Compound identification method and compound identification device

Abstract

To provide a method for classifying and estimating metal elements and compounds included in an unknown sample.SOLUTION: A compound identification method of the present invention performs factorization of spectral data for a standard sample of a compound species to be assumed to derive a factor and a factor score representing the degree of influence of the factor, and learning and setting in advance a score threshold for classifying the compound species according to the factor score in a decision tree and a factor application order to apply the factor score to the decision tree, calculates a factor score in an unknown sample by using a factor of the standard sample to spectral data of the unknown sample, and classifying the factor score of the unknown sample in the decision tree by using the score threshold according to the factor application order to identify the compound species included in the unknown sample.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a method for classifying and estimating metal elements and compounds contained in an unknown sample. [Background technology]

[0002] Decommissioning nuclear power plants requires measures to deal with contaminated water, remove fuel from spent fuel pools, remove fuel debris, and dispose of waste. In reactors with melted cores, molten fuel, control rods, and other materials are thought to have solidified into fuel debris. In reactors that use boron carbide (B4C) as a control material, borides, which are approximately twice as hard as oxides, are produced, making it efficient to remove fuel debris while distinguishing between metals, oxides, and borides. It has also been confirmed that there is a correlation between the boron concentration and hardness of borides. To study fuel debris removal methods, it is necessary to identify the material and hardness of unknown samples. Furthermore, the creation of material and hardness distributions within unknown samples is also being considered.

[0003] Methods for qualitative and quantitative analysis of elements in samples include inductively coupled plasma atomic emission spectrometry (ICP-AES) and laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). When a sample is externally energized, the elements contained in the sample emit light of element-specific wavelengths, and elemental analysis is performed by measuring the wavelength and amount of emitted light. ICP-AES can sometimes be inaccurate due to interference. Patent Document 1 also discloses an invention for a foreign matter analysis device that can accurately and easily analyze foreign matter contained in a sample.

[0004] Laser-induced breakdown spectroscopy (LIBS) is another method for elemental analysis, regardless of the state of the material, such as solid, liquid, or gas. LIBS is an atomic emission spectroscopy method that uses a high-energy pulsed laser as an excitation source. It enables analysis in a short time without complex pretreatment processes and is capable of simultaneously measuring elements ranging from light elements (boron and oxygen) to heavy elements (metals). Therefore, it is expected to be applied to obtaining elemental information on high-level waste, hardness measurement, and criticality control based on boron distribution information. Patent Document 2 also discloses an invention for a method for calculating the composition of oxygen and boron in materials whose main component is metal, as well as the hardness of the material.

[0005] In LIBS, a high-peak-power laser with a pulse width of less than a nanosecond is irradiated onto the target, instantly forming a plasma at tens of thousands of degrees Celsius. The target is atomized and excited by the plasma, and as the plasma disappears, the excited atoms relax to their ground state, emitting element-specific fluorescent light. This is measured to identify and quantify the target element. Fluctuations in the laser output and the plasma generated by the elements that make up the target are subject to minute fluctuations in time and space, resulting in fluctuations in the measured values.

[0006] In the case of LIBS, background due to the plasma state (time, space, etc.) and, in the case of energy discrimination processing, scattering components caused by the interaction of radiation with matter have a significant effect on the measurement values. In spectroscopic measurements that use radiation such as lasers, X-rays, and neutrons as probes, although there are various factors and their influence varies in magnitude, a function to cancel or correct them is desired to achieve highly accurate measurements. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Patent No. 6507757 [Patent Document 2] Japanese Patent Publication No. 2020-067337 Summary of the Invention [Problem to be solved by the invention]

[0008] However, when analyzing unknown solid samples, such as alloys and melts containing many elements, spectral data with many emission wavelengths is obtained, making it difficult to distinguish between elements and compounds. For example, the emission lines of iron (Fe) and zirconium (Zr) overlap so much that quantitative analysis can be hindered.

[0009] Furthermore, in measurements using spectrometers that use not only lasers but also X-rays, neutrons, etc., the dark current is measured with the spectrometer closed to eliminate its influence. In order to eliminate the influence of the measurement environment (air environment, inert environment, etc.), some methods use blank data (measurements without a sample) to eliminate the influence. However, this method cannot be applied when these influences are not added to the device functions or when similar samples do not exist.

[0010] In addition, when using laser-based optical emission spectrometry (LIBS and LA-ICP-MS) or neutrons, which have self-shielding, it is necessary to evaluate sample-specific errors, but when it comes to neutrons and plasma emission, the irradiation energy is often not controllable.

[0011] In particular, in LIBS measurements, the spectral baseline (a continuous line formed by connecting points thought to be the background) can fluctuate depending on the ambient gas. For example, an argon (Ar) atmosphere has high luminescence sensitivity, which can produce white noise that broadens the baseline over a wide wavelength range. In addition, multiple detectors are used to detect light with wavelengths from 190 nm to 900 nm, but because the detectable wavelength range varies depending on the characteristics of the optical system, a shift in the baseline occurs at the wavelength where the detector switches.

[0012] As such, there is a demand for reducing background noise components (variation errors such as dark current in spectrometers) in spectral analyses such as spectroscopic analysis and radiation analysis, optimizing probe conditions for light and radiation on the object being measured in various analyses (reducing damage caused by the probe to the object being measured), and identifying various compounds (including metals, ceramics, etc.).

[0013] Therefore, an object of the present invention is to provide a method for classifying and estimating metal elements and compounds contained in an unknown sample. Another object is to evaluate and correct sample-specific errors in classifying and estimating metal elements and compounds. [Means for solving the problem]

[0014] In order to solve the above-mentioned problems, the compound identification method of the present invention is characterized in that it factorizes spectral data of a standard sample of an expected compound species to derive factors and factor scores representing the influence of the factors, learns and sets in advance a score threshold for classifying the compound species using the factor scores in a decision tree and a factor application order for applying the factor scores to the decision tree, calculates factor scores for an unknown sample using the factors of the standard sample for the spectral data of the unknown sample, and classifies the factor scores of the unknown sample using the score threshold in the decision tree in accordance with the factor application order, thereby identifying the compound species contained in the unknown sample.

[0015] In the compound identification method, the spectral data is corrected to a baseline approximated by a penalized asymmetric least squares method so that errors are evaluated more severely when the measured value is below the estimated value than when it is above the estimated value.

[0016] In the compound identification method, the factorization is characterized in that a matrix Y with K rows and N columns, which is composed of the number of wavelength samples (wavelength dimension) K and the number of samples N of the spectrum data, is decomposed using M specified non-negative basis vectors, and the influences of the obtained factors are derive as the factor scores.

[0017] In the compound identification method, the decision tree is characterized in that factors and score thresholds are derived in advance such that a cost function using impurity becomes a minimum value when classifying the compound types.

[0018] The compound identification method is characterized in that the factor scores are mapped to estimate the components and composition ratios of the unknown sample.

[0019] The compound identification device of the present invention is characterized by comprising an irradiation device that irradiates the sample with a laser, a spectroscope that resolves light emitted from the sample irradiated with the laser, and a plurality of detectors that acquire spectral data generated by the spectroscope by dividing it into wavelengths within a predetermined range, and identifying compounds contained in the sample using the spectral data input from the detectors. [Effects of the Invention]

[0020] According to the present invention, it is possible to classify and estimate the metal elements and compounds contained in an unknown sample. By decomposing the spectral data of known substances into multiple factors in advance and training the system to classify elements and compounds using a decision tree, the components of the unknown sample can be classified.

[0021] Furthermore, by evaluating and correcting for sample-specific errors in advance, metal elements and compounds can be classified and estimated with high accuracy. Post-processing correction is also possible, and errors, including white noise, can be removed. [Brief explanation of the drawings]

[0022] [Figure 1] 1 is a schematic view of a measuring device in a compound identification method according to the present invention. [Figure 2] 1 is a flowchart showing the flow of the preprocessing step and the learning step of the compound identification method of the present invention. [Figure 3] FIG. 1 is a diagram illustrating the correction of spectral data in the compound identification method of the present invention. [Figure 4] FIG. 1 is a diagram illustrating the correction of spectral data in the compound identification method of the present invention. [Figure 5] FIG. 1 is a diagram illustrating factorization of spectral data in the compound identification method of the present invention. [Figure 6] FIG. 1 is a diagram illustrating factorization of spectral data in the compound identification method of the present invention. [Figure 7] FIG. 1 is a diagram illustrating classification by a decision tree in the compound identification method of the present invention. [Figure 8] FIG. 1 is a diagram illustrating classification by a decision tree in the compound identification method of the present invention. [Figure 9] FIG. 1 is a diagram illustrating classification by a decision tree in the compound identification method of the present invention. [Figure 10] FIG. 1 is a flowchart showing the flow of operational steps of a compound identification method according to the present invention. [Figure 11] FIG. 1 shows the results of classification by the compound identification method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0023] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Note that components having the same functions will be assigned the same reference numerals, and repeated description thereof may be omitted.

[0024] This invention relates to an algorithm for two processes in spectroscopic measurement systems that use radiation such as visible light, infrared light, X-rays, gamma rays, and neutrons: a process for reducing background fluctuations caused by the stability of the measurement system, and a process for identifying the spectrum of the measurement target. This algorithm consists of a machine learning method that uses measurement data (including predicted data) as training data, and a data processing system that results from this method. Its implementation is expected to improve measurement accuracy and robustness. Furthermore, because it can be implemented as software processing, it is highly adaptable to analysis programs installed in existing spectroscopic analysis devices.

[0025] To improve the analytical accuracy of element identification and quantification using LIBS (Laser-Induced Breakdown Spectroscopy), parameters (reference parameters) that characterize the plasma state (time, space, etc.) are measured simultaneously in addition to the elemental emission spectrum, and the correlation between the two is used for correction processing. Specifically, reference parameters that are strongly correlated with the intensity and shape of the atomic emission spectrum to be measured are set. Traditionally, these parameters have been set by highly skilled and experienced technicians who have grasped the fluctuating components and characteristics. To achieve high analytical accuracy, evaluation that understands the spectral fluctuation mechanisms is necessary.

[0026] Machine learning is used to set the reference parameters. Conventional methods search for the spectrum of the measurement target as a peak with a certain width, whereas this invention applies a machine learning method (NMF: Non-negative Matrix Factorization) to big data (all spectral components consisting of sets of wavelength-count values ​​(light intensity values) obtained from various analytical devices, etc.) that is difficult for humans to handle, and constructs the reference parameters. By using all spectral components instead of signal peaks, it is possible to evaluate the spectral intensity, shape, etc. when identifying unknown samples.

[0027] This invention is inspired by the methods for extracting specific sounds such as violins and violas from an orchestra, or for extracting a specific individual from a crowd, and has been applied to a measurement device. The algorithm of this invention uses spectral data of standard samples processed in the preprocessing phase, trained using NMF and decision trees in the training phase, and then applied to unknown samples in the operation phase. [Example]

[0028] 1 is a schematic view of a measurement device in the compound identification method of the present invention. The measurement device 100 includes a chamber 200, an irradiation device 300, a spectrometer 400, a detector 410, and a control device 220, and performs LIBS measurement.

[0029] The chamber 200 is a vacuum vessel or the like, and the sample 110 is placed inside after reducing the pressure using a vacuum pump 210. As the sample 110, a known sample such as a standard sample 110a is mainly used in the learning stage, and an unknown sample 110b to be measured and analyzed is mainly used in the operational stage. The standard sample 110a is a sample whose composition is known in advance, such as a metal element or compound, while the unknown sample 110b is a sample whose composition is unknown, such as fuel debris brought in from a nuclear power plant.

[0030] The irradiation device 300 is a laser head or the like equipped with a laser oscillator, and irradiates the sample 110 in the chamber 200 with irradiation light 310 such as a pulsed laser. The irradiation light 310 is an Nd:YAG laser or the like with a wavelength of 1064 nm, which uses a crystal of YAG (yttrium, aluminum, garnet) doped with Nd (neodymium) as a medium.

[0031] The surface of the sample 110 is turned into plasma by irradiating it with irradiation light 310, and emitted light 320 is detected when the surface returns to the ground state. The emitted light 320 is collimated using a collimator or the like and then incident on a spectrometer 400. The spectrometer 400 is a prism, diffraction grating, or the like, and separates the light into wavelengths. Spectral data 420 representing the intensity distribution is acquired by arranging multiple detectors 410 side by side, dividing the data into wavelengths within a predetermined range, and sent to the control device 220.

[0032] The control device 220 is a computer or the like, which controls each device and classifies and analyzes the sample 110 from the acquired spectral data 420. In addition, a sequencer 230, an electrically driven stage 240, a cooling water circulation device, etc. may be provided as necessary.

[0033] For example, a stage 240 carrying the sample 110 is moved and scanned over multiple points under the control of a sequencer 230. The laser irradiation position is obtained from the position information (coordinate information of the X, Y, and Z axes) of the stage 240, and shape information (height information) of the sample 110 is obtained using a displacement meter. The coordinates are mapped by a control device 220 to visualize the shape and intensity information in three dimensions.

[0034] Figure 2 is a flowchart showing the flow of the preprocessing and learning stages of the compound identification method. The preprocessing stage includes standard sample preparation (setting the compound type of the standard sample) L1, spectral data acquisition (baseline estimation) L2, and the learning stage includes NMF execution (acquiring the factors and factor scores of the standard sample) L3, and decision tree rule generation (setting the score threshold and factor application order) L4.

[0035] In the preprocessing and learning stages, the expected compound species are set and a standard sample 110a is prepared, the acquired spectrum data 420 of the standard sample 110a is factorized to derive the factors and factor scores of the standard sample 110a, and the score threshold for classifying the compound species by the factor scores in the decision tree and the factor application order for applying the factors to the decision tree are learned and set in advance as rules for the decision tree.

[0036] The preprocessing stage consists of standard sample preparation L1 and spectral data acquisition L2 in Figure 2. In the standard sample preparation L1 step, compound types such as metal elements and compounds that are expected to be components of the unknown sample 110b to be measured are set in advance, and these metal elements and compounds are prepared as standard sample 110a. Standard sample 110a serves as data prepared in advance to determine the influence and extent of each factor when identifying the compound types contained in unknown sample 110b.

[0037] In the spectral data acquisition step L2, the spectral data 420 sent from the detector 410 (including data previously stored in a storage device, etc.) is input to the control device 220. Since the spectral data 420 contains noise and the like, and the baseline 430 (baseline) is disturbed, baseline estimation is performed in the control device 220. In the baseline estimation, the spectral data 420 is corrected so that the baseline 430 becomes smooth.

[0038] Next, an example of the process related to the above-mentioned pretreatment step obtained by LIBS measurement will be described with reference to Figures 3 and 4. Figures 3 and 4 are diagrams for explaining the correction of spectral data in the compound identification method.

[0039] Figure 3(a) shows a spectrum acquired by dividing the LIBS measurement system into multiple measurement wavelength ranges and using a spectrometer with high sensitivity for each range to broaden the bandwidth (190 nm to 900 nm). As shown in Figure 3(a), it can be seen that a shift occurs in the baseline 430 at the boundary of the wavelength range of the detector 410. This shift causes a discontinuity in the baseline 430 due to the sensitivity and dark current of the detector 410.

[0040] In wavelength ranges where discontinuities occur, the measurement results contain many uncertain factors, making the results unreliable. Also, if the target has a spectrum in the discontinuous part, it is not possible to calculate the accurate area, and in some cases it becomes impossible to distinguish the spectrum.

[0041] 3(b) is an enlarged spectrum of a portion of the spectrum data 420 in FIG. 3(a). As shown in FIG. 3(b), unless optimal smoothing is applied to the detector resolution, the baseline will be adjusted more than necessary.

[0042] Therefore, the inventors have devised a method for correcting the deviation (discontinuity) of the baseline 430 at the boundary of the wavelength range of the multiple detectors 410 by asymmetric weighted penalty least squares smoothing (asymmetric least squares method with penalty terms) shown in FIG. 3(c). The asymmetric least squares method with penalty terms corrects the deviation (discontinuity) of the baseline 430 at the boundary of the wavelength range of the multiple detectors 410 by using the measured value y i Estimated baseline z for i The desirable baseline 430 is derived by minimizing the evaluation function so that the desirability (goodness) of the baseline 430 is not only minimized but also smoothed, minimizing the difference between the measured value and the estimated value.

[0043] From the viewpoint of evaluation, first, the square error Σ i (y i -zi ) 2 From the perspective of i ≦z i The error is very strict when y i >z i In addition, when z i Changes in fluctuations are evaluated strictly from the perspective of fluctuations in w i is the asymmetric weight, and y i >z i Then p and y i ≦z i Then, it becomes (1-p). p (p∈[0,1]: p is a very small value) is the value of w depending on whether the measured value exceeds the estimated value. i μ is a regularization parameter that controls the fluctuation (non-smoothness) of the estimated value Σ i (Δ 2 z i ) 2 is the penalty term. 2 represents the second-order differential operator, and w i , μ are hyperparameters that are empirically determined by the designer. The optimal solution can be found by varying the parameters.

[0044] This causes the spectral data 420 to be corrected to a baseline 430 that is approximated using a penalized asymmetric least squares method so that errors are assessed more severely when the measured values ​​are below the estimated values ​​than when the measured values ​​are above the estimated values.

[0045] Figure 4 shows a spectrum obtained by correcting the spectrum data 420 in Figure 3 using the correction shown in Figure 3(c). Figure 4(a) shows a spectrum obtained by correcting the spectrum data 420 in Figure 3(a), and Figure 4(b) shows an enlarged view of a portion of the spectrum data 420 in Figure 4(a).

[0046] As shown in Figure 4(a), the smoothed baseline 430 after correction eliminates the discontinuity of the spectrum in Figure 3(a), and background removal makes it easier to identify the shape and intensity of the spectral data 420. This improves analytical accuracy in terms of element identification and quantification, making spectral evaluation, which was previously difficult, possible. Note that this method is not limited to the LIBS measurement in this example, but can be applied to all measurement devices that use a spectrometer 400.

[0047] In addition to asymmetric weight penalty least squares smoothing, the method also applies pattern recognition removal from the baseline 430 under sample-free conditions to improve the performance specific to various analytical devices and reduce sample dependency. This method is applicable not only to analytical methods using a laser light source such as Raman spectroscopy and LA-ICP-MS, but also to all measurement devices 100 that use a spectrometer 400.

[0048] An example of removal using pattern recognition is to perform pattern recognition on the spectrum obtained when measuring only the measurement environment under conditions without a sample (blank run), and then remove it at the baseline. Generally, this is done by simply subtracting the conditions under which there is no sample, but by making full use of pattern recognition, it is possible to correct for the baseline, which fluctuates with each measurement.

[0049] 5 and 6 are diagrams illustrating the factorization of spectral data in the compound identification method. In the NMF execution L3 step, the control device 220 performs non-negative matrix factorization (NMF) on the corrected spectral data 420.

[0050] NMF uses machine learning to decompose all overlapping emission spectrum data into a specified number of factors, and then uses an algorithm to determine which factors to focus on and what thresholds to use in a decision tree to correctly classify the substances for a group of sample data represented by pairs of factor scores and substance names, and classifies elements and compounds from unknown sample 110b by dividing them into related factors.

[0051] The components in the unknown sample 110b are estimated by converting the spectrum of the known substance's spectral data 420 into a low-dimensional factor score vector using a dimension reduction technique in advance, and training an algorithm that classifies elements, compounds, etc. using a decision tree.

[0052] In the learning stage, the spectral data 420 (K-dimensional vector group) shown in Figure 5(a) is collected and divided into M factors 440 using NMF, reducing the dimension so that it can be easily handled by a decision tree. NMF can be applied to any type of data as long as the spectral data can be expressed as a matrix of 0 and positive values.

[0053] As shown in FIG. 5(b), N pieces of measurement data Y (y1 to y N ) is approximately decomposed into factors H and factor scores U. i is vector data (intensity) of K dimensions (for example, K = 11,562 types of wavelengths) and is expressed as a K × N (K rows, N columns) matrix. Factor H is a representation of some factor common to the K-dimensional vector data as M basis vectors (for example, M = 3 factors), and is expressed as a K × M (K rows, M columns) matrix. Factor score U is a coupling coefficient for N pieces of measurement data Y and M pieces of factors H, and is expressed as an M × N (M rows, N columns) matrix. Note that M is a hyperparameter.

[0054] In the case of the standard sample 110a, a matrix consisting of the number of wavelength samples (wavelength dimension) and the number of samples in the spectrum data 420 is factorized with non-negative values, and the obtained factors (for example, the first to third factors) are set as factors 440 of the standard sample 110a, and coefficients representing the degree of influence of each factor 440 on the standard sample 110a are derived as factor scores 450. The factor scores 450 indicate which factor 440 has the greatest influence on the measurement data.

[0055] As an example using LIBS, four types of compounds, primarily borides, were selected, and 37 pieces (N = 37 × 4) of spectral data 420 were extracted from each of them as the standard sample 110a, resulting in a high-dimensional sample (number of wavelength dimensions for each spectrum, k = 11,562). The results of reducing the dimensionality of these samples using NMF (using three factors 440 with different feature values ​​and a factor score 450 as reference parameters) are shown below. As shown in Figure 6, after taking into account the influence of the factor score 450 on each factor 440, the measured spectral data (Figure 5(a)) can be expressed as a linear sum of the three factors.

[0056] The factor 440 may standardize the spectral data 420 using the mean value and standard deviation to align the intensity scale for each wavelength. If standardization is not required, this step may be omitted. Note that the sample is susceptible to damage due to laser irradiation and has inherent self-shielding in radiation measurements, and standardization can mitigate these issues.

[0057] 7 to 9 are diagrams illustrating classification using a decision tree in the compound identification method. In the decision tree rule generation step L4, the control device 220 creates a decision tree 500 using the factor score 450 of the standard sample 110a as a feature, and sets a score threshold 460 for classification by the factor score 450 when applying the factors 440 of the standard sample 110a, and a factor application order for determining which factor 440 of the standard sample 110a to apply first for classification using the decision tree 500.

[0058] The decision tree 500 is a classification method that judges "True" or "False" based on the reference parameters obtained by NMF, and is a search algorithm that correctly classifies substances using the focused factor 440 (reference parameter) and the score threshold 460 of the factor score 450 used for the judgment for the learning data.

[0059] The combination of NMF and decision tree 500 in the learning stage can be applied to analytical methods with spectral data 420 such as dynamic light scattering and inductively coupled plasma mass spectrometry (ICP-MS), as well as any type of data as long as it can be expressed in a matrix of "0" and "positive values." It is also expected to be applicable to image data such as TEM (transmission electron microscope), SEM (scanning electron microscope), and associated EDX (energy dispersive X-ray analysis).

[0060] As shown in FIG. 7(a), in order to perform conditional branching in the decision tree 500, a cost function J is used to calculate the impurity of the factor score 450 of the standard sample 110a, and a score threshold t that minimizes the cost function J is selected. k (1≦k≦m) is derived. The cost function J is calculated using Gini impurity, etc., where k is the factor number, m is the number of factors 440, and m left is the number of factors to divide into one side (left subset), m right is the number of factors (right subsets) to divide into the other, G left is the impurity of the left subset, G right is the impurity of the right subset.

[0061] If only one compound exists in a subset, it is a pure subset, and if multiple compounds are mixed, it is an impure subset. In this case, the impurity of a subset is expressed by the Gini impurity or cross entropy, which are based on the ratio of each compound contained in the subset.

[0062] As shown in FIG. 7( b), when the compound species (atmosphere) are Ar and N2, a score threshold 460 (e.g., 114.155) of the factor score 450 is derived for the factor 440 on the horizontal axis side, and in the decision tree 500, if the factor score 450 is smaller than the score threshold 460, the factor 440 is classified into the left subset (estimated to be N2), and if the factor score 450 is greater than the score threshold 460, the factor 440 is classified into the right subset (estimated to be Ar).

[0063] Figure 8(a) shows an example in which boride spectra (four compound types: CrB, FeB, NiB, and ZrB) measured by LIBS are decomposed into factors 440, feature quantities (factor scores 450) are extracted, and a decision tree 500 is created. The order of application of factors 440 is assumed to be the second factor, the third factor, and the first factor.

[0064] At node N1 of the decision tree 500, the score threshold 460 of the factor score 450 in the second factor is derived as 108.615, and as shown in Figure 8(b), compounds with a score greater than the score threshold 460 are classified as compounds containing Cr and B, and the rest are classified at node N2.

[0065] At node N2 of the decision tree 500, the score threshold 460 of the factor score 450 in the third factor is derived as 65.837, and as shown in Figure 9(a), compounds with scores greater than the score threshold 460 are classified as compounds containing Ni and B, and the rest are classified at node N3.

[0066] At node N3 of the decision tree 500, the score threshold 460 of the factor score 450 in the first factor is derived as 117.886, and as shown in Figure 9(b), compounds with scores greater than the score threshold 460 are classified as compounds containing Zr and B, and the rest are classified as compounds containing Fe and B.

[0067] Chromium (Cr) is classified at the beginning of decision tree 500 rather than zirconium (Zr) or iron (Fe), which have relatively large spectra, and decision tree 500 shows a more accurate classification than expected by humans. This is thought to be due to the fact that Zr and Fe contain many peaks that are affected by interference (the accuracy rate decreases when classification is simply based on signal peaks without interference, as in conventional methods).

[0068] The control device 220 stores the factors 440 and factor scores 450 obtained from the standard sample 110a, as well as the score threshold of the factor score 450 and the order of factor application of the factors 440 as rules for the decision tree 500, in a storage device such as a database, and creates a compound identification index. In the learning stage, it is determined what kind of standard sample 110a the data is from and how the factor scores 450 are for each factor 440, and in the operation stage, this is used as a compound identification index when classifying the unknown sample 110b.

[0069] 10 is a flowchart showing the flow of the operational steps of the compound identification method. The operational steps include steps of unknown sample measurement S1, spectral data correction S2, factorization processing S3, and decision tree classification processing S4. Each factor 440 and its factor score 450 of the standard sample 110a are prepared in advance as compound identification indices, and the unknown sample 110b is classified by sorting it based on the factor score 450 for each factor 440 of the unknown sample 110b.

[0070] In the operational stage, the factor score 450 for the unknown sample 110b is calculated for the spectral data 420 of the unknown sample 110b using the factors 440 of the standard sample 110a, and the factor score 450 for the unknown sample 110b is divided by the score threshold 460 in a decision tree 500 according to the order of application of the factors 440, and the class of the unknown sample 110b is estimated, thereby identifying the compound species contained in the unknown sample 110b.

[0071] In the unknown sample measurement S1 step, the measurement device 100 irradiates the unknown sample 110b with a laser, and the spectrum data 420 sent from the detector 410 is input to the control device 220.

[0072] In the spectral data correction step S2, the control device 220 also performs baseline correction on the spectral data 420 of the unknown sample 110b in the same manner as in the spectral data acquisition step L2.

[0073] In the factorization process S3 step, the control device 220 performs factorization by NMF on the corrected spectral data 420 of the unknown sample 110b using the factors 440 of the standard sample 110a, and calculates the factor score 450 of the unknown sample 110b. Note that standardization may be performed as necessary.

[0074] In the decision tree classification process S4 step, the control device 220 classifies the factor scores 450 of the unknown sample 110b into classes of factors 440 based on the rules (learned algorithm) of the decision tree 500. In the decision tree 500, the factor scores 450 of the unknown sample 110b are classified using the score threshold 460 in accordance with the factor application order of the factors 440, and the class to which the unknown sample belongs is estimated, thereby identifying the compound species contained in the unknown sample 110b.

[0075] By applying the decision tree 500, the factor scores 450 are associated with the compound types. Common factors 440 (factor 1 to factor m) are derived from all the spectral data 420 of the standard sample group, and each spectral data 420 is converted into a factor score 450 through each factor. Then, the compound types are classified through the decision tree 500.

[0076] In fact, when an unknown sample 110b was measured using a eutectic compound of stainless steel (SUS) and boron carbide (B4C), the accuracy rate was 100% for four types of test data x eight pieces.

[0077] Figure 11 shows the results of classification using the compound identification method. The factor scores 450 may be mapped to estimate the components and composition ratios of the unknown sample 110b. As shown in Figure 11(a), it is also possible to estimate whether a simulated sample contains elements contained in a certain compound. Also, as shown in Figure 11(b), it is possible to estimate the composition ratio to determine whether a sample contains a large amount of CrB2.

[0078] By controlling the coordinates on two axes, the X and Y axes, and measuring the distance between the stage and the sample using a displacement meter, information on the Z axis height can be obtained and the shape of the sample can be determined. Coordinate control is easy using an electric stage. LIBS measurement results can be obtained by pinpointing the position of the sample with the X, Y, and Z axes known, and performing LIBS measurements. The range of the emission spectrum of a characteristic element can be selected from the LIBS measurement results, and the area value and height can be calculated. To display mapping information of elements in two dimensions, the X and Y coordinate information can be set on two axes, and the selected elements can be mapped by displaying the intensity of the emission spectrum of the selected element in color.

[0079] According to the present invention, it is possible to classify and estimate the metal elements and compounds contained in an unknown sample. By decomposing the spectral data of known substances into multiple factors in advance and training the system to classify elements and compounds using a decision tree, the components of the unknown sample can be classified.

[0080] Furthermore, by evaluating and correcting for sample-specific errors in advance, metal elements and compounds can be classified and estimated with high accuracy. Post-processing correction is also possible, and errors, including white noise, can be removed.

[0081] In this invention, reference parameters that characterize the plasma state (time, space, etc.) are used to decide whether to accept or reject the atomic emission spectrum to be measured. Furthermore, machine learning is used to set the reference parameters. Whereas conventional methods search for the spectrum to be measured as a peak with a certain width, this invention uses NMF to machine-learn big data (all spectral components obtained from various analytical devices, etc.) that is difficult for humans to handle, and constructs the reference parameters. As a result, by processing all spectral components rather than processing by signal peaks as in conventional methods, the amount of data used for evaluation is reduced by 10 3 As a result, it has become possible to evaluate spectral intensity and shape, which was previously difficult, and there is a prospect of establishing high analytical accuracy.

[0082] Depending on the type, baseline correction used in preprocessing can be applied to each spectrum using functions such as asymmetric weighted penalty least-squares smoothing or machine learning such as pattern recognition. If there is an environment where databases from public research institutions or known samples can be measured as training data, it is possible to retroactively improve the performance of various analytical instruments already in possession (not only analytical methods using laser light sources such as Raman scattering and LA-ICP-MS, but all measuring instruments that use spectroscopy) and reduce sample dependency.

[0083] All that is required is to use databases (spectroscopy and radiation spectrum information) from public research institutions and measurements of standard samples with known material compositions as training data, and it is possible to easily improve the performance specific to various analytical instruments and reduce sample dependency without having to modify the hardware, such as by implementing it in the instrument's analysis software.A wide variety of applications are expected, including not only spectroscopic measurements using light sources such as lasers, X-rays, and neutrons as probes, but also electron microscopes (transmission electron microscopes: TEM-EDX, scanning electron microscopes: SEM-EDX) equipped with energy dispersive X-ray analysis (EDX) as a spectroscopic function, and electron probe microanalyzers (EPMA) that visualize spectroscopic data as surface information.

[0084] Although the embodiments of the present invention have been described above, the present invention is not limited to these. For example, NMF is a machine learning technique used to separate orchestral sounds into factors such as violins and violas, and can create and classify characteristic sounds (reference parameters) of violins and violas. In this way, the present invention can be applied to a variety of big data.

[0085] Furthermore, the material distribution and hardness distribution within the unknown sample may be created, and may be applied to multivariate analysis. [Explanation of symbols]

[0086] 100: Measuring equipment 110: Sample 110a: Standard sample 110b: Unknown sample 200: Chamber 210: Vacuum pump 220: Control device 230: Sequencer 240: Stage 300: Irradiation device 310: Irradiation light 320:Emission light 400: Spectrometer 410: Detector 420: Spectral data 430: Baseline 440:Factor 450: Factor score 460: Score threshold 500: Decision Tree

Claims

1. A control device comprising: factorizing spectral data of a standard sample of a predicted compound species to derive factors and factor scores representing the influence of the factors; learning and setting in advance a score threshold for classifying the compound species using the factor scores in a decision tree and a factor application order for applying the factor scores to the decision tree; calculating factor scores for the unknown sample using the factors of the standard sample for the spectral data of the unknown sample, and classifying the factor scores of the unknown sample using the decision tree in accordance with the factor application order and the score threshold, thereby identifying the compound species contained in the unknown sample; A compound identification method characterized by:

2. The spectral data is corrected to a baseline fitted using a penalized asymmetric least squares method so that errors are assessed more severely when the measured values ​​are below the estimated values ​​than when they are above them. The compound identification method according to claim 1 .

3. The factorization involves decomposing a K-row, N-column matrix consisting of K wavelength dimensions and N sample numbers of the spectrum data using M specified non-negative basis vectors, and deriving the influences of the factors obtained as the factor scores.

3. The compound identification method according to claim 1 or 2.

4. The decision tree derives factors and score thresholds that minimize a cost function using impurity when classifying the compound species.

4. The compound identification method according to claim 1, wherein the compound is a nucleotide sequence.

5. Executing the compound identification method according to any one of claims 1 to 4, A compound identification device characterized by:

6. an irradiation device that irradiates the unknown sample with a laser; a spectrometer for resolving light emitted from the unknown sample irradiated by the laser; a detector for acquiring spectral data generated by the spectrometer; Identifying compounds contained in the unknown sample using the spectral data input from the detector; 6. The compound identification device according to claim 5.

Images