Underwater ore in-situ detection method and system based on combination of LIBS and pulse Raman

Abstract

The application relates to the technical field of underwater ore detection, and particularly provides an underwater ore in-situ detection method and system based on LIBS and pulse Raman combination, the method comprising the following steps: starting a LIBS-pulse Raman combined spectrum acquisition device, acquiring LIBS spectrum and pulse Raman spectrum of the same position on different ores; extracting spectrum features and fusing the extracted spectrum features into a spectrum feature matrix; establishing an ore classification model based on a machine learning method; when underwater ore identification is performed, LIBS spectrum and pulse Raman spectrum of the same position on the ore are acquired, and the ore type is identified through the ore classification model; through the combination of LIBS-pulse Raman spectrum technology, the present application realizes the simultaneous detection of element information and molecular information in underwater ores, extracts spectrum features of the two kinds of spectrum to combine into a spectrum matrix, and establishes an ore classification model through a machine learning method, so that the accuracy of underwater ore identification is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of underwater ore identification technology, and in particular to an underwater ore in-situ detection method and system based on a combination of LIBS and pulsed Raman spectroscopy. Background Technology

[0002] Deep-sea metallic mineral resources (such as polymetallic nodules, cobalt-rich crusts, and polymetallic sulfides) have been proven to have enormous exploitation value. Traditional sampling techniques bring seabed samples back to the laboratory for analysis, accurately obtaining chemical composition information from a specific location on the seabed. However, this information acquisition has a significant time lag, failing to provide real-time, rapid feedback and scientific control during the mining process, and is difficult to conduct long-term, large-scale seabed surveys. Therefore, future deep-sea mineral resource development urgently requires the development of rapid, accurate, and environmentally friendly deep-sea in-situ exploration technologies.

[0003] Laser-induced breakdown spectroscopy (LAS) is a spectroscopic analysis method that focuses a high-peak-power pulsed laser onto the sample surface, causing the laser to interact with the material, inducing breakdown, generating plasma, and emitting light. Qualitative and quantitative analysis of the test sample is achieved by analyzing the emission spectrum of the plasma, and elemental information can be provided by analyzing the position of spectral peaks. Laser Raman spectroscopy is a non-destructive molecular spectroscopy technique. Based on the interaction between excitation light and molecules in the sample, the sample molecules produce scattered light of different wavelengths. These scattered photons provide information about the vibrational and rotational energy levels of the molecules, thus Raman spectroscopy can provide information about the molecular structure of the sample.

[0004] Laser-induced breakdown spectroscopy (LIBS) and laser Raman spectroscopy (LAS), as emerging and promising technologies, offer advantages such as requiring no sample pretreatment, enabling real-time in-situ analysis of sample composition, and being non-destructive or minimally destructive to the sample. However, existing single LIBS or Raman techniques, when applied to underwater ore detection, provide incomplete sample information and limit detection capabilities. Therefore, we propose a combined LIBS and pulsed Raman in-situ detection method and system for underwater ore. Summary of the Invention

[0005] The purpose of this invention is to provide an underwater in-situ ore detection method and system based on LIBS and pulsed Raman combined, so as to solve the problems of single ore detection components and low identification accuracy of existing underwater ore detection technologies.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for in-situ underwater ore detection based on LIBS and pulsed Raman spectroscopy, the method comprising:

[0008] The LIBS-pulse Raman combined spectral acquisition device was activated to collect laser-induced breakdown spectra and pulsed Raman spectra at the same location on different ores;

[0009] Extract spectral features and fuse them into a spectral feature matrix;

[0010] Establish an ore classification model based on machine learning methods;

[0011] When identifying underwater ores, laser-induced breakdown spectra and pulsed Raman spectra at the same location on the underwater ores are obtained, and the types of ores are identified using an ores classification model.

[0012] Furthermore, the extracted spectral features are fused into a spectral feature matrix by matrix addition.

[0013] Furthermore, the method for establishing an ore classification model includes the following steps:

[0014] The number of principal components of the partial least squares function that minimizes the root mean square error is determined by leave-one-out cross-validation. This number is denoted as the optimal number of principal components. A PLS model between the spectral feature matrix and the ore type is then established, which is the ore classification model.

[0015] Furthermore, the method for identifying ore types using ore classification models includes the following steps:

[0016] To obtain the laser-induced breakdown spectrum and pulsed Raman spectrum of underwater ores;

[0017] Spectral features of laser-induced breakdown spectrum and pulsed Raman spectrum are extracted and fused into a spectral feature matrix;

[0018] The obtained fused spectral matrix is ​​input into the ore classification model, and the ore type is determined based on the output of the ore classification model.

[0019] This invention also discloses an underwater ore in-situ detection system, capable of realizing the underwater ore in-situ detection method based on LIBS and pulsed Raman combined as described in any of the above-mentioned methods, including:

[0020] The LIBS-pulse Raman combined spectral acquisition device is used to acquire laser-induced breakdown spectrum and pulsed Raman spectrum at the same location in underwater ore.

[0021] The spectral feature extraction unit is used to extract spectral features and fuse the spectral feature matrix.

[0022] The ore classification unit is used to load an ore classification model to classify ores.

[0023] Furthermore, the LIBS-pulse Raman combined spectral acquisition device includes:

[0024] A pulsed laser, used to emit pulsed laser light;

[0025] Laser attenuator, used to modulate the energy of pulsed laser to create plasma and Raman scattering on the surface of minerals;

[0026] The spectral signal receiving unit is used to receive laser-induced breakdown spectra and pulsed Raman spectra;

[0027] A digital pulse generator is used to control a pulsed laser and a spectral signal receiving unit;

[0028] When probing underwater ore, a pulsed laser emits laser light, which is then irradiated onto the ore to be classified after being adjusted by a laser attenuator. When acquiring the laser-induced breakdown spectrum, the laser attenuator adjusts the energy to high to form plasma, and the spectral signal receiving unit collects the laser-induced breakdown spectrum. A digital pulse generator controls the pulsed laser and the spectral signal receiving unit to acquire the spectrum with a delay. When acquiring the pulsed Raman spectrum, the laser attenuator adjusts the energy to low to form Raman scattering, and the spectral signal receiving unit collects the pulsed Raman spectrum. A digital pulse generator controls the pulsed laser and the spectral signal receiving unit to acquire the spectrum synchronously.

[0029] Furthermore, the spectral signal receiving unit includes:

[0030] Eyeglass frames;

[0031] A spectrometer is used to receive spectral signals;

[0032] Enhanced charge-coupled devices for controlling the switching of spectrometers;

[0033] The second and third lenses are sequentially arranged along the direction from the ore to the spectrum;

[0034] A detachable long-pass filter is connected between the second and third lenses. When acquiring laser-induced breakdown spectra, the long-pass filter is removed; when acquiring Raman spectra, the long-pass filter is installed.

[0035] Furthermore, the spectral signal receiving unit also includes:

[0036] A laser positioning unit is used to locate the focal point of a laser beam.

[0037] In summary, the present invention has the following advantages compared with the prior art:

[0038] The underwater ore in-situ detection method based on the combination of LIBS and pulsed Raman spectroscopy disclosed in this invention can simultaneously detect elemental and molecular information in underwater ores by combining LIBS-pulse Raman spectroscopy, thereby obtaining more comprehensive chemical composition information of underwater ores. Furthermore, the use of time-resolved pulsed Raman spectroscopy effectively avoids interference from fluorescence background. By extracting features from LIBS and Raman spectra and combining them into a spectral feature matrix, a ore classification model is established using machine learning methods, which greatly improves the accuracy of underwater ore identification. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating the steps of an underwater in-situ ore detection method based on a combination of LIBS and pulsed Raman spectroscopy, as disclosed in an embodiment of the present invention.

[0040] Figure 2 This is a flowchart of one of the steps in the underwater in-situ ore detection method based on LIBS and pulsed Raman combined, as disclosed in an embodiment of the present invention.

[0041] Figure 3 This is a typical laser-induced breakdown spectrum of an ore obtained by the underwater in-situ ore detection method based on LIBS and pulsed Raman spectroscopy disclosed in an embodiment of the present invention.

[0042] Figure 4 The image shows a typical ore Raman spectrum collected by the underwater in-situ ore detection method based on LIBS and pulsed Raman combined, as disclosed in an embodiment of the present invention.

[0043] Figure 5 This is a schematic diagram of the underwater ore in-situ detection system disclosed in an embodiment of the present invention.

[0044] Figure 6 This is a schematic diagram of a specific embodiment of the underwater ore in-situ detection system disclosed in this invention.

[0045] Figure 7 for Figure 6 A flowchart of the disclosed implementation method for acquiring spectral information. Detailed Implementation

[0046] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0047] Example 1

[0048] like Figure 1 As shown, one embodiment of the present invention provides an underwater in-situ ore detection method based on LIBS and pulsed Raman spectroscopy, the method comprising:

[0049] Step S100: Start the LIBS-pulse Raman combined spectral acquisition device to acquire laser-induced breakdown spectra and pulsed Raman spectra at the same location on different ores;

[0050] Specifically, in this embodiment, the LIBS-pulse Raman joint spectral acquisition device is activated, and the laser is directed to irradiate the ore located underwater. When the high-energy laser is focused on the ore, plasma is generated. When the low-energy laser irradiates the ore, Raman scattering is generated. The laser-induced breakdown spectrum is generated by acquiring the emission spectrum of the plasma, and the pulsed Raman spectrum is generated by acquiring the spectrum when Raman scattering occurs.

[0051] Laser-induced breakdown spectra and pulsed Raman spectra of several known ores were collected, such as siderite, anhydrite, barite, and calcite.

[0052] It should be noted that when collecting laser-induced breakdown spectra and pulsed Raman spectra, the laser-induced breakdown spectra and pulsed Raman spectra of the same ore are collected from the same locations.

[0053] Step S200: Extract spectral features and fuse the extracted spectral features into a spectral feature matrix;

[0054] Specifically, in this embodiment, spectral features are extracted from the acquired spectral information, and the extracted spectral features are fused into a spectral feature matrix;

[0055] In this embodiment, the spectral features are extracted using the Continuous Projection Algorithm (SPA).

[0056] Specifically, in this step, spectral feature extraction is performed using SPA (Successive Projection Algorithm). SPA is a forward feature variable selection method. SPA utilizes vector projection analysis, projecting wavelengths onto other wavelengths and comparing the magnitudes of the projection vectors. The wavelength with the largest projection vector is selected as the candidate wavelength. Then, the final feature wavelength is selected based on the correction model. SPA selects the variable combination containing the least redundant information and the least collinearity. The specific operation steps are as follows:

[0057] Input spectral information respectively in MATLAB, where the spectral information is the laser-induced breakdown spectrum or pulsed Raman spectrum obtained in step 100, perform a projection operation on the spectral information and apply it to Xcal, where Xcal is the modeling set data in subsequent applications, perform mean centering and auto-scaling processing on the data in Xcal to increase the difference between ore spectra, and during the auto-scaling processing, the parameter of the auto-scaling processing defaults to 1;

[0058] Set the initial band k(0), select the number of variables N, the spectral matrix has J columns, and the spectral matrix is a matrix composed of several spectra, with each spectrum constituting one column of the matrix;

[0059] The spectral matrix in this step refers to a spectral matrix formed by collecting several pieces of spectral information at the same position of the ore;

[0060] Step a, denote the initial iteration vector as x k(0) , and the number of variables to be extracted is N, and the spectral matrix has J columns;

[0061] Step b, arbitrarily select one column (the j-th column) of the spectral matrix, and assign the j-th column of the modeling set to x j , which is denoted as x k(0) ;

[0062] Step c, denote the set of positions of the unselected column vectors as S, where S is calculated by formula (1):

[0063]

[0064] Step d, calculate the projection of x based on formula (2) respectively j onto the remaining column vectors, which is denoted as P xj :

[0065]

[0066] Step e, extract the spectral wavelength of the maximum projection vector;

[0067] k(n)=arg(max(||P xj ||)), j∈S, formula (3)

[0068] Step f, let x j =P x , j∈S;

[0069] Step g, n=n+1, if n<N, return to step c.

[0070] Finally, the extracted variable is x k(i)i = 0, 1, ..., N-1, corresponding to k(n) and N in each iteration, establish a multiple linear regression (MLR) model to obtain the root mean square error of cross-validation (RMSECV) of the modeling set, corresponding to different candidate subsets, where the minimum root mean square error of cross-validation (RMSECV) value corresponds to the optimal values ​​of k(n) and N.

[0071] The selected variables are sorted according to their correlation, and the feature points extracted from the laser-induced breakdown spectrum and the pulsed Raman spectrum are fused to form a spectral feature matrix.

[0072] Each spectral wavelength corresponds to a projection vector. The correlation is determined by the proportion of each projection vector to all projection vectors. The larger the proportion, the stronger the correlation. The extracted spectral features are arranged from largest to smallest based on their correlation.

[0073] The feature points extracted from laser-induced breakdown spectrum and pulsed Raman spectrum are fused to form a new spectral feature matrix;

[0074] For example, if the matrix formed by the feature points of the laser-induced breakdown spectrum is A, and the matrix formed by the feature points of the pulsed Raman spectrum is B, then the fused spectral matrix is ​​A+B.

[0075] Specifically, matrix A is a row of a and column J, with the first column being the spectral wavelength and the remaining columns being the spectral intensity; matrix B is a row of b and column J, with the first column being the spectral wavenumber and the remaining columns being the spectral intensity. Then the fused spectral matrix is ​​a (a+b) row and column J matrix.

[0076] Step S300: Establish an ore classification model based on machine learning methods;

[0077] Specifically, in this embodiment, a partial least squares (PLS) model is established for ore classification. Specifically, the plsregress function in MATLAB is used to find the number of principal components corresponding to the minimum value of RMSECV in the spectral feature matrix using leave-one-out cross-validation. This number is denoted as the optimal number of principal components. The PLS model between the spectral feature matrix and the ore type is established in this way, which is the ore classification model.

[0078] For example, finding the optimal principal component number corresponding to the spectral feature matrix formed by ores such as siderite, anhydrite, barite, and calcite, and classifying known ores by modeling with the optimal principal component number;

[0079] It should be noted that, in this embodiment, ore classification models can also be established in other ways, such as random forest algorithm and principal component regression algorithm. The specific implementation methods of each algorithm are existing technologies and will not be described in detail here.

[0080] Step S400: When identifying underwater ore, obtain the laser-induced breakdown spectrum and pulsed Raman spectrum at the same location on the underwater ore, and identify the type of ore using the ore classification model;

[0081] Specifically, such as Figure 2 As shown, the method for identifying ore types using an ore classification model includes the following steps:

[0082] Step S410: Obtain the laser-induced breakdown spectrum and pulsed Raman spectrum of the underwater ore;

[0083] Specifically, the laser-induced breakdown spectrum and pulsed Raman spectrum of underwater ore are obtained through the LIBS-pulse Raman combined spectral acquisition device in step S100.

[0084] Step S420: Extract the spectral features of LIBS and pulsed Raman spectra, and fuse the spectral features into a spectral feature matrix;

[0085] Specifically, spectral features of LIBS and pulsed Raman spectra are extracted in step S200 and fused into a spectral feature matrix;

[0086] Step S430: Input the obtained spectral feature matrix into the ore classification model, and determine the ore type based on the output of the ore classification model;

[0087] Specifically, in this step, the spectral feature matrix is ​​input into the ore classification model, and the ore classification model outputs the predicted probability. By selecting the ore type corresponding to the maximum probability, underwater ore can be identified.

[0088] In summary, the underwater ore in-situ detection method based on the combined LIBS and pulsed Raman spectra disclosed in this invention acquires the laser-induced breakdown spectrum and pulsed Raman spectrum of underwater ore. The laser-induced breakdown spectrum obtains elemental information at a certain location in the underwater ore, while the pulsed Raman spectrum obtains molecular information at the same location. The spectral features of the two spectra are extracted and combined into a spectral feature matrix. An ore classification model is established using machine learning methods. When classifying ore, combining the two spectra for ore identification can greatly improve the accuracy of ore identification.

[0089] To verify the accuracy of the underwater ore classification method based on LIBS and pulsed Raman combined disclosed in this embodiment of the invention, a verification experiment was designed in the laboratory. The laser-induced breakdown spectrum and pulsed Raman spectrum of typical ores were used to extract spectral features using the method in step S200, and then the method in step S300 was used to establish ore classification models. The known ores were then classified to obtain the accuracy.

[0090] Specifically, typical ores such as siderite, anhydrite, barite, and calcite were placed underwater, and the LIBS and pulsed Raman spectra of the typical ores were collected by a LIBS-pulse Raman combined spectroscopy acquisition device.

[0091] The spectral features of LIBS spectrum and pulsed Raman spectrum are extracted respectively using the method in step S200. The spectral features of LIBS spectrum are divided into a first training set and a first prediction set as the first control group. The spectral features of pulsed Raman spectrum are divided into a second training set and a second prediction set as the second control group. The spectral feature matrix fused in the manner of step S200 is divided into a third training set and a third prediction set as the experimental group.

[0092] A first classification model is built using the first training set through step S300; a second classification model is built using the second training set through step S300; and a third classification model is generated using the third training set through step S300, which serves as the experimental group.

[0093] Run the first classification model to classify the first prediction set and calculate the accuracy of the first classification model in identifying ores; run the second classification model to classify the second prediction set and calculate the accuracy of the second classification model in identifying ores; run the third classification model to classify the third prediction set and calculate the accuracy of the third classification model in identifying ores.

[0094] The accuracy of ore identification is calculated using formula (1):

[0095]

[0096] like Figure 3 This is a typical laser-induced breakdown spectrum of underwater ores, in which... Figure 3 The laser-induced breakdown spectra of siderite, anhydrite, barite, and calcite are shown. Figure 4 This is a typical underwater mineral pulse Raman spectrum, in which, Figure 4 The pulsed Raman spectra of siderite, anhydrite, barite, and calcite are shown.

[0097] Based on laboratory data, as shown in Table 1, the accuracy rate of ore classification by laser-induced breakdown spectroscopy is 90.44%, the accuracy rate of ore classification by pulsed Raman spectroscopy is 92.89%, and the accuracy rate of ore classification by the LIBS-pulsed Raman spectroscopy fusion method disclosed in this embodiment of the invention can reach 98.44%. Therefore, the underwater ore in-situ detection method based on the combination of LIBS and pulsed Raman spectroscopy disclosed in this embodiment of the invention can greatly improve the accuracy of underwater ore identification.

[0098] Table 1. Accuracy of different ore classification methods

[0099] accuracy 90.44％， 92.89％ 98.44％

[0100] Example 2

[0101] This invention also discloses an underwater ore in-situ detection system, capable of realizing any of the above-described underwater ore in-situ detection methods based on LIBS and pulsed Raman spectroscopy, such as... Figure 5 As shown, the system 500 includes:

[0102] The LIBS-pulse Raman combined spectral acquisition device 510 is used to acquire laser-induced breakdown spectra and pulsed Raman spectra of ores.

[0103] The spectral feature extraction unit 520 is used to extract spectral features and fuse the spectral feature matrix.

[0104] Ore classification unit 530 is used to load an ore classification model to classify ores;

[0105] Specifically, in this embodiment, when establishing an ore classification model, the LIBS-pulse Raman combined spectral acquisition device 510 acquires the LIBS spectrum and pulse Raman spectrum of a typical ore, and then inputs the LIBS spectrum and pulse Raman spectrum into the spectral feature extraction unit 520. The spectral feature extraction unit 520 extracts the spectral features of the LIBS spectrum and pulse Raman spectrum, and fuses the extracted spectral features into a spectral feature matrix. The fused spectral matrix is ​​sent to the ore classification unit 530, and the ore classification unit 530 establishes an ore classification model based on the fused spectral matrix.

[0106] When identifying ore, the LIBS spectrum and pulsed Raman spectrum of underwater ore are acquired by the LIBS-pulse Raman combined spectral acquisition device 510, and then sent to the spectral feature extraction unit 520 to generate a fused spectral matrix. The ore classification unit 530 classifies the ore based on the fused spectral matrix, thereby identifying the ore.

[0107] Furthermore, such as Figure 6 As shown, the LIBS-pulse Raman combined spectral acquisition device 510 includes:

[0108] Pulsed laser 1, used to emit pulsed laser light;

[0109] Laser attenuator 3 is used to adjust the energy of the pulsed laser to form plasma and Raman scattering on the ore surface;

[0110] The spectral signal receiving unit is used to receive laser-induced breakdown spectra and pulsed Raman spectra;

[0111] Digital pulse generator 17 is used to control the pulsed laser and the spectral signal receiving unit;

[0112] When probing underwater ore, the digital pulse generator 17 controls the pulsed laser to emit laser light, which is then irradiated onto the ore to be classified after being adjusted by the laser attenuator 3. When acquiring the laser-induced breakdown spectrum, the laser attenuator 3 adjusts to high energy to form plasma, and the spectral signal receiving unit collects the laser-induced breakdown spectrum. The digital pulse generator 17 controls the pulsed laser 1 and the spectral signal receiving unit to acquire the spectrum with a delay. When acquiring the pulsed Raman spectrum, the laser attenuator 3 adjusts to low energy to form Raman scattering, and the spectral signal receiving unit collects the pulsed Raman spectrum. The digital pulse generator 17 controls the pulsed laser 1 and the spectral signal receiving unit to acquire the spectrum synchronously.

[0113] Specifically, in this embodiment, such as Figure 7 As shown, the digital pulse generator 17 is electrically connected to the pulse laser 1 and the spectral signal receiving unit. The laser from the pulse laser 1 is reflected by the first reflector 3 and irradiates the laser attenuator 3. The laser emitted from the laser attenuator 3 irradiates the beam expanding and collimating optical path 4. The laser emitted from the beam expanding and collimating optical path 4 irradiates the first lens 6 through the second reflector 5 and then irradiates the ore 7.

[0114] The laser attenuator 3 is composed of a zero-order half-wave plate and a polarizing prism, and the laser attenuator 3 can adjust the laser energy.

[0115] Furthermore, the spectral signal receiving unit includes:

[0116] Eyeglass frames;

[0117] A spectrometer and an ICCD (intensity-coupled device) 12 are used to receive spectral signals;

[0118] A second lens 8 and a third lens 10 are sequentially arranged along the direction from the ore to the spectrum;

[0119] A long-pass filter 9 is detachably connected between the second lens 8 and the third lens 10. When acquiring laser-induced breakdown spectra, the long-pass filter 9 is removed; when acquiring pulsed Raman spectra, the long-pass filter 9 is installed.

[0120] The spectrometer and ICCD12 are connected to the third lens 10 via optical fiber 11 to acquire spectral information;

[0121] The spectrometer and ICCD12 are electrically connected to a computer control system 18, and the spectral feature extraction unit 520 and the ore classification unit 530 are computer programs located within the computer control system 18.

[0122] Furthermore, the spectral signal receiving unit also includes:

[0123] A laser positioning unit is used to locate the focal point of the laser.

[0124] In this embodiment, as Figure 6 As shown, the laser positioning unit consists of a positioning laser 13, a CCD camera 16, a fourth lens 15, and a rotating attenuator 14. The optical axis of the CCD camera 16 is aligned with the axis of the fourth lens 15. The second reflector 5 is a semi-transparent and semi-reflective mirror. During laser positioning, the positioning laser emitted by the positioning laser 13 is irradiated onto the ore 7 via the rotating attenuator 14. The CCD camera 16 observes the laser point on the ore 7 through the second reflector 5 and the fourth lens 15 to adjust the focal point of the laser emitted by the pulse laser 1 on the ore 7, so that the laser emitted by the positioning laser 13 and the laser emitted by the pulse laser 1 coincide on the ore 7. The CCD camera 16 and the computer control system 18 display the focal point of the pulse laser 1.

[0125] Specific methods for obtaining spectral information include Figure 7 As shown, firstly, the computer control system 18 is turned on, and the pulsed laser 1 and digital pulse generator 17 are started. The frequency and delay of the laser flash and Q-switch are set. Then, the positioning laser 13 and industrial camera 16 are started to determine whether the laser spot of the pulsed laser 1 on the ore 7 coincides with the positioning laser of the positioning laser 13. If they coincide, the spectrometer ICCD 12 is started to prepare for spectrum acquisition. If they do not coincide, the focusing point of the pulsed laser 1 on the ore 7 is adjusted to change the focusing position on the sample surface so that they coincide. In the spectrometer software on the computer control system 18, parameters such as gate width, gain, and number of acquisitions are set to acquire the pulsed Raman spectrum of the underwater ore. The long-pass filter 9 is removed, and the laser attenuator 3 is adjusted to change the laser energy. Parameters are set in the computer control system 18 to perform LIBS spectrum detection at the same point. After acquisition, the data is analyzed and processed by the computer control system 18.

[0126] Example 3

[0127] The present invention also discloses a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to implement the underwater in-situ mineral detection method based on LIBS and pulsed Raman spectroscopy as described in Example 1.

[0128] In the several embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0129] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0130] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0131] In a typical configuration of an embodiment of the present invention, the electronic device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0132] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash-RAM). Memory is an example of computer-readable media.

[0133] Readable storage media include both permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data.

[0134] Examples of storage media for electronic devices include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media do not include non-transitory computer-readable media, such as modulated data signals and carrier waves.

[0135] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is merely an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

Claims

1. A method for in-situ underwater ore detection based on a combination of LIBS and pulsed Raman spectroscopy, characterized in that, The method includes the following steps: The LIBS-pulse Raman combined spectral acquisition device was activated to collect laser-induced breakdown spectra and pulsed Raman spectra at the same location on different ores; Spectral features are extracted and fused into a spectral feature matrix. The method for extracting spectral features is a continuous projection algorithm. Establish an ore classification model based on machine learning methods; When identifying underwater ores, laser-induced breakdown spectra and pulsed Raman spectra at the same location on the underwater ore are obtained, and the ore types are identified using an ore classification model. The method for establishing an ore classification model includes the following steps: The number of principal components of the partial least squares function that minimizes the root mean square error is determined by leave-one-out cross-validation. This number is denoted as the optimal number of principal components. A PLS model between the spectral feature matrix and the ore type is then established, which is the ore classification model. The LIBS-pulse Raman combined spectral acquisition device includes: A pulsed laser, used to emit pulsed laser light; Laser attenuator, used to modulate the energy of pulsed laser to create plasma and Raman scattering on the surface of minerals; A spectral signal receiving unit is used to receive laser-induced breakdown spectra and pulsed Raman spectra. The spectral signal receiving unit includes a frame, a spectrometer, an enhanced charge-coupled device (CCD), and a long-pass filter detachably connected between a second lens and a third lens. The spectrometer is used to receive spectral signals; the enhanced CCD is used to control the spectrometer's on / off state; the long-pass filter is removed when acquiring laser-induced breakdown spectra; and the long-pass filter is installed when acquiring pulsed Raman spectra. A digital pulse generator is used to control a pulsed laser and a spectral signal receiving unit; When probing underwater ore, a pulsed laser emits laser light, which is then irradiated onto the ore to be classified after being adjusted by a laser attenuator. When acquiring the laser-induced breakdown spectrum, the laser attenuator adjusts the energy to high to form plasma, and the spectral signal receiving unit collects the laser-induced breakdown spectrum. A digital pulse generator controls the pulsed laser and the spectral signal receiving unit to acquire the spectrum with a delay. When acquiring the pulsed Raman spectrum, the laser attenuator adjusts the energy to low to form Raman scattering, and the spectral signal receiving unit collects the pulsed Raman spectrum. A digital pulse generator controls the pulsed laser and the spectral signal receiving unit to acquire the spectrum synchronously.

2. The underwater in-situ ore detection method based on LIBS and pulsed Raman spectroscopy as described in claim 1, characterized in that, The extracted spectral features are fused into a spectral feature matrix by matrix addition.

3. The underwater in-situ ore detection method based on LIBS and pulsed Raman spectroscopy as described in claim 1, characterized in that, The method for identifying ore types using ore classification models includes the following steps: To obtain the laser-induced breakdown spectrum and pulsed Raman spectrum of underwater ores; Spectral features of laser-induced breakdown spectrum and pulsed Raman spectrum are extracted and fused into a spectral feature matrix; The obtained spectral feature matrix is ​​input into the ore classification model, and the ore type is determined based on the output of the ore classification model.

4. An underwater in-situ ore detection system, characterized in that, The underwater in-situ ore detection method based on LIBS and pulsed Raman spectroscopy, as described in any one of claims 1-3, comprises: The LIBS-pulse Raman combined spectral acquisition device is used to acquire laser-induced breakdown spectrum and pulsed Raman spectrum at the same location in underwater ore. The spectral feature extraction unit is used to extract spectral features and fuse the spectral feature matrix. The ore classification unit is used to load an ore classification model to classify ores.

5. The underwater ore in-situ detection system according to claim 4, characterized in that, The spectral signal receiving unit further includes: A laser positioning unit is used to locate the focal point of a laser beam.

Images