Automated libs-based system for minero-chemical analysis

Abstract

The present invention relates to an automated system based on laser-induced plasma spectroscopy, for minero-chemical, elemental and mineralogical analysis in samples with irregular surfaces, which comprises a system for the input, processing and output of information. The system of the invention generates results that provide the composition associated with the structure of a sample.

Description

[0001] AUTOMATED LIBS-BASED SYSTEM FOR THE ELEMENTAL AND MINERALOGICAL ANALYSIS OF IRREGULAR SURFACES

[0002] The present invention consists of an automated system based on laser-induced breakdown spectroscopy (LIBS) plasma spectroscopy technology, for mineralogical, elemental and mineralogical analysis of irregular surface samples.

[0003] This system constitutes a robust analytical tool for mineral samples, for example, in the field of mineral exploration. This technique offers advantages such as efficient analysis times, high-quality and representative information, among others.

[0004] The system of the invention allows obtaining results that not only indicate the composition or presence of elements and / or minerals, but also deliver as a result the composition associated with the structure of the sample (results layer), thanks to the development of a unique system of input, processing and output of information.

[0005] STATE OF THE ART

[0006] LIBS technology is widely used in the mining industry due to its ability to perform precise and rapid elemental analysis. This method allows for the determination of the chemical composition of mineralogical samples without the need for complex chemical pretreatment. LIBS provides detailed information on the distribution and concentration of chemical elements present on the surface of the samples, which is crucial for quality control, geological exploration, and the optimization of mining processes.

[0007] In a typical LIBS-based detection system, a high-energy laser pulse is focused by a lens onto the sample surface, generating a plasma, which is a partially ionized gas composed of excited atoms and free electrons. The light emitted by the plasma, which contains information about the elements present in the sample, is captured and directed to a spectrometer. In this device, the different wavelengths are separated and analyzed, producing a unique spectrum that acts as a fingerprint of the sample's elemental composition. The resulting spectral information is sent to a computer system for processing. This system compares the detected emission lines with certified databases, allowing the elements present to be identified and quantified in a matter of seconds.

[0008] 2LIBS technology stands out as an innovative method of elemental chemical analysis, offering multiple advantages, including: - no prior sample preparation required, - can analyze samples in any state of aggregation (solid, liquid or gaseous), - only a few milligrams of ablated material are needed to perform the analysis, - fast and accurate real-time analysis (in-situ and online).

[0009] Relevant prior art patent documents and a general comparison with the proposed invention are mentioned below.

[0010] Document C 10S824825A discloses an invention in the field of laser spectroscopy detection and mineral classification, describing a method and device for classifying waste minerals based on laser-induced breakdown spectroscopy (LIBS). The device comprises a mineral pulverizer, a filter, a conveying device, a sample shape inspection system, a LIBS detection system, a mechanical mineral sorting and collection system, and a measurement and control system. The pulverizer and filter are used to crush bulk waste minerals into smaller mineral blocks of similar size and shape. The sample shape inspection system is used to obtain the average heights and limit dimensions of the samples and to further control the focusing distance and scanning range of the LIBS detection system.The LIBS detection system is used to detect the composition of the ore material and determine its grade. The mechanical classification system transports the ore and waste rock to separate ore collection tanks based on the detection results. The measurement and control system is used to provide unified coordination control over all functional units of the device.

[0011] However, the data input does not include equipment for analyzing irregular surfaces, nor does it consider inputting information related to a mineralogical spectral database. Regarding data processing, the methodology is not comparable given the differences in the type of input data. And as for the data output, it does not provide results that integrate compositional parameters associated with a structure (results layer). Furthermore, it only provides the mineral grade, not the multi-element composition. It also does not provide mineralogical information or distribution maps, only point measurements.

[0012] 3Document EP3605062A1 relates to the area of ​​machine learning, in particular machine learning for the analysis of high or super resolution spectroscopic data, which normally comprises the analysis of highly complex substance samples / mixtures and / or low resolution data, e.g., laser-induced decomposition spectroscopy (LIBS).The present invention relates to a computational self-learning method for characterizing one or more constituents of a sample from electromagnetic spectral information of that sample. This method shifts the paradigm associated with prior art methods by using only sub-optical spectral information, that is, by obtaining the resolution of the spectral information and thus being able to extract spectral lines—determining a spectral line position—from that spectral information, thereby avoiding all the uncertainty associated with pixel-based methods. A computational apparatus configured to implement this method is also an object of the present invention.

[0013] However, the method for capturing sample information and other input data is not developed. Furthermore, there are differences in how this information is processed and obtained. Specifically, it does not provide a result that integrates compositional parameters associated with a structure (results layer). It does not provide mineralogical information or distribution maps, only point measurements. And it does not perform measurements on irregular surfaces.

[0014] US patent 2.015346103A1 describes a laser-induced breakdown spectroscopy (LIBS) system and a method for generating spectral images of a sample surface. The system uses a high-frequency pulsed laser to produce a train of laser pulses directed toward the sample by a scanning mechanism. This mechanism allows the laser beam to be moved across the sample surface in a controlled manner. Each laser pulse generates a LIBS signal from a specific point on the surface, which is captured by a spectrometer. The resulting spectra are correlated with the exact coordinates of each laser beam position. From this correlation, a two-dimensional (2D) map is obtained that represents both the spatial distribution and chemical composition of the sample surface. This map is used to construct a detailed spectral image of the sample surface.

[0015] However, the data input does not include equipment for analyzing irregular surfaces, nor does it consider inputting information related to a mineralogical spectral database. The processing and analysis of the data to obtain sample results is not specifically developed. Furthermore, the output data does not correspond to actual results.

[0016] 4 that integrate composition parameters associated with a 3D structure (results layer). It also does not provide multi-element or mineralogical information, nor a distribution map on irregular surfaces.

[0017] Document WO2G22243020A1 describes a method for compositional analysis using laser-induced breakdown spectroscopy (LIBS). This method involves providing a sample with a specific surface area for analysis. The ablation point is then moved to multiple positions on the sample surface, following a path defined by a series of adjacent arcs. These arcs extend from one edge of the area to the opposite edge, efficiently covering the surface. An energy source is pulsed to deliver an electromagnetic energy beam to excise material at the ablation point. The emission spectra generated in response to the pulsed energy source are captured and analyzed to determine the chemical composition at each point on the surface.

[0018] However, the data input does not include equipment for analyzing irregular surfaces, nor does it consider inputting information related to a mineralogical spectral database. The processing and analysis of the data to obtain sample results is not specifically developed. And the output data does not include results that integrate compositional parameters associated with a structure (results layer).

[0019] Patent CA3G5393GA1 provides an integrated method and system for characterizing the elemental and molecular composition of a material using a combination of laser-based infrared (IR) spectroscopy and laser-induced breakdown spectroscopy (LIBS). The approach involves performing both measurements at the same analysis point on the material. IR spectroscopy allows for the identification and characterization of the material's molecular composition, while LIBS analyzes its elemental composition. From these data, 2D and 3D profiles of a material sample can be obtained.

[0020] However, the data input does not include equipment for analyzing irregular surfaces, nor does it consider the input of information related to a mineralogical spectral database. The processing and analysis of the data to obtain sample results are not specifically developed. Finally, the output only provides elemental information via LIBS. Furthermore, it does not perform a 3D scan of the sample (results layer). Document WO2015ÍÍ93850A1 describes a method for calibration or testing in laser-induced breakdown spectroscopy (LIBS) aimed at improving the accuracy and reliability of chemical analysis. The procedure consists of using a series of reference samples with known concentrations of an analyte. These samples are analyzed to obtain their respective LIBS spectra.For each reference LIBS spectrum, the intensity of at least one spectral line corresponding to the analyte is measured, the variation of which depends on its concentration. Likewise, the maximum amplitude of at least one saturation line of a reference element present in the sample is measured. Based on the relationship between these measurements and known analyte concentrations, a calibration curve for the analyte is constructed.

[0021] However, the data input does not include equipment for analyzing irregular surfaces, nor does it consider inputting information related to a mineralogical spectral database. The processing and analysis of the data to obtain sample results is not specifically developed. Finally, the output does not include results that integrate compositional parameters associated with a structure (results layer). Furthermore, the output only considers elemental concentration for a given matrix, uses the intensity of atomic lines instead of AI-based models, and does not provide an elemental and mineralogical distribution map of the concentration.

[0022] WO2022.2S2982.A1 describes a system for the online classification of mineral samples based on mineralogical analysis. This system includes a conveyor that moves a flow of mineral samples along a conveyor path, a LIBS module, a height measuring device, a focus controller, and an airflow system. The LIBS module projects a LIBS laser beam along an optical path focused on an analysis point with dimensions on the order of one or more mineral components of the mineral samples and captures the generated LIBS light signal. The height measuring device and focus controller work together to adjust the laser focus in real time to move the analysis point perpendicular to the conveyor according to the height of the mineral samples crossing the optical path. The airflow system moves particles away from the optical path.A processing unit performs mineralogical analysis of the mineral sample based on the LIBS light signal.

[0023] However, the information input does not consider a 3D mapping of the surface, nor the input of information related to a database of mineralogical spectra.

[0024] 6Furthermore, given the differences in the type of input data, the data processing methodology is not comparable. Finally, the output data does not correspond to the multi-element concentration of the sample (results layer). Additionally, it analyzes samples of mineral concentrates, not rocks, and does not provide a 3D elemental and / or mineralogical distribution map.

[0025] According to the analysis of the documents previously identified as part of the state of the art, it is concluded that none of them compromise the novelty of the present invention, since it incorporates unique characteristics not addressed in the reviewed background.

[0026] The main differences of the invention include: (i) specialized equipment configuration, designed specifically for the analysis of samples with irregular surfaces, (ii) advanced data processing, based on a database of mineralogical spectra, allows for the efficient integration of elemental and mineralogical information, (iii) optimization in the delivery of results, generating fast, accurate and detailed results, with a reduced margin of error and significantly shorter analysis times.

[0027] Although the documents identified as part of the prior art describe the use of LIBS technology as a key feature, none comprehensively addresses the three key components of this invention: information input, processing, and output. This distinctive integration underscores the originality of the proposal and its ability to overcome the limitations of existing solutions.

[0028] These differences demonstrate that, even considering the document closest to the prior art, it would not be possible for a person skilled in the art to develop a technology like the one presented here. Nor would it be possible by combining two or more relevant documents, due to the lack of a holistic approach like the one that characterizes this invention.

[0029] In addition to representing a significant advance in LIBS-based analysis, this technology offers remarkable advantages: it allows for more accurate analysis of irregular surface samples, significantly reduces analysis time, delivers more complete information with a limited margin of error, and integrates elemental and mineralogical species data linked to the composition and structure of the samples.

[0030] The versatility of this invention makes it applicable in various industrial sectors. However, it is particularly noteworthy in mining, metallurgy, geology, and similar fields, where the

[0031] The ability to obtain accurate and representative information in real time is crucial for the efficient characterization of materials in these sectors.

[0032] BRIEF DESCRIPTION OF THE INVENTION

[0033] As already mentioned, the present invention corresponds to an "Automated LIBS-based system for minerachemical, elemental and mineralogical analysis of irregular surfaces" (LIBS@IA), where a unique system for input, processing and output of information has been developed.

[0034] Regarding data input, the system features significant configuration improvements, specifically tailored for surface mapping and LIBS measurements on materials with irregular surfaces. These improvements include the ability to integrate and compare the acquired data with a mineralogical spectral database, thereby expanding the accuracy and scope of the analysis.

[0035] Regarding information processing, this encompasses data storage and integration, as well as elemental and mineralogical mapping based on artificial intelligence. This process includes the development of mathematical models that depend directly on the input data generated by the developed equipment, ensuring a more accurate and efficient analysis.

[0036] Regarding data output, the system generates detailed reports that provide information on the elemental and / or mineralogical composition of the samples, linked to specific characteristics of mineral structures or other similar materials. These reports are designed to be clear and precise, facilitating the interpretation of the results and their application in various industrial or research contexts.

[0037] BRIEF DESCRIPTION OF THE FIGURES

[0038] Figure 1 represents part of the information input system of the invention. Figure 2 shows the conceptual schematic of the LIBS-3D device.

[0039] Figure 3 shows the sequential operation of the system of the invention

[0040] Figure 4 shows the system's data analysis principle

[0041] Figure 5 shows a first set of images analyzed using the LIBS system

[0042] Figure 6 shows a second set of images analyzed using the LIBS system

[0043] DETAILED DESCRIPTION OF THE INVENTION

[0044] As already noted, the LIBS-based system of the invention comprises:

[0045] A) An information input system, configured for surface map acquisition and LIBS measurement, specifically for application on irregular surfaces. This process is complemented by the integration of a mineralogical spectral database for precise comparisons and analyses. The system includes:

[0046] Surface map acquisition: Allows capturing the topography and physical characteristics of the analyzed surfaces.

[0047] LIBS measurement, Generates elemental spectral information directly from the sample, providing its chemical composition.

[0048] Mineralogy spectra database, Provides a structured repository to compare the data obtained, optimizing the identification and characterization of the analyzed components.

[0049] B) An information processing system, configured to manage, integrate, and model the acquired data, ensuring efficient and accurate processing through the use of artificial intelligence (AI). This system combines information storage with the development of advanced mathematical models that depend directly on the data captured by the input system. The system's capabilities include:

[0050] Data storage: Securely and structurally preserves all captured information to ensure its availability and traceability.

[0051] AI-powered data integration and modeling: Uses advanced algorithms to merge different types of information and generate elemental and mineralogical maps, optimizing the analysis and interpretation of results.

[0052] C) An information output system, configured to generate detailed reports that provide accurate data on the elemental and / or mineralogical composition associated with specific characteristics of mineral structures or similar features. This system is optimized to communicate the results clearly and in a visually understandable manner. The system includes:

[0053] Integrated results, detailed information that combines elemental and mineralogical composition, correlated with specific sample structures.

[0054] Advanced visualization, quantitative graphic representations in 2D and 3D format, illustrating the elemental and mineralogical distribution of the analyzed sample, facilitating its interpretation and use.

[0055] The information input system (A), configured for the acquisition of surface maps and LIBS measurements, for application on irregular surfaces, which integrates the comparison with a database of mineralogical spectra, comprises the following components:

[0056] 1) High definition (HD) camera: captures optical images of the sample, which will serve as a reference to identify the scanning area or region of interest (ROI) to be scanned by the height sensor and the LIBS system.

[0057] 2) Laser distance sensor: generates a surface map in the form of a point cloud, with a spatial resolution of 200 pm and an acquisition speed of 1 cm / s. The laser pulse is directed at the sample at each scan point, according to the desired resolution, for the x and y spatial dimensions. Furthermore, this laser allows the optical focal distance for LIBS to be assigned and maintained at each point. This sensor operates in coordination with a translation base that enables three-dimensional scanning (xyz axes) of the surface.

[0058] 3) Pulsed laser emission source (1064 nm): Generates pulses of up to 300 mJ / pulse. The laser beam passes through two expanders (lenses L1 and L2), which amplify and concentrate the beam, and is then redirected vertically by a mirror MI towards a focusing lens L3, which concentrates the beam at a focal point F1 for sample ablation.

[0059] 4) Emission collection and focusing system: A parabolic mirror M2, acting as a collector, concentrates the emission generated by the ablation, which is directed to a lens L4. This lens L4 acts as a collector and allows the collection and reduction of the resulting radiation width to the next lens L5. This lens L5 acts as a concentrator and focuser, allowing the spectrum obtained to be adjusted to a focal length F2 at the input of an optical fiber for transport of the optical spectrum.

[0060] 105) Multichannel optical fiber: Composed of seven sub-fibers, for each channel of a UV-VIS spectrometer, for the reception and obtaining of spectral data allowing the acquisition and recording of data, using sampling control via NI LabView Data Analysis.

[0061] 6) Focusing and Collection System: Mounted on a vertical translation base, it allows for the measurement of multiple points on the sample using predefined trajectories (e.g., zigzag). Capable of taking N shots at M points, this system requires synchronization with a microcontroller unit (MCU) and periodic calibration using reference standards. The information processing system (B) is designed to manage, store, and integrate data, providing advanced elemental and mineralogical mapping using artificial intelligence (AI). This system incorporates specific mathematical models that depend on both the input information and the acquisition system parameters. Its key components are as follows:

[0062] 1) Mineralogy Spectra Database: Validated using internal rock samples chemically characterized by traditional analytical techniques, such as microscopy and X-ray diffraction. This database ensures accurate identification of the elements present in the analyzed samples. For example, the NIST-LIBS database is used as the primary reference, supplemented by academic studies to identify specific elements in selected samples.

[0063] 2) Local data storage and logging station: Ensures data security and availability through regular backups. Processing is performed in closed and reliable environments, using advanced tools such as MATLAB and Python IDEs.

[0064] 3) Data integration: The database is adapted to meet the requirements of the analysis platform. This process includes data correction, imputation of missing values, and accurate assignment of labels based on wavelength and detected elements.

[0065] 4) AI-based analysis: The system uses artificial intelligence algorithms to identify critical features in the data. This includes generating clusters, maximizing differences between classes, and detecting patterns that allow for the identification of key elements in the analyzed samples.

[0066] 11. The information output system (C) is designed to generate reports detailing the elemental and / or mineralogical composition, linked to the characteristics of mineral structures or similar materials. This system uses data matrices to record intensity measurements at different wavelengths, in TDMS (Technical Data Management Streaming) format. Example: characteristic spectral signatures of the elements present.

[0067] LIBS@IA system workflow:

[0068] The sequential process of the system is organized into the following stages:

[0069] 1) Sample positioning Alignment and fixing of the sample in the analysis device.

[0070] 2) HD Image Capture: Obtaining a detailed visual reference of the sample to determine its structure and area of ​​interest.

[0071] 3) Scanning area selection: Defining the region of interest (ROI) in the sample.

[0072] 4) 3D profile scanning with IR laser: Generation of a three-dimensional map by scanning the surface with an infrared laser.

[0073] 5) 3D point cloud capture: Creation of a three-dimensional model of the analyzed surface.

[0074] 6) LIBS Measurement: Application of laser ablation spectroscopy in the selected region to analyze its composition.

[0075] 7) LIBS data storage: Registration and management of the data obtained in a secure and structured environment.

[0076] 8) Data processing and modeling: Advanced analysis, including AI techniques, to interpret and model the collected data.

[0077] 9) 3D map visualization: Generation of quantitative three-dimensional representations that show the elemental and mineralogical distribution in the sample.

[0078] Regarding the system analysis principle of the invention (LIBS@IA), it is based on a structured approach to data analysis, distributed across the following key stages:

[0079] 1. 3D point cloud capture (stage 5):

[0080] o Configured to store point-by-point position (XY) and height (Z) information. o Enables accurate reconstruction of the three-dimensional profile of the scanned surface.

[0081] 2. LIBS measurement in the selected area (step 6):

[0082] 12Designed to perform two sequential laser pulses for each position analyzed. This improves the quality of the analysis by maximizing the consistency and accuracy of the data obtained.

[0083] 3. LIBS data storage (stage 7):

[0084] or Configured to capture spectral information at each point analyzed, associating it with XY coordinates.

[0085] or Generates an organized data matrix that facilitates subsequent analysis.

[0086] 4. Data processing and modeling (stage 8):

[0087] o It includes a sequential analysis divided into two processes:

[0088] a) Model creation:

[0089] x Background noise removal.

[0090] ■ Data normalization to ensure uniformity.

[0091] a Selection of the spectral range associated with elements and / or mineralogical phases of interest.

[0092] » Division of data into sets for calibration and validation.

[0093] or Validation of the model using specific merit figures.

[0094] b) Sample prediction:

[0095] KApplication of preprocessing techniques such as noise removal and normalization.

[0096] ■ Selection of the target mineralogical element or phase.

[0097] "Quantitative prediction of concentrations based on the model."

[0098] 5. 3D Map Visualization (Stage 9):

[0099] or Configured to reconstruct and present three-dimensional maps of elemental and / or mineralogical composition.

[0100] o It provides detailed and quantitative visual representations that facilitate interpretation and analysis.

[0101] CONCLUSION ON THE PRESENT INVENTION

[0102] As previously described, this invention represents a significant advancement compared to similar technologies, standing out for its improvements in information input, processing, and output. The LIBS (Laser-Induced Plasma Optical Emission Spectroscopy) based system incorporates key innovations, such as the ability to analyze irregular surfaces, which was not possible until now. Furthermore, it offers:

[0103] * Greater accuracy: Thanks to the integration of spectral data and artificial intelligence modeling.

[0104] "Reduced analysis time: Providing faster and more efficient results. * More complete and accurate information: Includes elemental and mineralogical composition linked to sample structures."

[0105] * Limited error margins: Optimization in model calibration and validation. The invention has significant application in the mining, metallurgical, and geological industries, enabling the rapid and representative characterization of samples associated with these sectors. Furthermore, its flexibility allows for implementation in other technical fields requiring surface analysis and characterization.

[0106] In summary, this technology not only overcomes the limitations of available tools, but also sets a new standard in terms of functionality, accuracy, and versatility for spectroscopic analysis across multiple industries.

[0107] EXAMPLE

[0108] The LIBS-based automated system for the mineralogical analysis of irregular surfaces has been tested on rock samples of varying composition, revealing mineralogical phases such as pyrite (FeS2), chalcopyrite (CuFeS2), quartz (SiO2), and bornite (Cu5FeS4), among others. These samples were previously characterized using classical reference techniques, such as inductively coupled plasma mass spectrometry (ICP-MS) and quantitative scanning electron microscopy (QEMSCAN), to determine the concentration of the elements and mineral phases present. These characterized samples are then used to calibrate AI-based models such as nearest neighbors (KNN) and artificial neural networks (ANN).Once the calibration models have been generated and validated, unknown samples are predicted with their corresponding associated error, comparing with the classical technique.

[0109] 14. Based on predictive models, the concentration of species of interest is estimated in a given area, generating distribution images in false-color scales according to the desired level of interest. Finally, based on the number of pixels associated with the different mineral phases present on the irregular surface scanned by the LIBS technique, the concentration of each of the chemical species of interest is estimated.

[0110] The generated result is delivered in the form of distribution maps for each elemental or mineralogical species of interest, compared with the optical image of the sample and the corresponding error associated with the measurement.

[0111] 15

Claims

CLAIMS 1. An automated LIBS-based system for the mineralogical, elemental, and mineral analysis of irregular surfaces, CHARACTERIZED in that it comprises: A) an information input system configured for the acquisition of surface maps and LIBS measurement, which integrates comparison with a database of mineralogical spectra, B) an information processing system configured for data storage, data integration, and elemental and mineralogical mapping by AI, C) an information output system configured to deliver reports that provide information regarding the elemental composition and / or molecules associated with characteristics of mineral structures or similar.

2. The system of claim 1, CHARACTERIZED in that the information input system (A) comprises: a pulsed laser where the light beam passes through a lens L1 that amplifies the beam, and then through a lens L2 that concentrates the beam, A mirror M1 concentrates and sends the beam vertically to a lens L3 that concentrates the optical beam at a specific point at a focal distance F1 from the sample being analyzed, generating its ablation. A parabolic mirror M2 concentrates the emission generated by the ablation, which is directed to a lens L4 that collects and reduces the width of the resulting radiation to a lens L5, allowing the spectrum obtained at the input of an optical fiber to be adjusted to a focal distance F2 for transporting the optical spectrum; where a focusing and light collection system mounted on a vertical translation base allows the measurement of different points on a sample, making it possible to perform N shots per M measurement points, which requires external synchronization to a microcontroller unit (MCU), and calibration processes for standards.

3. The system of claim 2, CHARACTERIZED in that the information input system (A) further comprises: a laser distance sensor; assigns a focal distance allowing a motion control system or translation base to establish a defined and constant height for a configured focal point.

4. The system of claim 2, CHARACTERIZED in that the optical fiber corresponds to a fiber composed of seven subfibers for each channel of a UV-VIS spectrometer, for the reception and obtaining of spectral data allowing the acquisition and recording of data, using sampling control via NI LabView Data Analysis.

5. The system of claim 1, CHARACTERIZED in that the information processing system (B) comprises the development of mathematical models dependent on the input information and the information input system.

6. The system of claim 1, CHARACTERIZED in that the information processing system (B) comprises: The database of mineralogical spectra, local and external functions, allows for the correct identification of different elements present in the analyzed samples; storage; the data is recorded at the local station. Integration: the database is modeled under parameters required by the analysis platform. AI analysis allows the recognition of relevant study features.

7. The system of claim 6. CHARACTERIZED in that the mineralogical spectral database uses the NIST LIBS database as the main reference and academic studies, for elements to be searched.

8. The system of claim 6, CHARACTERIZED in that storage involves making backup copies, works on closed programming environments for its processing (MATLAB, Python IDEs).

9. The system of claim 6, CHARACTERIZED in that the integration involves data correction, imputation and adaptation of labels by wavelength / elements.

10. The system of claim 6, CHARACTERIZED in that the AI ​​analysis involves different groupings, maximizing differences between classes to explore the presence of objective elements.

11. The system of claim 1, CHARACTERIZED in that the information output system (C) configured to deliver reports comprises spectrometer arrays for recording counts for different wavelengths, in TDMS format.

12. Procedure to be executed in the LIBS-based automated system for the mineralogical, elemental and mineralogical analysis of irregular surfaces as described in the preceding claims, CHARACTERIZED in that it comprises the following steps: a) positioning of the sample, b) HD image capture of the sample, c) selection of scanning area, d) 3D profile scanning with IR laser, e) 3D point cloud capture, f) LIBS measurement of the selected area, g) LIBS data storage, h) data processing and modeling, i) visualization of quantitative elemental and mineralogical 3D maps.

13. Method according to claim 12, CHARACTERIZED in that the 3D point cloud capture e), is configured for the storage of position (XY) and height (Z) information point by point, allowing the reconstruction of the 3D profile of the scanned surface.

14. Method according to claim 12, CHARACTERIZED in that the LIBS measurement of the selected area f), is configured to perform two laser pulses for each position sequentially.

15. Method according to claim 12, CHARACTERIZED in that the data LIBS storage g), is configured for the capture of spectral information for each point analyzed in XY coordinates and generation of a data matrix.

16. Method according to claim 12, CHARACTERIZED in that the data processing and modeling h) is configured to perform sequential data analysis for: a) model creation, b) sample prediction. 1817. Process according to claim 16, CHARACTERIZED in that the creation of models a) comprises: background noise removal standardization Selection of spectral range for each element and / or mineralogical phase of interest; division of data into calibration and validation validation of the model by merit figures.

18. Method according to claim 16, CHARACTERIZED in that the sample prediction b) comprises: background noise removal standardization selection of element and / or mineralogical phase of interest concentration prediction.

19. Method according to claim 12, CHARACTERIZED in that the visualization of quantitative elemental and / or mineralogical 3D maps i) is configured for the reconstruction and visualization of elemental and / or mineralogical 3D maps. 19

Images