Method for detecting and analyzing mining elements composition based on laser-induced breakdown spectroscopy

Abstract

This invention discloses a method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy, belonging to the technical field of laser-induced breakdown spectroscopy detection and analysis. The method includes: exciting multiple points on the sample surface and acquiring spectral signals and spatial coordinates at each point; constructing an elemental distribution probability map based on all point coordinates and signals; extracting the gradient field of signal intensity varying with coordinates from the spectrum, and marking regions with gradients exceeding a threshold as heterogeneous interfaces; performing layered virtual sectioning based on the heterogeneous interface distribution density to generate multiple independent sub-blocks; normalizing and superimposing all spectral signals within each block to obtain the block's characteristic spectrum; and matching the characteristic spectra of each block with a standard mineral characteristic spectral library to determine the dominant element type and relative content in each block. This invention enables refined and non-destructive analysis of the elemental distribution within mineral samples.

Description

Technical Field

[0001] This invention relates to the field of laser-induced breakdown spectroscopy detection and analysis technology, specifically to a method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy. Background Technology

[0002] In mineral elemental analysis, when using laser-induced breakdown spectroscopy (LAS-PS) to analyze ore samples, the conventional approach is to select a small number of discrete points on the sample surface for excitation, and then average or analyze the collected spectral signals individually to infer the elemental composition and content of the entire sample. However, because mineral samples generally exhibit uneven mineral distribution and random mixing of different mineral phases, the results from a small number of discrete points are insufficient to reflect the overall elemental distribution of the sample. This is especially true in ores with complex mineral coexistence relationships and where beneficial elements are present in a fine-grained disseminated manner; information obtained from single-point or random multi-point excitation has a significant elemental dependency.

[0003] Existing technologies have shortcomings in utilizing the spatial dimension of spectral signals. Spectral data acquired from multiple locations are isolated, failing to establish a statistical correlation between spectral features and actual spatial locations, thus losing information about the continuous distribution and abrupt changes in the spatial structure of elements on the sample surface. Sample analysis is often simplified to comparing the spectrum of a single location with a standard spectral library, lacking quantitative analysis of the spatial variations in the spectrum. This results in the inability to effectively identify boundaries between different mineral phases and to correlate elemental composition analysis with the inherent geological structural characteristics of the sample. When different minerals intertwine to form complex interface structures, the global average spectral matching method is highly susceptible to missing trace elements and inaccurate estimations of the relative abundance of major elements.

[0004] Mineral sample composition analysis actually faces two levels of technical challenges. The first is how to robustly extract probabilistic information that characterizes the continuity and abrupt changes in element distribution from a large number of spatially discrete spectral signals. The second is how to divide the sample into analytical units that conform to its inherent mineral pattern based on the structural characteristics of element distribution, so that subsequent spectral matching can be performed in independent blocks with mineralogical significance, thereby obtaining more accurate dominant element types and relative contents in a geological sense. Summary of the Invention

[0005] The technical problem to be solved by this invention is to overcome the shortcomings of existing laser-induced breakdown spectroscopy technology in the analysis of mineral samples, which has low utilization rate of spectral spatial structure information. This invention provides a method that can automatically identify heterogeneous interfaces on the sample surface and perform layered virtual sectioning of the sample based on the spatial statistics and gradient change characteristics of multi-point spectral signals. This allows the qualitative and quantitative analysis of elemental composition to be established on an analytical unit consistent with the internal mineral pattern of the sample, thereby obtaining information on the distribution and relative content of dominant elements that are closer to the actual occurrence state.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The present invention discloses a method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy. This method uses a laser-induced breakdown spectrometer to excite multiple points on the surface of a mineral sample, acquiring the plasma spectral signal corresponding to each point, and simultaneously recording the spatial coordinates of each point on the sample surface. As one technical solution of the present invention, the laser-induced breakdown spectrometer is controlled to move along a preset serpentine scanning path on the surface of the mineral sample, triggering laser pulses at fixed spatial intervals during the movement. Preferably, this fixed spatial interval is dynamically adjusted according to the expected mineral particle size of the mineral sample, ensuring that the distance between adjacent excitation points is less than the average diameter of the expected mineral particles, thereby ensuring a sufficient density of spectral sampling points for the small mineral particles. Each time a laser pulse is triggered, the spectrometer records the plasma emission spectral line generated at the current point, and simultaneously, a position sensor synchronized with the laser-induced breakdown spectrometer reads the three-dimensional spatial coordinates of the current point. After converting the plasma emission spectral line of each point into a digital spectral signal, it is associated and stored with the three-dimensional spatial coordinates of that point.

[0007] After obtaining the spatial coordinates and spectral signals of all points, an elemental distribution probability map of the sample surface is constructed based on the spatial coordinates of all points and the corresponding plasma spectral signals. During the construction process, for each preset target element, the characteristic spectral line intensity value corresponding to that target element is extracted from the digital spectral signal of each point. All points are projected onto a two-dimensional planar grid on the sample surface according to their three-dimensional spatial coordinates. Each grid cell contains the characteristic spectral line intensity values ​​of one or more points falling within that grid cell. For each grid cell, the statistical median of the characteristic spectral line intensity values ​​of all points falling within that grid cell is calculated, and this statistical median is used as the probability value of the target element's presence in that grid cell. For each grid cell, the standard deviation of the characteristic spectral line intensity values ​​of all points falling within that grid cell is also calculated. When the standard deviation exceeds a preset fluctuation threshold, the grid cell is marked as an intensity-unstable cell. When constructing the elemental distribution probability map, the probability value of the intensity-unstable cell is replaced with the arithmetic mean of the probability values ​​of all adjacent grid cells, thereby suppressing the interference of abnormal local signal fluctuations on the overall distribution judgment. For blank areas within each grid cell where no points were collected, inverse distance weighted interpolation is used to supplement the probability values ​​of that grid cell, thereby obtaining continuous and complete spatial distribution information of the elements. The probability values ​​of each grid cell are arranged according to their grid positions to generate a two-dimensional probability distribution matrix of the target element on the sample surface, i.e., the element distribution probability map.

[0008] The gradient field of element signal intensity as a function of spatial coordinates is extracted from the element probability distribution spectrum. Specifically, the horizontal and vertical differences in the two-dimensional probability distribution matrix of the element probability distribution spectrum are performed to obtain the intensity change rate in the horizontal and vertical directions for each grid cell. Then, based on the horizontal and vertical intensity change rates of each grid cell, the comprehensive gradient magnitude of that grid cell is calculated. Grid cells with a comprehensive gradient magnitude greater than a preset gradient threshold are marked as boundary candidate cells. Adjacent boundary candidate cells are connected to form a continuous region, which is then used as a heterogeneous interface region. This process can accurately identify the locations where abrupt changes occur in the composition or structure of different minerals in the sample.

[0009] Based on the distribution density of heterogeneous interface regions on the sample surface, the mining sample is virtually sectioned in layers to generate multiple independent sub-sample analysis blocks. Specifically, the total number of grid cells occupied by all heterogeneous interface regions on the sample surface is counted, and the area coverage density of the heterogeneous interface regions on the entire sample surface is calculated. When the area coverage density is greater than a first density threshold, it indicates that the sample has a complex internal structure and well-developed interfaces. In this case, the sample surface is divided into regions using the geometric center of each heterogeneous interface region as the dividing point. Each Thiessen polygon corresponds to a sub-sample analysis block, ensuring that the block boundaries closely match the natural heterogeneous interfaces. When the area coverage density is less than or equal to the first density threshold, it indicates that the sample is relatively homogeneous internally. In this case, the sample surface is divided into a predetermined number of square blocks using a uniform grid method, with each square block serving as a sub-sample analysis block. This adaptive sectioning strategy ensures high compositional homogeneity within the blocks to be analyzed subsequently.

[0010] The plasma spectral signals within each subsample analysis block are normalized and superimposed to obtain the block characteristic spectrum corresponding to each subsample analysis block. The processing method is as follows: for any subsample analysis block, the digital spectral signals of all points falling within the boundary of that subsample analysis block are extracted. Each extracted digital spectral signal is then normalized according to its wavelength channel, ensuring that the sum of the intensity values ​​of each digital spectral signal is a unit value. Preferably, during intensity normalization, the square root of the sum of the squares of the intensity values ​​of all wavelength channels in that spectral signal is used as the normalization benchmark to eliminate absolute intensity differences between different excitation points. Next, the intensity values ​​of all normalized digital spectral signals are accumulated on the same wavelength channel to obtain an accumulated spectral signal. This accumulated spectral signal is then normalized again to ensure that the sum of the intensity values ​​of the accumulated spectral signal is a unit value. This re-normalized accumulated spectral signal is used as the block characteristic spectrum corresponding to that subsample analysis block. After this double normalization and superposition process, the obtained block feature spectrum can effectively suppress random noise and highlight the common spectral features within the block.

[0011] Preferably, after obtaining the block characteristic spectrum corresponding to each sub-sample analysis block, the spectral angle cosine of the block characteristic spectrum corresponding to each sub-sample analysis block and the block characteristic spectrum corresponding to the adjacent sub-sample analysis blocks is calculated to obtain the spectral similarity between adjacent blocks. When the spectral similarity between adjacent blocks is higher than the similarity threshold, the adjacent blocks are merged into the same geological unit, and the boundaries of the sub-sample analysis blocks are adjusted according to the division results of all geological units to generate updated sub-sample analysis blocks for subsequent matching. Through this spectral similarity-driven block merging operation, homogeneous regions that were originally over-segmented can be re-aggregated, making the final analysis blocks more closely resemble the actual geological unit distribution.

[0012] The characteristic spectra of each block are matched with a standard mineral element characteristic spectral library to determine the dominant element species and their relative abundance in each subsample analysis block. During the matching process, multiple standard characteristic spectral lines and their standard intensity ratios for each standard mineral element are read from the standard mineral element characteristic spectral library. For any block's characteristic spectrum, the presence of a peak matching the position of a standard characteristic spectral line is detected at each wavelength position. When a matching peak is detected, its actual intensity value is recorded. Based on the actual intensity values ​​of multiple standard characteristic spectral lines for the same standard mineral element, the actual intensity ratio of that standard mineral element in the block's characteristic spectrum is calculated. The consistency ratio between the actual intensity ratio and the standard intensity ratio is checked. When the consistency deviation is less than a preset deviation threshold, the block is determined to contain that standard mineral element, and the actual intensity ratio is used as the relative abundance of that element. This multi-spectral line joint matching and intensity ratio consistency check effectively avoids element misidentification that may occur with single-spectral line matching, significantly improving the accuracy of qualitative and quantitative analysis of mineral elements.

[0013] After traversing all sub-sample analysis blocks and obtaining the dominant element types and their relative contents for each sub-sample analysis block, all sub-sample analysis blocks are spliced ​​together according to their spatial arrangement to generate a heat map of elemental composition distribution for the entire mineral sample. Connected regions with continuous distribution of the same element type and relative contents all above the content threshold are identified from the elemental composition heat map. These connected regions are marked as mineralized enrichment areas, providing a direct basis for subsequent mineral resource evaluation and target area delineation.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: The construction of the elemental probability distribution map uses the statistical median to characterize the intensity probability of the characteristic spectral lines of the target element within each spatial grid cell. Compared to directly using single-point intensity values ​​or arithmetic mean, the statistical median is insensitive to individual extreme intensity values ​​caused by microscopic unevenness of the sample surface, small fluctuations in laser energy, and mineral micro-regional compositional segregation, and can robustly reflect the probability of the element's presence in the local spatial region. Based on this, inverse distance weighted interpolation is applied to the blank areas in the grid, allowing continuous probability estimates to be obtained for locations that could not be directly sampled due to excitation point spacing and scanning path limitations. The resulting two-dimensional probability distribution matrix completely covers the entire analytical surface of the sample, preserving the spatial gradient and abrupt changes in elemental intensity, and avoiding breaks or false boundaries in subsequent gradient analysis due to sparse sampling points.

[0015] The gradient field is extracted from the elemental distribution probability map and the heterogeneous interface region is marked. The intensity change rate is obtained by performing differential operations on the horizontal and vertical directions on the spatial grid using probability values, thereby calculating the comprehensive gradient amplitude of each grid cell. This two-dimensional differential gradient extraction method is insensitive to direction and can capture boundaries between different mineral phases along arbitrary orientations, not just interfaces in a specific scanning direction. Grid cells exceeding the gradient threshold are connected into continuous regions, directly identifying mineral phase contact zones or microcrack filling zones on the sample surface indicated by elemental abrupt changes, thus replacing visual judgment or empirical partitioning with objective numerical criteria. Based on this, a region partitioning strategy is dynamically selected according to the distribution density of the heterogeneous interface region. When the density is high, Thiessen polygons are used with the geometric center of the heterogeneous interface as the seed point, causing the block boundaries to converge to the mineral phase abrupt change zone; when the density is low, uniform grid partitioning is used to ensure the regularity of the analysis cells. This division method ensures that the block characteristic spectra generated by subsequent normalization and overlay correspond to geological units with relatively uniform internal composition. During spectral matching, the dominant mineral elements and their proportions of the unit can be resolved from the overlay spectra. Finally, by combining the analysis results of all blocks according to their spatial location, the elemental composition distribution heatmap and the spatial range of the mineralized enrichment area of ​​the entire mineral sample can be directly presented. Attached Figure Description

[0016] Figure 1 This is a flowchart of a method for detecting and analyzing mineral element composition based on laser-induced breakdown spectroscopy.

[0017] Figure 2 This is a flowchart of a method for constructing elemental distribution probability maps based on laser-induced breakdown spectroscopy.

[0018] Figure 3 This is a flowchart of the process for extracting and analyzing heterogeneous interface regions and dividing blocks based on element distribution probability maps.

[0019] Figure 4 This is a flowchart of the process for generating characteristic spectra of subsample analysis blocks.

[0020] Figure 5 It is a two-dimensional elemental distribution probability profile of the target elements Fe and Cu in the mineral sample.

[0021] Figure 6 It is a schematic diagram of the comprehensive gradient magnitude and heterogeneous interface region of the elemental distribution probability map of a mining sample.

[0022] Figure 7 This is a schematic diagram of the relative content distribution of copper and the mineralization enrichment area.

[0023] Figure 8 This is a diagram showing the boundary and total spectral intensity distribution of the subsample analysis blocks before and after merging and adjustment. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] See Figure 1 This invention provides a method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy. In this method, a multi-point excitation of the mineral sample surface is first performed using a laser-induced breakdown spectrometer, acquiring the corresponding plasma spectral signal at each point and simultaneously recording the spatial coordinates of each point on the sample surface. Based on the acquired spatial coordinates and corresponding plasma spectral signals of all points, an elemental distribution probability map of the sample surface is constructed. The gradient field, showing the variation of elemental signal intensity with spatial coordinates, is extracted from this elemental distribution probability map, and regions with gradient values ​​exceeding a preset threshold are marked as heterogeneous interface regions. Then, based on the distribution density of heterogeneous interface regions on the sample surface, the mineral sample is virtually segmented into multiple independent sub-sample analysis blocks. All plasma spectral signals within each sub-sample analysis block are normalized and superimposed to obtain the block characteristic spectrum corresponding to each sub-sample analysis block. Finally, the characteristic spectrum of each block is matched with a standard mineral element characteristic spectral library to determine the dominant element species and their relative abundance in each sub-sample analysis block.

[0026] Example 1

[0027] See Figure 2 In practice, the laser-induced breakdown spectrometer (LASDS) moves along a preset serpentine scanning path across the surface of the mineral sample. During this movement, laser pulses are triggered at fixed spatial intervals for excitation. These fixed spatial intervals are dynamically adjusted based on the expected mineral particle size of the sample, ensuring that the distance between adjacent excitation points is less than the average diameter of the expected mineral particles. Each time a laser pulse is triggered, the LASDS records the plasma emission spectral line generated at the current point, while a position sensor synchronized with the LASDS reads the three-dimensional spatial coordinates of that point. After converting the plasma emission spectral line at each point into a digital spectral signal, the digital spectral signal is associated with and stored along with the corresponding three-dimensional spatial coordinates.

[0028] For each preset target element, the characteristic spectral line intensity value corresponding to the target element is extracted from the digital spectral signal at each point. The extraction method is as follows: locate the pre-defined characteristic spectral line wavelength position of the target element on the wavelength axis of the digital spectral signal, and read the spectral intensity value at that wavelength position as the characteristic spectral line intensity value of the target element at that point. The target element can be one or more types, and one or more characteristic spectral lines are pre-defined for each target element.

[0029] All points are projected onto a two-dimensional planar grid on the sample surface according to their corresponding three-dimensional spatial coordinates. The two-dimensional planar grid consists of multiple grid cells of the same size. During the projection process, the horizontal coordinates of each point are matched with the coordinates of the grid cells in the two-dimensional planar grid to determine the grid cell in which the point falls. For each grid cell, the characteristic spectral line intensity values ​​of the target element corresponding to all points falling within that grid cell are collected to form a set of intensity values. If no points fall within a certain grid cell, the grid cell temporarily does not have characteristic spectral line intensity values.

[0030] For each grid cell with a set of intensity values, the statistical median of all values ​​in the intensity value set is calculated, and this statistical median is used as the probability value of the target element's existence in the grid cell. Using the statistical median as the probability value can reduce the impact of abnormally high or low characteristic spectral line intensity values ​​in the intensity value set on the overall probability representation of the grid cell.

[0031] For each blank grid cell with no points falling into it, inverse distance weighted interpolation is used to supplement the probability value of the blank grid cell. Inverse distance weighted interpolation estimates the probability value of the blank grid cell using grid cells with existing probability values. A search radius is pre-defined, and the grid cells with known probability values ​​covered by the search radius are used as reference grid cells. The interpolated probability value of the blank grid cell is... The probability values ​​of the reference grid cells are calculated using a distance-inverse weighted average, as shown in the formula:

[0032] in, This represents the probability value of the blank grid cells to be interpolated and filled. This represents the total number of reference grid cells with known probability values ​​selected within the search radius; Indicates the first The probability value of each reference grid cell; Indicates the first Euclidean distance from the center point of a reference grid cell to the center point of a blank grid cell; Indicates the power exponent of distance decay. The value of is limited to positive integers, and is set during implementation. The value of is 2, which makes the probability weight of the reference grid cell decrease in an inverse square proportion as the distance increases.

[0033] After calculating the statistical median of each grid cell with a set of intensity values, the standard deviation of all values ​​in the intensity value set is also calculated. When the calculated standard deviation exceeds a preset fluctuation threshold, the grid cell is marked as an intensity-unstable cell. The preset fluctuation threshold is determined based on the range of spectral intensity fluctuations of the same type of target element in homogeneous samples from historical experimental data, taking into account the instrument noise level of the laser-induced breakdown spectrometer. When constructing the elemental distribution probability map of the target element, for grid cells marked as intensity-unstable cells, the probability value originally obtained from the statistical median of the grid cell is replaced with the arithmetic mean of the probability values ​​of all grid cells adjacent to the intensity-unstable cell. All adjacent grid cells refer to all grid cells in the two-dimensional planar grid that share at least one vertex or one edge with the intensity-unstable cell.

[0034] Finally, the probability values ​​of each grid cell are arranged according to the row and column positions of the two-dimensional planar grid to form a two-dimensional probability distribution matrix of the target element on the sample surface. This two-dimensional probability distribution matrix is ​​the element distribution probability map corresponding to the target element. All the above steps are performed for each preset target element to generate its corresponding element distribution probability map.

[0035] See Figure 5 In the figure, the horizontal axis represents the horizontal spatial coordinates in millimeters, ranging from 0 to 50; the vertical axis represents the vertical spatial coordinates in millimeters, also ranging from 0 to 50. The legend shows the probability distribution contours of two different target elements. The solid line represents the probability distribution contour of the target element Fe, and the dashed line represents the probability distribution contour of the target element Cu. Both are contour lines representing the two-dimensional elemental distribution probability map formed by collecting multi-point spectral signals of mineral samples, extracting characteristic spectral line intensities, and mapping them using a laser-induced breakdown spectrometer in Example 1.

[0036] As shown in the figure, the probability distribution profile of the target element Fe is located in the lower right region of the sample surface. The profile is a closed, concentric curve with distinct layers, indicating that Fe is densely distributed in the horizontal direction (20 mm to 35 mm) and the vertical direction (15 mm to 40 mm), with the probability gradually decreasing. This suggests that the Fe content in this region is high and exhibits a relatively continuous enrichment distribution. The uniform spacing between the layers indicates that the gradient of the Fe probability value changes relatively gently.

[0037] The probability distribution profile of the target element Cu is represented by dashed lines, showing two distinct and independent enrichment regions: one located in the upper left region (approximately 15 mm horizontally and 35 mm vertically), and the other in the lower right region (approximately 35 mm horizontally and 15 mm vertically). Both Cu enrichment regions exhibit closed circular outlines with few and relatively dispersed outline layers, indicating that the Cu element distribution on the sample surface exhibits spatial discreteness, and that there is a clear spatial separation between the two enrichment regions.

[0038] As described in Example 1, the two-dimensional probability distribution profile used in the figure reflects the probability of the target element's presence obtained by statistically analyzing the median intensity values ​​of characteristic spectral lines at each point and interpolating them. The continuity and hierarchy of the Fe element distribution profile indicate that the element has a stable and concentrated distribution area on the sample surface. The two distinct partitions of the Cu element distribution correspond to the spatial heterogeneity of the Cu element distribution in the sample, reflecting the heterogeneous interface and local enrichment characteristics of different elements in the sample.

[0039] This figure effectively illustrates the spatial probability distribution characteristics of two key target elements on the sample surface, providing basic data support for extracting gradient fields to identify heterogeneous interfaces and performing zoning analysis in subsequent embodiments.

[0040] Example 2

[0041] See Figure 3 In specific implementation, the gradient field of element signal intensity as a function of spatial coordinates is extracted from the element probability distribution map. The element probability distribution map is a two-dimensional probability distribution matrix corresponding to each target element obtained in the above embodiment. The row and column directions of the two-dimensional probability distribution matrix correspond to two orthogonal spatial dimensions of the sample surface, respectively. A horizontal difference operation is performed on the two-dimensional probability distribution matrix to obtain the intensity change rate of each grid cell in the horizontal direction. The horizontal difference operation is performed as follows: for the grid cell located in row u and column v of the two-dimensional probability distribution matrix, the probability value of the grid cell located in row u and column v+1 is subtracted from the probability value of the grid cell located in row u and column v, and the difference is used as the intensity change rate of the grid cell in the horizontal direction; for the grid cell in the rightmost column, its intensity change rate in the horizontal direction is set to zero. A vertical difference operation is performed on the two-dimensional probability distribution matrix to obtain the intensity change rate of each grid cell in the vertical direction. The vertical difference operation is performed as follows: For the grid cell located in the u-th row and v-th column in the two-dimensional probability distribution matrix, the probability value of the grid cell located in the (u+1)-th row and v-th column is subtracted from the probability value of the grid cell located in the u-th row and v-th column. The difference is used as the intensity change rate of the grid cell in the vertical direction. For the grid cells in the bottom row, the intensity change rate in the vertical direction is set to zero.

[0042] The combined gradient magnitude of each grid cell is calculated based on its horizontal and vertical intensity variation rates. The formula for calculating the combined gradient magnitude is:

[0043] in, This represents the combined gradient magnitude of the grid cell in the u-th row and v-th column of the two-dimensional probability distribution matrix; The horizontal intensity variation rate of the grid cell in row u and column v is obtained from the aforementioned horizontal differential operation; The vertical intensity variation rate of the grid cell in row u and column v is obtained by the aforementioned vertical difference operation.

[0044] After calculating the comprehensive gradient magnitude of all grid cells, the comprehensive gradient magnitude of each grid cell is compared with a preset gradient threshold. The preset gradient threshold is determined by collecting a two-dimensional probability distribution matrix obtained from testing on a standard sample with a known homogeneous mineral composition, calculating the average and standard deviation of the comprehensive gradient magnitude of all grid cells corresponding to the standard sample, and setting the preset gradient threshold as the average of the comprehensive gradient magnitude of the standard sample plus three times the standard deviation of the comprehensive gradient magnitude of the standard sample. Grid cells with a comprehensive gradient magnitude greater than the preset gradient threshold are marked as boundary candidate cells. After traversing all grid cells and completing the marking, adjacent boundary candidate cells are connected to form continuous regions. The criterion for adjacency is that grid cells share at least one edge. Each continuous region obtained by connection is treated as a heterogeneous interface region. When there is an isolated boundary candidate cell that cannot be connected to other boundary candidate cells, the isolated boundary candidate cell is treated as a separate heterogeneous interface region.

[0045] After obtaining all heterogeneous interface regions on the sample surface, the total number of mesh cells occupied by all heterogeneous interface regions is counted. The area coverage density of the heterogeneous interface regions on the entire sample surface is then calculated. The area coverage density is calculated by dividing the total number of mesh cells by the total number of mesh cells in the two-dimensional planar mesh on the sample surface; the resulting ratio is the area coverage density.

[0046] The area coverage density is compared with a first density threshold. The first density threshold is a preset value, set as follows: based on statistical analysis of the heterogeneous interface distribution characteristics measured in a large number of mining samples, density boundary points that can distinguish between densely interwoven and sparsely isolated heterogeneous interfaces in the samples are selected. These density boundary points are used as the first density threshold, which is set to 0.3 during implementation. When the area coverage density is greater than the first density threshold, it indicates a dense distribution of heterogeneous interfaces. In this case, the geometric center of each heterogeneous interface region is used as the dividing point, and the Thiessen polygon algorithm is used to divide the sample surface into regions. The Thiessen polygon algorithm is executed as follows: the geometric center points of all heterogeneous interface regions are used as the generation point set, and the two-dimensional plane of the sample surface is convexly hulled, such that the distance from any position within each polygon region to its own generation point is less than the distance to other generation points. Each Thiessen polygon corresponds to a sub-sample analysis block. When the area coverage density is less than or equal to the first density threshold, it indicates that the heterogeneous interface is sparsely distributed. In this case, the sample surface is divided into a predetermined number of square blocks using a uniform grid method, with each square block serving as a sub-sample analysis block. The predetermined number is set in advance based on the actual sample size and analytical accuracy requirements. During implementation, the sample surface is uniformly divided into 8 rows × 8 columns, totaling 64 square blocks.

[0047] See Figure 6 The figure shows the comprehensive gradient amplitude spectrum of the two-dimensional spatial distribution of a target element on the surface of a mining sample, obtained based on laser-induced breakdown spectroscopy. The horizontal axis represents the horizontal spatial coordinates of the sample surface, ranging from 0 to 50 mm; the vertical axis represents the vertical spatial coordinates, also ranging from 0 to 50 mm. The grayscale image reflects the magnitude of the comprehensive gradient amplitude; the lighter the grayscale, the higher the gradient amplitude, indicating a more dramatic change in the element's signal intensity across its spatial location.

[0048] The boundary of the heterogeneous interface region is marked by a black dashed line (dotted line) in the figure. This boundary corresponds to a continuous region where the overall gradient amplitude exceeds a preset gradient threshold. The heterogeneous interface region is mainly concentrated in a spatial range of approximately 20 mm to 35 mm horizontally and 15 mm to 35 mm vertically. This region exhibits a high gradient amplitude, indicating significant variations in element signal intensity and obvious interface heterogeneity. The gradient amplitude decreases rapidly at the periphery of the heterogeneous interface region, transitioning to a homogeneous region with a lower gradient.

[0049] Furthermore, the boundaries of the sub-sample analysis blocks are marked with thin solid lines in the figure, and the blocks are divided into several uniform square grids. The sub-sample analysis blocks cover the entire sample surface, and the boundary spacing is uniformly distributed, which facilitates the subsequent normalization, overlay, and analysis of the spectral data within each block. This uniform division method conforms to the fixed grid division strategy used in Example 2 when the heterogeneous interface area coverage density is less than a preset first density threshold.

[0050] In summary, this figure clearly reflects the comprehensive gradient amplitude field extracted from the element distribution probability spectrum, the heterogeneous interface region is determined and marked by threshold, and the specific implementation steps of dividing the sub-sample analysis blocks using a uniform grid in combination with area coverage density determination are verified. This verifies the effectiveness and implementation effect of the technical solution for identifying heterogeneous interfaces and dividing blocks on the surface of mining samples in Example 2.

[0051] Example 3

[0052] See Figure 4 In specific implementation, for each sub-sample analysis block generated by the steps of the above embodiment, digital spectral signals of all points falling within the boundary range of the sub-sample analysis block are extracted. Each point's digital spectral signal is a spectral data point containing multiple wavelength channels; the same wavelength position in the digital spectral signals of different points corresponds to the same wavelength channel. The intensity values ​​of all extracted digital spectral signals are normalized one by one. The benchmark for intensity value normalization is the square root of the sum of the squares of the intensity values ​​of all wavelength channels in the processed digital spectral signal. The specific method of intensity value normalization is as follows: select a digital spectral signal, square the intensity value of each wavelength channel in the selected digital spectral signal, sum the squares of the intensity values ​​of all wavelength channels to obtain the sum of squares, and take the square root of the sum of squares. The result is used as the normalization benchmark value of the selected digital spectral signal. The formula for calculating the normalization benchmark value is:

[0053] in, This represents the normalized reference value corresponding to the s-th digital spectral signal. The sequence number of the digital spectral signal; This indicates the s-th digital spectral signal in the wavelength channel. The original intensity value at the location; This represents the t-th wavelength channel of the s-th digital spectral signal, where t ranges from 1 to... ; This represents the total number of wavelength channels contained in the s-th digital spectral signal. It involves processing the s-th digital spectral signal in each wavelength channel. Original intensity value at the location Divide by the normalized baseline value The s-th digital spectral signal in the wavelength channel is obtained. The normalized intensity value at point s is adjusted so that the sum of the intensity values ​​of the s-th digital spectral signal after normalization is adjusted to a unit value.

[0054] For all normalized digital spectral signals belonging to the same subsample analysis block, the intensity values ​​are accumulated on the same wavelength channel. The same wavelength channel refers to a channel with the same wavelength value. The accumulation process is as follows: A fixed wavelength channel is used, and the normalized intensity values ​​of each normalized digital spectral signal within the subsample analysis block are added to that wavelength channel to obtain the accumulated intensity value. This accumulation operation is performed on all wavelength channels covered by the digital spectral signal to obtain a single accumulated spectral signal. The accumulated spectral signal corresponds to an accumulated intensity value on each wavelength channel.

[0055] The accumulated spectral signal is then subjected to overall intensity normalization again. The overall intensity normalization process is as follows: the sum of squares of the accumulated intensity values ​​for all wavelength channels in the accumulated spectral signal is calculated, and the square root of the sum is taken to obtain a secondary normalization reference value. The accumulated intensity value for each wavelength channel is then divided by the secondary normalization reference value to obtain the renormalized intensity value. The spectral data comprising the intensity values ​​of all wavelength channels after renormalization is used as the block characteristic spectrum corresponding to the subsample analysis block.

[0056] Example 4

[0057] In practice, a standard mineral element characteristic spectral library is pre-constructed. This library stores multiple standard characteristic spectral lines and their standard intensity ratios for various known standard mineral elements. Each standard mineral element corresponds to an element identifier, which is associated with the wavelength positions of a set of standard characteristic spectral lines and the corresponding standard relative intensities. The standard intensity ratios are determined by the ratios between the standard relative intensities of multiple standard characteristic spectral lines for the same standard mineral element.

[0058] For each subsample analysis block corresponding to the block characteristic spectrum obtained by the steps of the above embodiment, the block characteristic spectrum is matched one by one with each standard mineral element in the standard mineral element characteristic spectral library. The matching process performs the following operations: Select a standard mineral element from the standard mineral element characteristic spectral library, and read the wavelength positions of all standard characteristic spectral lines associated with the selected standard mineral element. On the block characteristic spectrum, detect whether there is a spectral peak that matches the wavelength position of the standard characteristic spectral line at each wavelength position. The wavelength position matching is determined as follows: when the normalized intensity value of the block characteristic spectrum at a certain wavelength position is simultaneously greater than the normalized intensity values ​​of the preceding and following wavelength positions adjacent to that wavelength position, and the absolute value of the wavelength deviation between that wavelength position and the wavelength position of the standard characteristic spectral line is less than the preset wavelength tolerance, it is determined that there is a matching spectral peak at the wavelength position of the standard characteristic spectral line of the selected standard mineral element in the block characteristic spectrum. The preset wavelength tolerance is set to three times the wavelength resolution of the laser-induced breakdown spectrometer.

[0059] When a matching peak is detected, the actual intensity value of the matching peak in the block's characteristic spectrum is recorded. The actual intensity value is the normalized intensity value of the matching peak. After obtaining the actual intensity values ​​of all standard characteristic lines that can match the peaks among all the standard characteristic lines corresponding to the selected standard mineral element, the actual intensity ratio of the standard mineral element in the block's characteristic spectrum is calculated based on the actual intensity values ​​of multiple matched standard characteristic lines of the same standard mineral element. The actual intensity ratio is calculated as follows: the standard characteristic line with the largest relative intensity among the multiple matched standard characteristic lines of the standard mineral element is selected as the reference line. The actual intensity value corresponding to each matched standard characteristic line is compared with the actual intensity value corresponding to the reference line to obtain a set of actual intensity ratio data.

[0060] The calculated actual strength ratio is then compared with the standard strength ratio stored in association with standard mineral elements to perform a consistency check. The consistency check method is as follows:

[0061] in, Indicates the consistency deviation value; This indicates the total number of standard characteristic spectral lines that have been matched among the selected standard mineral elements. The value of is an integer greater than or equal to 2; Indicates the sequence number of the matched standard characteristic spectral lines. The value ranges from 1 to ; Indicates the first The actual intensity value corresponding to the matched standard characteristic spectral lines; This represents the actual intensity value corresponding to the reference spectral line; Indicates the first The standard relative intensity corresponding to the matched standard characteristic spectral lines; This represents the standard relative intensity corresponding to the reference spectral line. When the calculated consistency deviation value... If the intensity is less than the preset deviation threshold, the selected standard mineral element is determined to exist in the subsample analysis block, and the actual intensity ratio is used as the relative content of the standard mineral element in the subsample analysis block. The preset deviation threshold is determined based on experimental repeatability verification. During implementation, the preset deviation threshold is set to 0.15, which corresponds to an average deviation of no more than 15% between the actual intensity ratio and the standard intensity ratio of each spectral line.

[0062] The process involves iterating through all subsample analysis blocks and performing the matching process described above for each subsample analysis block to obtain the dominant element type and its relative content for each subsample analysis block. The dominant element type is determined by the standard mineral elements that are successfully matched in the characteristic spectrum of the block. When the characteristic spectrum of a block matches multiple standard mineral elements, all of these standard mineral elements are used as the dominant element type of the subsample analysis block, each corresponding to its respective relative content.

[0063] After all sub-sample analysis blocks have been matched, they are stitched together according to their spatial arrangement on the sample surface. The spatial arrangement is determined by the grid position or Thiessen polygon position of the sub-sample analysis blocks during the division process in the aforementioned embodiment. During stitching, the dominant element type and its relative content for each sub-sample analysis block are mapped onto the spatial area occupied by that block, generating a heat map of elemental composition distribution for the entire mineral sample. The heat map uses different colors or shades of color to represent the levels of different dominant element types and their relative contents.

[0064] Connected regions with continuous distribution of the same element and relative abundances exceeding the content threshold are identified from the elemental composition heatmap. Continuous distribution of the same element means that the same standard mineral element is matched in adjacent sub-sample analysis blocks on the heatmap space. The content threshold is pre-set based on the economic boundary grade of mining operations. The identified connected regions are marked as mineralized enrichment areas.

[0065] See Figure 7The figure shows a heat map of elemental composition distribution obtained based on the method for detecting and analyzing the elemental composition of mineral samples described in Example 4. The horizontal axis represents the horizontal spatial coordinates of the sample surface, in millimeters, ranging from 0 to 50 mm; the vertical axis represents the vertical spatial coordinates, in millimeters, ranging from 0 to 50 mm. The grayscale values ​​in the figure represent the relative content of copper (Cu). Brighter grayscale values ​​indicate higher Cu content, and darker grayscale values ​​indicate lower Cu content. The legend is located above the graph, indicating the "relative Cu content" corresponding to the gray blocks and the "mineralized enrichment zone" corresponding to the dashed boxes.

[0066] Two distinct high-grayscale regions are visible in the figure, located at approximately 16 mm, 35 mm and 34 mm, 17 mm respectively, indicating that the relative Cu content in these two spatial regions is significantly higher than in other regions. Each high-grayscale region is enclosed by a dashed box, representing a mineralized enrichment zone identified in Example 4 by matching the block characteristic spectrum with the standard mineral element characteristic spectral library. The boundaries of the two mineralized enrichment zones are approximately circular, demonstrating the connectivity and relative uniformity of Cu distribution within the regions.

[0067] The remaining areas in the graph have lower gray levels, indicating that the Cu content is low or close to the background level. This heatmap was generated by stitching together the dominant element types and their relative contents from multiple sub-sample analysis blocks as described in Example 4. It reflects the spatial distribution characteristics of Cu on the surface of the mining sample and can intuitively identify mineralized enrichment areas where the content is higher than a set content threshold.

[0068] Overall, the heat map clearly shows the spatial non-uniformity of Cu distribution and its enrichment areas on the sample surface, providing important spatial information on elemental composition for subsequent mineral exploration and resource assessment.

[0069] Example 5

[0070] In specific implementation, after obtaining the block characteristic spectrum corresponding to each sub-sample analysis block through the steps of the above embodiment, the spectral similarity calculation and block merging adjustment operation of adjacent sub-sample analysis blocks are performed. For any two spatially adjacent sub-sample analysis blocks, one sub-sample analysis block is designated as the first sub-sample analysis block, and the other is designated as the second sub-sample analysis block. The block characteristic spectrum corresponding to the first sub-sample analysis block is designated as the first block characteristic spectrum, and the block characteristic spectrum corresponding to the second sub-sample analysis block is designated as the second block characteristic spectrum. Both the first and second block characteristic spectra are normalized spectral vectors, the dimension of which is equal to the total number of wavelength channels, and each element in the spectral vector corresponds to the normalized intensity value at a wavelength channel.

[0071] The cosine of the spectral angle between the characteristic spectra of the first block and the second block is calculated as the spectral similarity between adjacent blocks. The formula for calculating the cosine of the spectral angle is:

[0072] in, This represents the cosine value of the spectral angle between the characteristic spectra of the first block and the characteristic spectra of the second block. The value of the cosine value ranges from 0 to 1. The closer the value is to 1, the smaller the angle between the two spectral vectors and the higher the similarity. This indicates the total number of wavelength channels in the characteristic spectrum of the block; The first in the characteristic spectrum of the block One wavelength channel, The value ranges from 1 to ; This indicates the characteristic spectrum of the first block in the wavelength channel. The normalized intensity value at that location; This indicates the characteristic spectrum of the second block in the wavelength channel. The normalized intensity value at that location.

[0073] After calculating the spectral similarity between adjacent blocks, the spectral similarity is compared with a pre-set similarity threshold. The similarity threshold is determined as follows: Block characteristic spectra of multiple sub-sample analysis blocks obtained using the same analytical procedure on homogeneous samples with a known uniform single mineral composition are collected. The average value of the spectral angle cosines between each pair of these block characteristic spectra is calculated, and the average value minus twice the standard deviation is used as the similarity threshold. When the spectral similarity between adjacent blocks is higher than the similarity threshold, it indicates that the first and second sub-sample analysis blocks have a high degree of consistency in elemental composition characteristics. In this case, the first and second sub-sample analysis blocks are merged into the same geological unit.

[0074] For all adjacent sub-sample analysis blocks, spectral angle cosine calculation and similarity threshold comparison are performed sequentially to determine the merging of each pair of adjacent sub-sample analysis blocks. During the merging process, a regional growth merging method is adopted: a sub-sample analysis block not yet assigned to any geological unit is selected as the starting block. The spectral similarity between the starting block and each spatially adjacent sub-sample analysis block is checked. Adjacent sub-sample analysis blocks with spectral similarity higher than the similarity threshold are merged into the geological unit where the starting block is located. Then, using the newly merged sub-sample analysis block as a benchmark, the adjacent sub-sample analysis blocks of the newly merged sub-sample analysis block are checked again, and the spectral similarity comparison and merging operation is repeated until no new adjacent sub-sample analysis blocks meet the merging conditions. Subsequently, the next unassigned sub-sample analysis block is selected as the new starting block, and the above merging process is repeated until all sub-sample analysis blocks are assigned to their respective geological units.

[0075] After all sub-sample analysis blocks have been divided into geological units, the boundaries of the sub-sample analysis blocks are adjusted according to the division results. The boundary adjustment method is as follows: the internal boundaries between multiple original sub-sample analysis blocks belonging to the same geological unit and adjacent to each other are eliminated, and the outer boundaries of all original sub-sample analysis blocks belonging to the same geological unit are taken as the boundaries of the updated sub-sample analysis blocks, generating updated sub-sample analysis blocks. The updated sub-sample analysis blocks are used for subsequent matching operations with the standard mineral element characteristic spectral library. The matching operation adopts the matching and content determination method described in the above embodiment.

[0076] See Figure 8 The figure shows the two-dimensional distribution of total spectral intensity on the surface of a mining sample within a spatial coordinate range of 0 to 50 mm in both the horizontal and vertical directions. The legend uses shades of gray to represent the total spectral intensity at different locations; brighter gray levels indicate higher spectral intensity, reflecting the relative abundance of elemental components in that region of the sample. Overall, the central region of the sample (approximately 20 to 35 mm horizontally and 15 to 30 mm vertically) exhibits a distinct spectral intensity peak, forming a clear high-intensity cluster, while the spectral intensity at the edges is relatively lower, appearing as darker gray levels.

[0077] The figure shows the "original subsample analysis block boundaries" marked with black dashed lines. These boundaries are represented by a uniformly divided grid, forming boundary lines at approximately 12.5 mm, 25 mm, and 37.5 mm horizontally and vertically, dividing the entire sample surface into several uniform subsample analysis blocks. This division corresponds to the subsample analysis blocks initially generated in the above embodiment.

[0078] Furthermore, the "updated subsample analysis block boundary" is marked by a thick solid line alternating with black dashed lines and dotted lines in the figure, located at approximately 25 mm horizontally and 25 mm vertically. The original block boundaries were merged and adjusted, forming a larger and more irregular analysis block. This boundary adjustment reflects the spectral similarity calculation results based on adjacent subsample analysis blocks in Example 5. By merging adjacent blocks with spectral similarity exceeding a preset threshold, regional growth and merging of multiple original subsample analysis blocks were achieved, forming a new geological unit boundary.

[0079] As shown in the figure, the updated sub-sample analysis block boundaries reduce the number of blocks compared to the original boundaries. Particularly in the high-spectral-intensity block at the center of the sample, the previously subdivided sub-blocks have been merged into a larger block, reflecting a high degree of consistency in spectral characteristics within this region. The adjusted boundary division more closely reflects the true spatial distribution of the sample's elemental composition, facilitating more accurate spectral matching of mineral element characteristics and determination of relative content.

[0080] In summary, the data illustrated in the figure demonstrates the technical process described in Example 5, which involves using laser-induced breakdown spectral data to calculate the spectral angle cosine similarity between adjacent sub-sample analysis blocks, and then performing regional growth merging and boundary adjustment. This effectively optimizes the spatial division of sub-sample analysis blocks and improves the spatial accuracy and continuity of elemental composition analysis of mineral samples.

[0081] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy, characterized in that, Includes the following steps: The surface of the mining sample was excited at multiple points using a laser-induced breakdown spectrometer, and the plasma spectral signal corresponding to each point was collected. The spatial coordinates of each point on the sample surface were recorded simultaneously. Based on the spatial coordinates of all points and the corresponding plasma spectral signals, an elemental distribution probability map of the sample surface is constructed. The gradient field of element signal intensity as a function of spatial coordinates is extracted from the element distribution probability map, and the region in the gradient field whose gradient value exceeds a preset threshold is marked as a heterogeneous interface region. Based on the distribution density of heterogeneous interface regions on the sample surface, the mining sample is subjected to layered virtual sectioning to generate multiple independent sub-sample analysis blocks. The plasma spectral signals within each subsample analysis block are normalized and superimposed to obtain the block characteristic spectrum corresponding to each subsample analysis block. The characteristic spectra of each block were matched with the standard mineral element characteristic spectral library to determine the dominant element species and their relative abundance in each subsample analysis block.

2. The method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy according to claim 1, characterized in that, The specific method for exciting multiple points on the surface of mining samples using laser-induced breakdown spectroscopy, acquiring the plasma spectral signal corresponding to each point, and simultaneously recording the spatial coordinates of each point on the sample surface is as follows: The laser-induced breakdown spectrometer is controlled to move along a preset serpentine scanning path on the surface of the mineral sample, and laser pulses are triggered at fixed spatial intervals during the movement to excite it. Each time a laser pulse is triggered, the spectrometer records the plasma emission spectral lines generated at the current point, while the position sensor synchronized with the laser-induced breakdown spectrometer reads the three-dimensional spatial coordinates of the current point. After converting the plasma emission spectral lines at each point into digital spectral signals, they are associated with and stored in relation to the three-dimensional spatial coordinates of that point.

3. The method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy according to claim 2, characterized in that, The fixed spatial interval is dynamically adjusted according to the expected mineral particle size of the mining sample, so that the distance between adjacent excitation points is less than the average diameter of the expected mineral particles.

4. The method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy according to claim 2, characterized in that, The specific method for constructing the elemental distribution probability map of the sample surface based on the spatial coordinates of all points and the corresponding plasma spectral signals is as follows: For each preset target element, extract the characteristic spectral line intensity value corresponding to the target element from the digital spectral signal of each point; All points are projected onto a two-dimensional planar grid on the sample surface according to their three-dimensional spatial coordinates. Each grid cell contains the characteristic spectral line intensity values ​​of one or more points falling within that grid cell. For each grid cell, calculate the statistical median of the characteristic spectral line intensity values ​​of all points falling within that grid cell, and use this statistical median as the probability value of the existence of the target element in that grid cell; The probability values ​​of each grid cell are arranged according to their grid positions to generate a two-dimensional probability distribution matrix of the target element on the sample surface, which serves as the element distribution probability map.

5. The method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy according to claim 4, characterized in that, The specific method for extracting the gradient field of element signal intensity as a function of spatial coordinates from the element distribution probability map and marking regions in the gradient field whose gradient values ​​exceed a preset threshold as heterogeneous interface regions is as follows: By performing horizontal and vertical difference operations on the two-dimensional probability distribution matrix in the element distribution probability map, the intensity change rate in the horizontal direction and the intensity change rate in the vertical direction of each grid cell are obtained. Calculate the comprehensive gradient magnitude of each grid cell based on the horizontal and vertical intensity change rates of each grid cell. Mesh cells with a combined gradient magnitude greater than a preset gradient threshold are marked as boundary candidate cells. Adjacent boundary candidate cells are connected to form a continuous region, which is then used as a heterogeneous interface region.

6. The method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy according to claim 5, characterized in that, Based on the distribution density of heterogeneous interface regions on the sample surface, the specific method for performing layered virtual sectioning of the mining sample to generate multiple independent sub-sample analysis blocks is as follows: The total number of grid cells occupied by all heterogeneous interface regions on the sample surface is counted, and the area coverage density of heterogeneous interface regions on the entire sample surface is calculated. When the area coverage density is greater than the first density threshold, the sample surface is divided into regions using the geometric center of each heterogeneous interface region as the dividing point, and each Thiessen polygon corresponds to a sub-sample analysis block. When the area coverage density is less than or equal to the first density threshold, the sample surface is divided into a predetermined number of square blocks in a uniform grid manner, and each square block is used as a sub-sample analysis block.

7. The method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy according to claim 6, characterized in that, The specific method for obtaining the block characteristic spectrum of each subsample analysis block by normalizing and superimposing all plasma spectral signals within each subsample analysis block is as follows: For any sub-sample analysis block, extract the digital spectral signals of all points falling within the boundary of the sub-sample analysis block; Each extracted digital spectral signal is normalized according to its wavelength channel, so that the sum of the intensity values ​​of each digital spectral signal is a unit value. The intensity values ​​of all normalized digital spectral signals are accumulated on the same wavelength channel to obtain the accumulated spectral signal. The accumulated spectral signal is normalized again to make the sum of the intensity values ​​of the accumulated spectral signal a unit value. The normalized accumulated spectral signal is then used as the block characteristic spectrum corresponding to the analysis block of the subsample.

8. The method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy according to claim 7, characterized in that, When normalizing the intensity values ​​of a digital spectral signal, the square root of the sum of the squares of the intensity values ​​of all wavelength channels in the spectral signal is used as the normalization reference.

9. The method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy according to claim 7, characterized in that, The specific method for matching the characteristic spectra of each block with the standard mineral element characteristic spectral library to determine the dominant element types and their relative abundances in each sub-sample analysis block is as follows: Read multiple standard characteristic spectral lines and their standard intensity ratios for each standard mineral element from the standard mineral element characteristic spectral library; For any block of characteristic spectrum, detect whether there is a spectral peak that matches the position of the standard characteristic spectral line at each wavelength position. When a matching spectral peak is detected, record the actual intensity value of the spectral peak. Calculate the actual intensity ratio of the standard mineral element in the characteristic spectrum of the block based on the actual intensity values ​​of multiple standard characteristic spectral lines of the same standard mineral element. The consistency of the actual strength ratio and the standard strength ratio is checked. When the consistency deviation is less than the preset deviation threshold, it is determined that the standard mineral element exists in the block, and the actual strength ratio is used as the relative content of the element.

10. The method for detecting and analyzing mineral elemental composition based on laser-induced breakdown spectroscopy according to claim 9, characterized in that, After determining that the standard mineral element exists in the block and using the actual intensity ratio as the relative content of the element, perform the following operations: After traversing all subsample analysis blocks, the dominant element types and their relative contents for each subsample analysis block are obtained; All sub-sample analysis blocks are spliced ​​together according to their spatial arrangement to generate a heat map of the elemental composition distribution of the entire mineral sample. Identify connected regions in the elemental composition distribution heatmap where the same element is continuously distributed and the relative content is higher than the content threshold, and mark these connected regions as mineralized enrichment areas.

Images