the feature to the value of the proportion of fine particles, using a known coefficient (relational expression) to determine the proportion of fine particles adhering to the pieces of material x (step S3).

Step S1 is then performed at predetermined measurement cycles. Each time distance data is obtained in step S1; step S2 and step S3 are performed to determine the proportion of fine particles.

The fraction of fine particles on the pieces of material x is thus measured in real time with high precision.

In an embodiment, the present invention is used to measure the fraction of fines in coke x _c before it is loaded into a blast furnace. The present invention is applicable not only to this, but also to the measurement of the proportion of fine particles in pieces of material x of various kinds.

The present invention relates to a method suitable for measuring, in real time, the proportion of fine particles on pieces of material x (eg coke or ore) during transportation, for example, on a conveyor. The present invention is also applicable to measuring the proportion of fine particles on pieces of material x in a stationary state. This is done by moving the distance measuring device 1 to measure the three-dimensional shape of the pieces of material x in a stationary state.

Although the adhering fine particles are described as having a size of 1 mm or less, their size is not limited to this value, and it can be appropriately determined. For example, even if the adhering fine particles have a size of 2 mm or less (not 1 mm or less), the proportion of fine particles can be measured by first determining the relationship between the characteristic value of the feature and the proportion of fine particles.

Explanation of item numbers indicated in the drawings

1 - Device for measuring distance

2 - Computer device

2a - Computing tools

2b - Transformers

3 - Feeder

4 - Sieve

5 - Conveyor

6 - Standard deviation filter

7 - Output matrix

8 - Modal standard deviation

9 - Step or border between superimposed pieces of coke

10 - Superimposed pieces of coke

11 - Schematic diagram of the standard deviation filter applied to the surface of the coke

12 is a schematic diagram of a standard deviation filter applied to a step or boundary between coke chunks.

Claims (19)

1. A method for measuring the proportion of fine particles, including: step (S1) measuring the distance between the distance measuring device and the pieces of material (x); step (S2) calculating a feature characteristic value from the distance data obtained in step (S1); and a step (S3) of converting the feature characteristic value calculated in step (S2) to a fine particle proportion value in which the distance measuring device is a two-dimensional distance measuring device, and the characteristic feature value calculated in step (S2) represents the distance data in each pixel obtained in step (S1) using a two-dimensional distance measuring device. 2. The fine particle proportion measurement method according to claim 1, wherein the feature characteristic value calculated in step (S2) is based on standard deviations in a standard deviation matrix obtained by applying a standard deviation filter (standard deviation calculation filter) to distance data. in each pixel obtained in step (S1). 3. The fine particle proportion measurement method according to claim 2, wherein the feature characteristic value calculated in step (S2) is the modal standard deviation in the standard deviation matrix obtained by applying the standard deviation filter (standard deviation calculation filter) to the distance data in each pixel obtained in step (S1). 4. The fine particle proportion measurement method according to claim 2 or 3, wherein the filtering range of the standard deviation filter (standard deviation calculation filter) is 10 × 10 pixels or less. 5. The method of measuring the proportion of fine particles according to any one of paragraphs. 1-4, in which, in step (S1), a distance to the pieces of material (x) is measured with a two-dimensional distance measuring device mounted above the pieces of material (x). 6. Apparatus for measuring the proportion of fine particles, containing: a device (1) for measuring distance, configured to measure the distance to pieces of material (x); and a computer device (2) containing computing means (2a) and converting means (2b); where the computing means (2a) is configured to calculate the characteristic value of the feature from the distance data obtained by the device (1) for measuring the distance; where the converting means (2b) are configured to convert the characteristic value of the feature calculated by the computing means (2a) to a fine particle fraction value in which the distance measuring device is a two-dimensional distance measuring device, and the computing means (2a) calculates a characteristic feature value representing the distance data in each pixel obtained using the two-dimensional distance measuring device. 7. The fine particle proportion measurement apparatus according to claim 6, wherein the calculation means (2a) apply a standard deviation filter (standard deviation calculation filter) to the distance data in each pixel obtained using the 2D distance measuring device, and calculating a characteristic feature value based on the standard deviations in the resulting standard deviation matrix. 8. The fine particle proportion measurement apparatus according to claim 7, wherein the calculation means (2a) apply a standard deviation filter (standard deviation calculation filter) to the distance data in each pixel obtained using the 2D distance measuring device, and calculating the modal standard deviation in the resulting matrix of standard deviations as a characteristic feature value. 9. The fine particle proportion measurement apparatus according to claim 7 or 8, wherein the filtering range of the standard deviation filter (standard deviation calculation filter) is 10 × 10 pixels or less. 10. Apparatus for measuring the proportion of fine particles according to any one of paragraphs. 6-9, in which a two-dimensional distance measuring device is mounted above the pieces of material (x) and measures the distance to the pieces of material (x) with it.

Images