JP4941266B2 - Luminescence analyzer - Google Patents

Description

  The present invention relates to an emission analysis apparatus using various discharge methods such as spark discharge, arc discharge, inductively coupled plasma (ICP) discharge, and laser excitation.

  In an emission spectroscopic analyzer, generally, a solid sample, which is a metal or nonmetal, is given energy by arc discharge or spark discharge to evaporate and evaporate the sample, and the emitted light is introduced into a spectrometer. A spectral line having a wavelength peculiar to the element is extracted and detected. In particular, an emission spectroscopic analysis apparatus that uses spark discharge as an excitation source can perform highly accurate analysis. For example, in a production plant such as a steel material or a non-ferrous metal material, a composition analysis is performed in a produced metal body. Widely used for.

  When the emission spectroscopic analysis is performed on the solid sample, if there is a defect in the analysis portion, normal light emission does not occur and accurate analysis cannot be performed. Therefore, in the emission analyzer described in Patent Document 1, the analysis surface of the sample is photographed with a CCD camera, the presence or absence of a defect in the analysis portion is determined based on the photographed image, and the portion having no defect is extracted and analyzed. Like to do. In addition, since there may be a lack of reliability when the analysis is performed at only one location, the analysis is usually performed at a plurality of locations on the same analysis surface and the result is averaged (see, for example, Patent Document 2). ).

  When an emission analyzer is used in the above-mentioned production factory, for example, various samples such as samples with relatively many defects collected during refining and samples that need to be analyzed with high accuracy finally. It is sometimes necessary to change the number and position of analysis points on the analysis surface or the interval (distance) between the analysis points in accordance with the sample. However, in the conventional luminescence analysis apparatus, the analysis location is extracted according to a certain predetermined algorithm, and therefore the optimal analysis location may not always be extracted depending on the sample. In addition, it takes time to extract an unnecessarily large number of analysis points, which may reduce the analysis throughput.

JP 2005-69853 A Utility Model Registration No. 3131999

  The present invention has been made in order to solve the above-described problems, and the object of the present invention is to provide a light emission method capable of performing an emission analysis by extracting an appropriate analysis site in accordance with the type of sample and the purpose of analysis. It is to provide an analysis device.

In order to solve the above problems, the present invention provides an emission analyzer for performing an emission analysis on one or more analysis points on the surface of a solid sample.
a) Image analysis means for capturing an analysis surface of a solid sample and detecting a defective portion based on the captured image;
b) Preliminarily storing a plurality of types of analysis site extraction procedures for determining the position of one or more analysis sites on the analysis surface of the solid sample so as to avoid the defective sites using the defect information obtained by the image analysis means. Storage means to keep,
c) a selection means for selecting an analysis site extraction procedure to be used corresponding to the sample to be analyzed prior to performing the analysis;
d) At the time of emission analysis, an analysis site extraction procedure corresponding to the sample to be analyzed is acquired by the storage unit based on the selection information by the selection unit, and according to the procedure, one or more on the analysis surface of the sample is acquired. An analysis location extraction means for extracting the analysis location;
It is characterized by having.

  Further, as an embodiment of the emission analysis apparatus according to the present invention, an emission analysis apparatus that sequentially performs emission analysis on a plurality of samples, sample identification information is given to each sample, and sample identification information is provided by the selection means. The analysis site extraction procedure is associated with each of the analysis site extraction procedures, and the analysis site extraction means determines the sample identification information of the given sample to obtain the corresponding analysis site extraction procedure, and 1 on the analysis surface of the sample according to the procedure. Or it can be set as the structure which extracts a several analysis location.

  Here, the image analysis means photographs the analysis surface of the sample by an imaging unit such as a CCD camera, and detects a defective part by judging the difference between the normal part and the defective part based on the photographed image, for example, by brightness. Can be.

  The analysis location extraction procedure is an algorithm for determining the location of one or more analysis locations on the analysis surface, and is determined initially with a procedure for determining the number of analysis locations and the initial location of the analysis locations. And a procedure for determining the position of an analysis location in place of a defective portion at the location of the analysis location. That is, different types of analysis location extraction procedures, for example, the number of analysis locations to be extracted is different, the initial location of the analysis location is different, or there is a defect in the site where the analysis location is to be set. In some cases, the method of shifting the position of the analysis location is different to avoid it.

  First, the operator determines an appropriate analysis site extraction procedure for each of a plurality of samples to be analyzed according to the type of sample, the purpose of analysis, and the like, and sets the correspondence between the sample and the analysis site extraction procedure by the selection means. When analysis is started and a certain solid sample is given, the image analysis means detects a defect from the image of the analysis surface of the sample. The analysis location extraction means reads the analysis location extraction procedure corresponding to the sample from the storage means according to the setting by the selection means, and follows this procedure to avoid the defect indicated by the defect information obtained by the image analysis means. Extract multiple analysis points.

  According to the emission analysis apparatus according to the present invention, for example, when the number of analysis points is small as in an in-furnace analysis and there are many defects on the analysis surface of the sample, for example, the two analysis points are separated as much as possible on the analysis surface Can be set to In addition, when high analysis accuracy is required as in the case of evaluating the quality of a finished product, a large number of analysis points can be set while avoiding defective portions on the analysis surface. In this way, analysis locations can be set at optimal numbers and positions according to the type of sample, the purpose of analysis, etc., and analysis results with the required accuracy can be obtained while improving analysis throughput.

  An embodiment of an emission analyzer according to the present invention will be described below with reference to FIGS. FIG. 1 is a schematic configuration diagram of an emission analyzer according to the present embodiment.

  The emission analyzer includes an analysis unit 2 having a discharge electrode for performing emission analysis on a sample 6 conveyed by a robot arm 4 and an imaging unit including a CCD camera 12 for imaging an analysis surface of the sample 6. 10, a sample identification unit 18 for identifying the sample 6 to be transported, and a control / processing unit 20 connected to the analysis unit 2, the robot arm 4, the CCD camera 12, and the sample identification unit 18. For example, the control / processing unit 20 can be configured around a personal computer equipped with dedicated control / processing software, and an input unit 30 such as a keyboard and a display unit 31 such as a monitor are connected to each other.

  In the imaging unit 10, the sample 6 is held by the robot arm 4 so that the analysis surface stands vertically, and the CCD camera 12 is installed sideways so as to image the analysis surface from the front. Further, a half mirror 16 and an illumination lamp 14 are installed so as to illuminate the analysis surface from the photographing direction and the coaxial direction. The entire photographing unit 10 is surrounded by a shielding cover 8 in order to avoid external light that interferes with photographing. It should be noted that appropriate modifications such as using other imaging means such as a C-MOS camera instead of the CCD camera 12 are possible.

  The control / processing unit 20 reads the image data obtained by the CCD camera 12 and performs image analysis to detect a defective portion on the analysis surface of the sample 6, and on the analysis surface of the sample 6. An extraction information storage unit 22 for storing a plurality of extraction algorithms (corresponding to the analysis site extraction procedure in the present invention) for extracting one to a plurality of analysis locations, and the defect portion according to one of the plurality of extraction algorithms The analysis location extraction processing unit 23 extracts one or more analysis locations on the analysis surface of the sample 6, and the number assigned to each sample and the type of extraction algorithm used when extracting the analysis location for the sample Selection information storage unit 24 that stores the correspondence relationship between the robot arm 4 and the transfer control unit that controls the operation of the robot arm 4 based on the extracted position (coordinate) information of one or more analysis points. 5 includes an analysis control unit 26 for controlling the analysis unit 2 to perform emission analysis when the specimen 6 is set so as to be possible to analyze a given analysis point, as functional blocks.

  The sample identification unit 18 identifies the samples 6 that are sequentially given. Here, it is assumed that the sample number that is a serial number given to each sample 6 is read. For example, a bar code is used, etc. The method is not particularly limited as long as each sample can be specified.

  Next, an example of an analysis point extraction algorithm on the analysis surface of the sample 6 will be described with reference to FIG. The shape of the sample 6 is a flat cylindrical shape, and one end surface of the circular shape is the analysis surface 40. FIG. 4 is a plan view of the analysis surface 40, and (A) to (D) are schematic diagrams for explaining different extraction algorithms. Basically, the analysis portion is provided on a circumference 41 that is ½ of the radius r of the circular analysis surface 40. Here, it is assumed that the following four different extraction algorithms A, B, C, and D are prepared.

[Extraction algorithm A]
As shown in FIGS. 4 (A) to 4 (a), four locations that are 90 degrees apart from each other on the circumference 41 are set as initial analysis location candidates 42a, 42b, 42bc, and 42d. If there is no defect in this initial analysis location candidate, these 4 locations are determined as analysis locations. On the other hand, as shown in FIGS. 4A to 4B, when the defective portion 43 exists at the position of the initial analysis location candidate (42a in this example), the predetermined position is determined in the predetermined direction from the position. The next analysis location candidate 42a ′ is reset at a position separated by an angle (10 ° counterclockwise in this example), and if there is no defect at the location of the analysis location candidate 42a ′, the analysis location candidate 42a ′ and the other three analysis location candidates 42b, 42c, and 42d that originally had no defect are determined as analysis locations.

  If there is also a defect at the position of the analysis location candidate 42a ′, this time, the analysis location candidate is reset at a position away from the initial analysis location candidate 42a by 10 ° in the reverse direction (clockwise direction). Then, it is determined whether or not there is a defective portion at that position. In this way, the analysis point candidates are gradually shifted (by a predetermined angle step) on the circumference 41 to a position where the defect portion does not exist (however, up to ± 90 °), and finally four analysis points are determined.

[Extraction algorithm B]
As shown in FIGS. 4B to 4A, the initial four analysis site candidates 42a, 42b, 42bc, and 42d are set in the same manner as the extraction algorithm A, and the initial analysis site candidates are defective. If there is no overlap with the part, these four locations are determined as analysis locations. When there is a defect in at least one of the analysis location candidates, as shown in FIGS. 4 (B)-(b), while maintaining the angular position relationship of the four analysis location candidates, that is, the angular interval of 90 °, The whole is rotated by a predetermined angle (10 ° counterclockwise in this example) to set four new analysis location candidates 42a ′, 42b ′, 42c ′, 42d ′. In this state, if there is no defect at the position of a new analysis location candidate, these 4 locations are determined as analysis locations. In this case, since the angular positional relationship between the four analysis points is maintained, it is effective in obtaining a representative value of the analysis result of the sample.

[Extraction algorithm C]
Basically, it is the same as the extraction algorithm B, but the analysis point candidates are not two points but two points across the center point of the analysis surface 40 (that is, the separation angle is 180 °). Determine the analysis location to maintain and avoid defects.

[Extraction algorithm D]
In this case, the maximum number of analysis points P is determined in advance. Usually, since it is used in multi-point analysis with more than four points, the maximum number of analysis points P is set to 6 or 8, for example, but P = 1 may be used. As shown in FIGS. 4D to 4A, the first analysis location candidate 42a is determined at an appropriately determined reference position (this is assumed to have an angle of 0 °), and a predetermined angle (in this example) is determined from the position in a predetermined direction. In this case, the next analysis location candidate 42b is set at a position separated by 10 ° in the counterclockwise direction, and the next analysis location candidate 42c is similarly set at a position away from the position by a predetermined angle. The analysis points up to the maximum number of analysis points P are determined. As shown in FIGS. 4D to 4B, when the defect portion 43 exists at the position of the analysis location candidate, the analysis location candidate is set at a position separated by a predetermined angle after skipping the defect portion 43.

  Thus, according to the four types of extraction algorithms A, B, C, and D described above, one to a plurality of analysis points can be set on the analysis surface 40 of the sample 6. Even when the same number of analysis points are set, the position may differ depending on the type of extraction algorithm (A, B, C, or D). You can set the location. Note that the above four types of extraction algorithms are merely examples, and it is natural that many other algorithms for extracting analysis points can be considered.

  In each extraction algorithm, it is preferable that the extracted analysis locations are extracted so as to be separated by a minimum distance interval appropriately defined. If a plurality of analysis positions are overlapped, the analysis accuracy is affected. However, it is possible to avoid the overlap of analysis positions. In addition, when a normal part is not found in each extraction algorithm, it is good to judge that it is a defective sample and reject it. As a result, since the defective sample can be discarded without being analyzed, a quick response to the next processing can be performed, and as a result, the throughput is greatly improved.

  Next, a series of operations for analyzing one sample in the emission analyzer of the present embodiment will be described with reference to the flowchart of FIG.

  Prior to the continuous analysis of a plurality of samples, the operator inputs the correspondence between the number of the sample to be analyzed and the type of extraction algorithm used when analyzing each sample from the input unit 30. An appropriate extraction algorithm may be designated according to the type of sample, the state of the sample, or the purpose of analysis. Specifically, for example, so-called in-furnace analysis has a demand for analysis of a small number of analysis points as far as possible because there are many defects on the analysis surface of the sample. The extraction algorithm B or C is effective for such purposes. On the other hand, when it is desired to perform a precise analysis to evaluate the quality of the finished product, the extraction algorithm D is effective because it is necessary to set as many analysis points as possible. Therefore, the operator selects an appropriate extraction algorithm according to the type of each sample and the purpose of analysis. The information input from the input unit 30 is held in the selection information storage unit 24 as a correspondence table between sample numbers and extraction algorithm types, for example, as shown in FIG.

  When the continuous analysis is started and a certain sample 6 is given to the apparatus (step S1), the sample identification unit 18 reads the sample number given to the sample 6 in order to identify the sample, and the sample number Is sent to the control / processing unit 20 (step S2).

  The sample 6 held by the robot arm 4 is transferred to the imaging unit 10, and the imaging unit 10 images the analysis surface 40 of the sample 6 (step S3). That is, the illumination light from the illumination lamp 14 is applied to the analysis surface 40, and the CCD camera 12 images the entire analysis surface 40. The obtained image data is sent to the image processing unit 21. The image processing unit 21 detects a defective part from the difference in luminance from the normal part, and the coordinate information indicating the position of the defective part is sent to the analysis location extraction processing unit 23. (Step S4).

  The analysis location extraction processing unit 23 refers to the selection information storage unit 24 for the sample number sent from the sample identification unit 18 to recognize the type of extraction algorithm used for the sample 6. Then, the corresponding extraction algorithm is read from the extraction information storage unit 22 (step S5). Then, according to the extraction algorithm and using the defect information, extraction of one or more analysis points on the analysis surface 40 of the sample 6 is executed (step S6). For example, when analyzing the sample of sample number <4>, since the extraction algorithm C is designated, two analysis locations that are 180 ° apart from each other are extracted by the above-described procedure.

  When the position of the analysis location is thus determined, coordinate information indicating the location of one or more analysis locations is sent to the transport control unit 25 and the analysis control unit 26. The transport control unit 25 controls the robot arm 4 based on the coordinate information, and positions the sample 6 so as to analyze one analysis site on the analysis surface 40 of the sample 6 (step S7).

  When positioning of the sample 6 is completed, under the control of the analysis control unit 26, a spark discharge is performed from the analysis electrode of the analysis unit 2 to a predetermined position on the analysis surface of the sample 6, and emission analysis corresponding to this is performed. (Step S8). Next, the analysis control unit 26 determines whether or not analysis of all necessary analysis locations has been completed (step S9). If there are still analysis locations remaining, the process returns to step S7 to analyze the next analysis location. To repeat the sample positioning and analysis execution.

  When all the analyzes for one sample 6 are completed in this way, the robot arm 4 discharges the sample 6 into, for example, an analyzed sample container (step S10). When performing continuous analysis of a large number of samples, the processing of steps S1 to S10 is repeated to collect emission analysis data for each sample.

  Note that all processes may not be performed automatically, but the process transition may be appropriately triggered by an operator's operation. For example, a new sample may be received or an image of the analysis surface of the sample may be captured by an operator's manual operation.

  In the present invention, the method for exciting the sample for emission analysis is not particularly limited, and naturally, for example, arc discharge, glow discharge, laser excitation or the like may be used in addition to spark discharge.

1 is a schematic configuration diagram of an emission analyzer that is one embodiment of the present invention. The figure which shows an example of the correspondence of a sample number and an extraction algorithm kind. The flowchart which shows the operation | movement at the time of analyzing one sample in the emission spectrometer of a present Example. The schematic diagram for demonstrating the analysis location extraction procedure in the emission spectrometer of a present Example.

Explanation of symbols

2 ... analyzing unit 4 ... robot arm 6 ... sample 8 ... shielding cover 10 ... imaging unit 12 ... CCD camera 14 ... illumination lamp 16 ... half mirror 18 ... sample identifying unit 20 ... control / processing unit 21 ... image processing unit 22 ... extraction Information storage unit 23 ... analysis location extraction processing unit 24 ... selection information storage unit 25 ... transport control unit 26 ... analysis control unit 30 ... input unit 31 ... display unit 40 ... analysis surface 41 ... circumferences 42a, 42b, 42c, 42d ... Analysis location candidate 43 ... Defect

Claims (2)

In an emission analyzer for performing emission analysis on one or more analysis points on the surface of a solid sample,
a) Image analysis means for capturing an analysis surface of a solid sample and detecting a defective portion based on the captured image;
b) Preliminarily storing a plurality of types of analysis site extraction procedures for determining the position of one or more analysis sites on the analysis surface of the solid sample so as to avoid the defective sites using the defect information obtained by the image analysis means. Storage means to keep,
c) a selection means for selecting an analysis site extraction procedure to be used corresponding to the sample to be analyzed prior to performing the analysis;
d) At the time of emission analysis, an analysis site extraction procedure corresponding to the sample to be analyzed is acquired by the storage unit based on the selection information by the selection unit, and according to the procedure, one or more on the analysis surface of the sample is acquired. An analysis location extraction means for extracting the analysis location;
An emission analysis apparatus comprising:   The emission analysis apparatus according to claim 1, wherein the emission analysis apparatus sequentially performs emission analysis on a plurality of samples, each sample identification information is given to each sample, and each sample identification information is analyzed by the selection means. A location extraction procedure is associated, and the analysis location extraction means obtains a corresponding analysis location extraction procedure by judging sample identification information of a given sample, and according to the procedure, one or a plurality of analysis location extraction procedures are obtained on the sample analysis surface. An emission analyzer characterized by extracting analysis points.

Images