Rare earth element analysis method, device, equipment, medium and product

Abstract

The invention relates to the technical field of element detection, and discloses a rare earth element analysis method, device and equipment, a medium and a product. The spectral data is obtained by performing LIBS detection on a rock sample of trace rare earth elements at a plurality of points; inputting the spectral data into an element analysis model to obtain a quantitative analysis result of the rare earth elements of the rock sample; wherein the element analysis model is a partial least square regression model which is trained based on a sample training set and passes the verification of a sample test set. By optimizing the data acquisition mode and establishing the trained and verified regression model, the problems of non-uniform distribution of trace rare earth elements in the rock and weak signals are effectively solved, and the accuracy of quantitative analysis is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of element detection, and particularly relates to a rare earth element analysis method, device, equipment, medium and product. Background Art

[0002] If the mineral components in the rock contain calcium-bearing rock-forming minerals - feldspar and hornblende, the calcium in them will undergo isomorphous substitution with rare earth elements, so that the rare earth elements are stored in the mineral lattice.

[0003] Laser induced breakdown spectroscopy (LIBS) is a technique that generates plasma spectra by exciting the surface of a sample with a high-energy pulsed laser. It can perform rapid, minimally invasive and in-situ detection, and is widely used in deep space exploration, mineral exploration, archaeological identification and other fields. Currently, conventional spectral analysis methods establish a model by analyzing the relationship between the intensity or peak area of characteristic spectral lines in the spectrum and the content of the element of interest, so as to predict the element content in a sample with unknown content of this element. In the above scheme, the choice of different quantitative analysis modeling methods has a great impact on the results of quantitative analysis. Therefore, there is an urgent need for a method of quantitative analysis modeling to accurately analyze rare earth elements. Summary of the Invention

[0004] In view of this, the present application provides a rare earth element analysis method, device, equipment medium and product, which can accurately perform quantitative analysis on rare earth elements. The technical solution is as follows.

[0005] In the first aspect, a rare earth element analysis method is provided. The method includes:

[0006] Obtain spectral data; the spectral data is obtained by performing LIBS detection on a rock sample of trace rare earth elements at multiple points.

[0007] Input the spectral data into an element analysis model to obtain a quantitative analysis result of the rare earth elements in the rock sample; wherein, the element analysis model is a partial least squares regression model trained based on a sample training set and verified by a sample test set.

[0008] In an optional implementation manner, the obtaining of the spectral data includes:

[0009] Move the three-dimensional moving platform to a plurality of specified positions in sequence to send high-energy pulsed laser to a plurality of points of the rock sample, so as to obtain the spectral data.

[0010] In an optional implementation manner, the sending of the high-energy pulsed laser to a plurality of points of the rock sample to obtain the spectral data includes:

[0011] Send high-energy pulsed lasers to multiple points of the rock sample to obtain candidate spectra corresponding to the multiple points respectively;

[0012] According to the deviation rates of the candidate spectra corresponding to the multiple points, select at least two candidate spectra corresponding to the points as the spectral data.

[0013] In an alternative embodiment, the sending high-energy pulsed lasers to multiple points of the rock sample includes:

[0014] For any target point among the multiple points, send multiple high-energy laser pulses to the target point of the rock sample to obtain multiple pulse spectra corresponding to the target point;

[0015] Perform an averaging process on the multiple pulse spectra to obtain the spectral data corresponding to the target point.

[0016] In an alternative embodiment, the method includes:

[0017] Obtain a sample training set and a sample test set;

[0018] Iteratively train a partial least squares regression model based on the training sample set to obtain a trained partial least squares regression model;

[0019] Verify the partial least squares regression model through the sample test set;

[0020] If it passes the verification, determine the partial least squares regression model as the element analysis model;

[0021] If it fails to pass the verification, iteratively train the partial least squares regression model again through the training sample set.

[0022] In an alternative embodiment, the iteratively training the partial least squares regression model based on the training sample set includes:

[0023] Obtain a number of candidate models; the number of candidate models are partial least squares regression models with different numbers of principal components respectively;

[0024] Based on the sample training set, perform cross-validation on the number of candidate models respectively to obtain training result indicators corresponding to the number of candidate models respectively;

[0025] Based on the training result indicators corresponding to the number of candidate models, select a partial least squares regression model with a target number of principal components among the number of candidate models, so as to iteratively train the partial least squares regression model with the target number of principal components based on the training sample set.

[0026] In a second aspect, a rare earth element analysis device is provided. The device includes:

[0027] A spectrum acquisition module for acquiring spectrum data. The spectrum data is obtained by performing LIBS detection on a rock sample containing trace rare earth elements at multiple points;

[0028] An element analysis module for inputting the spectrum data into an element analysis model to obtain a quantitative analysis result of the rare earth elements in the rock sample. Among them, the element analysis model is a partial least squares regression model trained based on a sample training set and verified by a sample test set.

[0029] In a third aspect, a computer device is provided. The computer device includes a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the above-mentioned rare earth element analysis method.

[0030] In a fourth aspect, a computer-readable storage medium is provided. A computer instruction is stored on the computer-readable storage medium, and the computer instruction is used to cause a computer to execute the above-mentioned rare earth element analysis method.

[0031] In a fifth aspect, a computer program product or a computer program is provided, including computer instructions, and the computer instructions are used to cause a computer to execute the above-mentioned rare earth element analysis method.

[0032] The technical solution provided by this application may include the following beneficial effects:

[0033] When rare earth element analysis is required, LIBS detection can be performed on a rock sample containing trace rare earth elements at multiple points to obtain spectrum data. At this time, by performing LIBS detection at multiple points, the non-uniformity of the distribution of trace rare earth elements in the rock sample can be captured, and the local deviation that may be brought by single-point sampling can be reduced, thereby improving the representativeness and stability of the overall data. Then, the spectrum data is input into the element analysis model. Since this element analysis model is obtained by training a partial least squares regression model using a training set and verifying its prediction effect through a test set, it has good prediction accuracy. In the above solution, by optimizing the data acquisition method and establishing a regression model that has been trained and verified, the problems of uneven distribution of trace rare earth elements and weak signals in rocks are effectively solved, and the accuracy of quantitative analysis is improved. Description of the Drawings

[0034] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0035] Figure 1 The flowchart of a rare earth element analysis method according to an embodiment of the present application is shown.

[0036] Figure 2 The flowchart of a rare earth element analysis method according to an embodiment of the present application is shown.

[0037] Figure 3 The schematic diagram showing the relationship between the root mean square error of cross - validation and the coefficient of determination and the number of principal components is shown.

[0038] Figure 4 The schematic diagram showing the prediction result with the sample spectral data as the test set is shown.

[0039] Figure 5 The logic diagram of a method for quantitatively analyzing trace rare earth elements in rocks based on LIBS combined with partial least squares regression according to an embodiment of the present application is shown.

[0040] Figure 6 It is the schematic diagram of the structure of a rare earth element analysis device provided by an embodiment of the present application.

[0041] Figure 7 It is the schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention. Specific Embodiments

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0043] In the description of the embodiments of the present application, the term "corresponding" may represent a direct or indirect corresponding relationship between two entities, may also represent an associated relationship between two entities, or may be a relationship such as indication and being indicated, configuration and being configured, etc.

[0044] Rare earth elements have a wide range of applications in fields such as geological exploration, aerospace, and national defense. They are divided into light, medium, and heavy rare earth elements and are known as the "industrial vitamins". If the mineral components in rocks contain calcium-bearing rock-forming minerals - feldspar and hornblende, the calcium in them will have a isomorphic substitution phenomenon with rare earth elements, thus enabling rare earth elements to be hosted in the mineral lattice. Relevant research shows that the rare earth content in the weathering crust of Emei basalt can meet the requirements of industrial indicators. Moreover, there are also considerable amounts of rare earth elements on the moon. Since a large amount of basalt magma erupted on the moon and flowed into the impact basins on the near side of the moon, forming mare lava plains, the basalt in them is called "mare basalt". The distribution of mare basalt is relatively consistent with the distribution range of the "Oceanus Procellarum-KREEP terrane", and the Oceanus Procellarum-KREEP terrane is rich in KREEP components (KREEP) with abundant K, P, and REE. Quantitative analysis of trace rare earth elements in rocks is of practical significance for exploring rare earth minerals, analyzing mineral components, and deep space exploration.

[0045] Laser induced breakdown spectroscopy (LIBS) is a technique that generates plasma spectra by exciting the surface of a sample with a high-energy pulsed laser. It can perform rapid, minimally invasive, and in-situ detection and is widely used in fields such as deep space exploration, mineral exploration, and archaeological identification. Currently, conventional spectroscopic analysis methods establish models by analyzing the relationship between the intensity of characteristic spectral lines or the peak area in the spectrum and the content of the element of interest, so as to predict the element content in samples with unknown element content. The method combined with machine learning can achieve quantitative analysis of rare earth elements, but the quantitative research on trace rare earth elements is not yet abundant. Therefore, it is necessary to select a suitable machine learning method based on LIBS for the quantitative analysis of trace rare earth elements.

[0046] At present, there are still some challenges in the quantitative analysis of trace rare earth elements: First, the distribution of rare earth elements in minerals is uneven, which may result in the absence of rare earth elements in the detection area and the inability to detect effective characteristic spectral lines; Second, the characteristic spectral lines of rare earth elements are rich and easily confused with other elements. The spectral line intensities of trace rare earth elements generally decrease. Especially after being incorporated into the rock matrix, some characteristic spectral lines of rare earth elements will be submerged by the spectral signals of the rock matrix. Selecting appropriate spectral bands is beneficial to improving the prediction effect; Second, choosing different quantitative analysis modeling methods has a great impact on the results of quantitative analysis. Therefore, making rock samples doped with trace rare earth elements available in the laboratory and combining the LIBS method with suitable machine learning methods are particularly important for realizing the quantitative analysis of trace rare earth elements. Among the existing machine learning methods, partial least squares is suitable for the spectral data analysis with not very large data volume. Usually, a single characteristic spectral line or multiple characteristic spectral lines are selected as single-variable or multi-variable inputs to the PLS model. To solve the problems of complex process and low accuracy in the quantitative analysis of trace rare earth elements in the rock matrix in the prior art, this application provides a rare earth element analysis method, which solves the problem of non-utilization of low-intensity characteristic spectral lines, improves the utilization efficiency of characteristic spectral lines, and ultimately improves the prediction effect.

[0047] Please refer to Figure 1 , which shows a method flowchart of a rare earth element analysis method according to an embodiment of this application. As Figure 1 shown, this method is executed by a computer device, and the rare earth element analysis method includes:

[0048] Step 101, obtain spectral data; this spectral data is obtained by performing LIBS detection on a rock sample of trace rare earth elements at multiple points.

[0049] In the embodiment of this application, the LIBS (Laser Induced Breakdown Spectroscopy) technology can be used to detect the rock sample. LIBS uses a high-energy pulsed laser to irradiate the surface of the sample, causing the sample to evaporate locally and instantaneously at high temperature and form a plasma. Each element in the plasma will emit spectral signals with specific wavelengths, and the characteristic spectral lines of trace rare earth elements will also be captured.

[0050] Moreover, trace rare earth elements in rocks usually show the characteristic of uneven distribution. Therefore, by performing LIBS detection at multiple points, more spectral data at different positions inside the sample can be obtained, thereby effectively reducing the measurement deviation caused by local non-uniformity and ensuring the representativeness and stability of the data.

[0051] Specifically, in actual operation, the laser parameters (such as pulse energy, laser frequency, spectrometer acquisition delay, and gate width, etc.) can also be adjusted to optimize the signal-to-noise ratio and ensure that weak trace element signals can be accurately captured.

[0052] Step 102: Input the spectral data into the elemental analysis model to obtain the quantitative analysis results of the rare earth elements in the rock sample. Among them, the elemental analysis model is a partial least squares regression model trained based on a sample training set and verified by a sample test set.

[0053] When obtaining the preprocessed spectral data collected at multiple points, the spectral data can be used as input to establish an elemental analysis model based on partial least squares regression. This model uses the data of rock samples with known concentrations (i.e., the training set) to establish the quantitative relationship between spectral features and rare earth element contents.

[0054] Specifically, first use samples containing known rare earth element concentrations to train the model through partial least squares regression to fit the linear relationship between each spectral variable (such as characteristic spectral line intensity or peak area) and element concentration.

[0055] To prevent overfitting of the model and at the same time verify the generalization ability of the model, then use a part of the samples (test set) to predict the trained model, and use the root mean square error (RMSE) and coefficient of determination (R 2 ) as evaluation indicators. Only when these indicators meet the expected requirements is the model considered applicable to actual quantitative analysis, and thus the partial least squares regression model that meets the expected requirements is used as the elemental analysis model.

[0056] The partial least squares regression model after training and verification can convert newly collected spectral data into specific rare earth element concentration values, thereby realizing the quantitative analysis of trace rare earth elements in rock samples.

[0057] In summary, when rare earth element analysis is required, LIBS detection can be performed on rock samples of trace rare earth elements at multiple points to obtain spectral data. At this time, by performing LIBS detection at multiple points, the uneven distribution of trace rare earth elements in the rock sample can be captured, reducing the local deviation that may be brought by single-point sampling, thereby improving the representativeness and stability of the overall data. Then, the spectral data is input into the elemental analysis model. Since the elemental analysis model is obtained by training the partial least squares regression model using the training set and verifying its prediction effect through the test set, it has good prediction accuracy. In the above solution, by optimizing the data acquisition method and establishing a trained and verified regression model, the problems of uneven distribution of trace rare earth elements and weak signals in rocks are effectively solved, and the accuracy of quantitative analysis is improved.

[0058] Please refer to Figure 2 , which shows the method flowchart of a rare earth element analysis method involved in an embodiment of the present application. As Figure 2As shown, this method is executed by a computer device, and this rare earth element analysis method includes:

[0059] Step 201, obtain a rock sample.

[0060] In the embodiments of the present application, before analyzing rare earth elements, certain process treatments need to be carried out on the rock sample to improve the accuracy of subsequent analysis of the rock sample. Taking a basalt sample as an example, basalt samples doped with different contents of samarium elements can be prepared through the following steps: Prepare a 10 mg / mL samarium chloride solution using samarium chloride hexahydrate powder, use a pipette to drop the samarium chloride solution into the basalt substrate according to the samarium element content required for each sample, add laboratory distilled water to submerge the basalt, use a mixer to mix the doped samples, and number them according to the samarium element content: Samples with samarium element contents of 50, 60, 70, 80, 90, 100, 110, 120, 130, and 140 ppm correspond to serial numbers 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 respectively. After drying the above samples, put them into an automatic tablet press and press them into tablets with boric acid as the substrate; the pressure range is up to 35.1 tons and down to 34.7 tons, and the pressing time is 2 minutes. The surface of the sample is flat and the inside is compact.

[0061] Step 202, move the three-dimensional moving platform to a plurality of specified positions in sequence to send high-energy pulsed lasers to a plurality of points of the rock sample to obtain spectral data.

[0062] In the embodiments of the present application, random points are taken on the surface of each sample for LIBS detection: Control the three-dimensional platform to move to a plurality of specified positions in sequence to move the sample so that the laser acts on different positions. The three-dimensional platform moves at a fixed step size, and the selectable moving step sizes include 0.5 mm, 1 mm, 2 mm, etc.

[0063] In a possible implementation manner, send high-energy pulsed lasers to a plurality of points of the rock sample to obtain candidate spectra respectively corresponding to the plurality of points;

[0064] According to the deviation rates of the candidate spectra respectively corresponding to the plurality of points, screen out at least two candidate spectra respectively corresponding to the points as the spectral data.

[0065] Further, for any target point among the plurality of points, send a plurality of high-energy laser pulses to the target point of the rock sample to obtain a plurality of pulsed spectra corresponding to the target point;

[0066] Perform an averaging process on the plurality of pulsed spectra to obtain the spectral data corresponding to the target point.

[0067] Specifically, the collected LIBS spectral data are input into the PLS model: there are 22 excitation positions for each sample, and each position is excited by a total of 50 lasers. The results of every 10 excitations are accumulated into one spectrum, and 5 spectra are averaged into one spectrum, that is, each position contributes 1 spectrum, each sample contributes 22 spectra, and 2 spectra with large deviations are eliminated, that is, each sample contributes 20 spectra. The 22 excitation positions can expand the LIBS detection area and avoid the situation where the characteristic spectral lines of the samarium element cannot be detected due to the uneven distribution of the samarium element; the elimination of the two spectra with large deviations is to look through the 22 average spectra and eliminate the two spectra with the strongest and weakest spectral intensities corresponding to the characteristic spectral lines with the strongest general spectral intensity; the average spectrum and the elimination of the two spectra with large deviations can both improve the spectral stability, thereby improving the prediction accuracy of the subsequent model analysis.

[0068] The LIBS equipment used in this embodiment is configured as follows: the laser light source is a Nd:YAG Q-switched laser with a wavelength of 1064nm, the laser energy is set to 140.1mJ, the laser pulse width is 6ns, the repetition frequency is 10Hz, and under standard atmospheric pressure and room temperature of 25°C, the spectrometer acquisition delay time is 0.5μs, and the spectrometer acquisition gate width time is 9μs. The sample is placed on an electrically controlled XYZ three-dimensional displacement platform in the sample chamber. The laser beam is coupled and focused to the sample chamber through the optical path, and after acting vertically on the sample surface for 0.5μs, it begins to collect the excited spectrum with a gate width time of 9μs. The excited spectrum is transmitted to the fiber optic spectrometer through the collection optical path and optical fiber.

[0069] Step 203: input the spectral data into an element analysis model to obtain a quantitative analysis result of the rare earth elements of the rock sample.

[0070] The element analysis model is a partial least squares regression model that is trained based on a sample training set and verified by a sample test set.

[0071] Specifically, in the embodiment of the present application, the element analysis model can be obtained in the following manner:

[0072] Get sample training set and sample test set;

[0073] Iteratively training the partial least squares regression model based on the sample training set to obtain a partial least squares regression model after training;

[0074] The partial least squares regression model is verified through the sample test set;

[0075] If the verification is passed, the partial least squares regression model is determined as the element analysis model;

[0076] If the verification fails, the partial least squares regression model is iteratively trained again using this training sample set.

[0077] In the embodiments of the present application, the structure of the PLS model includes a preprocessing part and a quantitative analysis part. The preprocessing part performs operations such as smoothing, baseline removal, and noise reduction on the input LIBS spectral data to reduce the influence of the energy fluctuation output by the laser during use and the air flow in the environment on the plasma; the quantitative analysis part divides the input spectral data into a training set and a test set, and establishes a model to predict the element content.

[0078] Further, when iteratively training the partial least squares regression model, it is first necessary to determine the number of principal components of the partial least squares regression model. The specific steps include:

[0079] Obtain a number of candidate models; the number of candidate models are partial least squares regression models with different numbers of principal components respectively;

[0080] Based on this sample training set, perform cross-validation on the number of candidate models respectively to obtain the training result indicators corresponding to the number of candidate models respectively;

[0081] Based on the training result indicators corresponding to the number of candidate models respectively, select a partial least squares regression model with the target number of principal components from the number of candidate models, so as to iteratively train the partial least squares regression model with the target number of principal components based on this training sample set.

[0082] In the process of determining the target number of principal components corresponding to the partial least squares regression model, a validation set is divided from the sample training set, the spectral data of other serial number samples is used as the training set, a PLS model is established and simply trained using the training set, and then the validation set is used to verify the PLS model to obtain the corresponding evaluation index.

[0083] Specifically, in the embodiments of the present application, the prediction effect of the PLS model (i.e., a number of candidate models) with each number of principal components can be evaluated using leave-one-out cross-validation. The indicators for evaluating the prediction effect are the root mean square error of cross-validation and the coefficient of determination of the prediction set, and the number of principal components with the best prediction effect is selected. As Figure 3 shown, when the number of principal components is 7, the coefficient of determination of leave-one-out cross-validation is the largest, and at the same time its root mean square error is the smallest. Therefore, the number of principal components is selected as 7 to establish the subsequent partial least squares regression model. The prediction result of quantitatively analyzing the content of samarium element is as Figure 4 shown.

[0084] Please refer to Figure 5 , which shows the logic diagram of a method for quantitatively analyzing trace rare earth elements in rocks based on LIBS combined with partial least squares regression involved in the embodiments of the present application. As Figure 5 shown, the method includes the following logical steps:

[0085] 1. Sample preparation: Prepare rock samples with different rare earth element contents. Specifically, different amounts of rare earth element solutions can be doped into rock powders to ensure relatively uniform distribution of rare earth elements in the samples. Subsequently, these samples are pressed or sintered into solids for subsequent laser-induced breakdown spectroscopy (LIBS) detection.

[0086] 2. Optimize LIBS acquisition parameters, specifically including laser pulse energy, spectrometer gate width time, and delay time. In this method, by adjusting these parameters, the spectral signal intensity of trace rare earth elements can be maximally increased, the interference of matrix element signals can be reduced, and the best signal-to-noise ratio can be obtained simultaneously.

[0087] 3. Spectral acquisition and preprocessing: Under the optimized laser parameters, perform LIBS detection on different samples and multiple detection points to obtain spectral data. Then, perform spectral preprocessing, using smoothing (removing high-frequency noise to obtain a smoother spectral line), noise reduction (reducing interference such as environmental noise and instrument fluctuations), and baseline removal (eliminating the influence of background radiation on peak intensity to highlight the characteristic spectral lines of trace rare earth elements) and other methods to provide higher-quality and purer spectral data for subsequent modeling, reducing the interference of noise and background signals.

[0088] 4. PLS model training and parameter optimization. During the model training process, the sample data with known concentrations can be divided into a training set (i.e., the above sample training set) and a test set (i.e., the above sample test set). Among them, the training set is used to train the model, and the test set is used to verify the prediction effect of the trained model.

[0089] Furthermore, to improve the training effect of the model, cross-validation can be used, that is, a validation set is divided from the training set, and the prediction effect of the model under latent variables is verified through cross-validation, so as to select the best number of principal components among different numbers of principal components, and establish the partial least squares regression model required for subsequent use based on the best number of principal components. The partial least squares regression model is iteratively trained through the training set. When a certain iterative training is completed, if the evaluation indicators obtained after testing the partial least squares regression model through the test set meet the requirements, it can be considered that the partial least squares regression model has completed training and has good prediction accuracy. Specifically, the evaluation indicators can use the root mean square error (RMSE) and the coefficient of determination (R 2 ) etc., which are used to measure the prediction accuracy and stability of the model. In the above process, the best number of principal components is selected through cross-validation, which not only avoids overfitting but also ensures that the model has good generalization ability.

[0090] 5. Model verification and quantitative analysis of rare earth elements, that is, using the test set samples to further verify the prediction effect of the model, ensuring high accuracy and stability on unknown samples. After the model verification is completed, the trained optimal PLS model can be applied to the spectral data of unknown rock samples to achieve quantitative analysis of trace rare earth elements.

[0091] In summary, when rare earth element analysis is required, LIBS detection can be performed on rock samples containing trace rare earth elements at multiple points to obtain spectral data. By performing LIBS detection at multiple points, the non-uniform distribution of trace rare earth elements in the rock samples can be captured, reducing the local deviation that may be caused by single-point sampling, thereby improving the representativeness and stability of the overall data. Then, the spectral data is input into the element analysis model. Since this element analysis model is trained using a training set for the partial least squares regression model and its prediction effect is verified through a test set, it has good prediction accuracy. In the above solution, by optimizing the data acquisition method and establishing a regression model that has been trained and verified, the problems of uneven distribution and weak signals of trace rare earth elements in rocks are effectively solved, and the accuracy of quantitative analysis is improved.

[0092] In an embodiment of the present application, a rare earth element analysis device is also provided. This device is used to implement the above embodiments and preferred implementation manners, and those that have been described will not be repeated. As used hereinafter, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0093] An embodiment of the present application provides a rare earth element analysis device. Figure 6 FIG. is a schematic structural diagram of a rare earth element analysis device provided by an embodiment of the present application. This device is set in a computer device and includes:

[0094] A spectrum acquisition module 601, configured to acquire spectral data; the spectral data is obtained by performing LIBS detection on rock samples containing trace rare earth elements at multiple points;

[0095] An element analysis module 602, configured to input the spectral data into an element analysis model to obtain a quantitative analysis result of the rare earth elements of the rock sample; wherein, the element analysis model is a partial least squares regression model trained based on a sample training set and verified through a sample test set.

[0096] In summary, when rare earth element analysis is required, LIBS detection can be performed on rock samples containing trace rare earth elements at multiple points to obtain spectral data. By performing LIBS detection at multiple points, the non-uniformity of the distribution of trace rare earth elements in the rock samples can be captured, reducing the local deviation that may be caused by single-point sampling, thereby improving the representativeness and stability of the overall data. Then, the spectral data is input into the element analysis model. Since this element analysis model is obtained by training a partial least squares regression model using a training set and validating its prediction effect through a test set, it has good prediction accuracy. In the above solution, by optimizing the data acquisition method and establishing a regression model that has been trained and validated, the problems of uneven distribution and weak signals of trace rare earth elements in rocks are effectively solved, and the accuracy of quantitative analysis is improved.

[0097] The further function descriptions of the above-mentioned various modules and units are the same as those in the corresponding embodiments above, and will not be repeated here.

[0098] The rare earth element analysis device in this embodiment is presented in the form of functional units. Here, the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0099] An embodiment of the present invention further provides a computer device, which has the above-mentioned Figure 6 shown rare earth element analysis device to analyze rare earth elements or train a model.

[0100] Please refer to Figure 7 , Figure 7 is a schematic structural diagram of a computer device provided by an optional embodiment of the present invention. As Figure 7 shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphic information in a graphical user interface on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a set of blade servers, or a multi-processor system). Figure 7 One processor 10 is taken as an example in

[0101] The processor 10 may be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 may further include a hardware chip. The above-mentioned hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The above-mentioned programmable logic device may be a complex programmable logic device, a field programmable gate array, a generic array logic, or any combination thereof.

[0102] Among them, the memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiments.

[0103] The memory 20 may include a program storage area and a data storage area. Among them, the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may further include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 may optionally include a memory remotely disposed relative to the processor 10, and these remote memories may be connected to the computer device through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0104] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive; the memory 20 may further include a combination of the above types of memories.

[0105] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0106] Embodiments of the present invention also provide a computer-readable storage medium. The methods according to the embodiments of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented as computer code that is originally stored in a remote storage medium or a non-transitory machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the methods described herein can be stored as such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by the computer, the processor, or the hardware, the methods shown in the above embodiments are implemented.

[0107] A part of the present invention can be applied as a computer program product, such as computer program instructions. When executed by a computer, through the operation of the computer, the methods and / or technical solutions according to the present invention can be called or provided. Those skilled in the art should be able to understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executes the instructions, or the computer compiles the instructions and then executes the corresponding compiled program, or the computer reads and executes the instructions, or the computer reads and installs the instructions and then executes the corresponding installed program. Herein, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to the computer.

[0108] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A rare earth element analysis method, characterized in that: The method comprises: Acquiring spectral data; the spectral data is obtained by performing LIBS detection on a rock sample of trace rare earth elements at multiple points; The spectral data is input into an element analysis model to obtain a quantitative analysis result of the rare earth elements of the rock sample; wherein the element analysis model is a partial least squares regression model trained based on a sample training set and verified by a sample test set.

2. The method according to claim 1, characterized in that The obtaining of spectral data comprises: The three-dimensional mobile platform is moved to a plurality of designated positions in sequence to send high-energy pulse lasers to a plurality of points of the rock sample to obtain the spectral data.

3. The method according to claim 2, characterized in that The step of sending high energy pulsed lasers to multiple points of the rock sample to obtain the spectral data comprises: Sending high-energy pulsed lasers to multiple points of the rock sample to obtain candidate spectra corresponding to the multiple points respectively; According to the deviation rates of the candidate spectra respectively corresponding to the multiple points, candidate spectra respectively corresponding to at least two points are screened out as the spectrum data.

4. The method according to claim 2, characterized in that: The step of sending high energy pulse laser to multiple points of the rock sample comprises: For any target point among the multiple points, sending multiple high-energy laser pulses to the target point of the rock sample to obtain multiple pulse spectra corresponding to the target point; The multiple pulse spectra are averaged to obtain spectrum data corresponding to the target point.

5. The method according to any one of claims 1 to 4, characterized in that: The method comprises: Get sample training set and sample test set; Iteratively training the partial least squares regression model based on the training sample set to obtain a partial least squares regression model after training; Verifying the partial least squares regression model through the sample test set; If the verification is passed, the partial least squares regression model is determined as the element analysis model; If the verification fails, the partial least squares regression model is iteratively trained again using the training sample set.

6. The method according to claim 5, characterized in that The iterative training of the partial least squares regression model based on the training sample set includes: Obtaining several candidate models; the several candidate models are partial least squares regression models with different numbers of principal components; Based on the sample training set, cross-validate the several candidate models respectively to obtain training result indicators corresponding to the several candidate models respectively; Based on the training result indicators respectively corresponding to the several candidate models, a partial least squares regression model of the target number of principal components is selected from the several candidate models, so as to iteratively train the partial least squares regression model of the target number of principal components based on the training sample set.

7. A rare earth element analysis device, characterized in that: The device comprises: A spectrum acquisition module is used to acquire spectrum data; the spectrum data is obtained by performing LIBS detection on rock samples of trace rare earth elements at multiple points; The element analysis module is used to input the spectral data into the element analysis model to obtain the quantitative analysis results of the rare earth elements of the rock sample; wherein the element analysis model is a partial least squares regression model trained based on a sample training set and verified by a sample test set.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the rare earth element analysis method according to any one of claims 1 to 6 by executing the computer instructions.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the rare earth element analysis method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the rare earth element analysis method according to any one of claims 1 to 6.

Images