Spectral data real-time processing and quantitative analysis system based on LIBS (Laser-induced Breakdown Spectroscopy) technology

Abstract

The invention relates to the technical field of laser-induced breakdown spectroscopy, in particular to an LIBS (laser-induced breakdown spectroscopy) technology-based real-time processing and quantitative analysis system for spectral data, which comprises a laser emission module, a spectrum acquisition module, a real-time processing unit, a quantitative analysis module, an environmental parameter monitoring module and a remote communication module, the sub-second response speed is realized through an FPGA parallel processing architecture, the signal processing precision is improved by adopting an improved wavelet threshold algorithm and a self-adaptive peak position detection algorithm, and a multi-element analysis error is controlled to be lt through a matrix effect correction algorithm; 5%. The system is provided with an environmental parameter monitoring module and a remote communication module, achieves environmental interference suppression and remote data transmission, is suitable for rapid element quantitative analysis scenes in industries such as metallurgy, environmental protection and geological exploration, and has the remarkable advantages of being high in real-time processing capacity, high in quantitative precision, high in anti-interference capacity and the like.

Description

Technical Field

[0001] This invention relates to the field of laser-induced breakdown spectroscopy, and in particular to a real-time processing and quantitative analysis system for spectral data based on LIBS technology. Background Technology

[0002] Laser-induced breakdown spectroscopy (LIBS), as an emerging atomic emission spectroscopy technique, offers significant advantages such as no sample pretreatment required, simultaneous multi-element analysis, and remote detection. However, existing LIBS systems suffer from significant shortcomings in real-time processing capabilities, quantitative accuracy, and interference resistance.

[0003] Existing technologies generally suffer from the following common problems: Weak real-time processing capability: Traditional systems use general-purpose processors for data processing, and the processing speed is limited by the processor performance, with analysis cycles lasting several minutes. Insufficient quantitative accuracy: The characteristic peak extraction algorithm is easily affected by matrix effects, and the error is large when analyzing multiple elements simultaneously; Poor anti-interference ability: Interference factors such as ambient temperature and laser energy fluctuations can cause measurement results to drift; Low system integration: Each functional module is designed separately, resulting in a large system size that is not conducive to on-site deployment. Summary of the Invention

[0004] The purpose of this invention is to provide a real-time processing and quantitative analysis system for spectral data based on LIBS technology, in order to solve the technical problems existing in the prior art.

[0005] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: A real-time processing and quantitative analysis system for spectral data based on LIBS technology, comprising: Laser emission module: Employs an Nd:YAG solid-state laser with a pulse width ≤3ns, model Spectra-Physics Quanta-Ray PRO-Series, operating wavelength 1064nm±1nm, repetition rate continuously adjustable from 1-1000Hz via FPGA control, single-pulse energy stability ≤±1.5%, beam quality M 2 ≤1.2; Spectrum acquisition module: includes a CMOS detector array, specifically using a Sony IMX250 sensor with a pixel size of 8μm×8μm, an effective pixel count of 5120×3840, a quantum efficiency of ≥45%@532nm band, a dynamic range of 72dB, and a dark current of ≤0.1e- / s; Real-time processing unit: Integrates a Xilinx Zynq UltraScale+ MPSoC chip to achieve parallel data processing, and is configured with: (1) Background subtraction submodule: An improved wavelet thresholding algorithm is adopted, with a threshold function λ=σ√(2logN) / √(1+αt), where σ is the noise standard deviation calculated in real time by the median absolute deviation method, N is the data length of 2048 points, and α is the dynamic adjustment coefficient that is automatically adjusted from 0.1 to 1.0 according to the signal-to-noise ratio. (2) Feature extraction submodule: Combining continuous wavelet transform and adaptive peak position detection algorithm, the continuous wavelet transform adopts Mexican hat wavelet basis, and the scale parameter is dynamically adjusted in the range of 1-64 according to the signal-to-noise ratio. The adaptive peak position detection achieves peak position positioning error <0.05nm by calculating the zero crossing point of the first derivative and using cubic spline interpolation. (3) Data caching submodule: A dual-port RAM with a capacity of 256KB is used to achieve seamless connection of data streams; (4) Outlier detection submodule: Based on the improved Grubbs criterion, it dynamically identifies and removes outlier data points. The improved Grubbs criterion judges the outlier by calculating the deviation of the data points and combining it with the 95% confidence threshold. Quantitative Analysis Module: Includes a multi-element standard curve library and a matrix effect correction unit. The multi-element standard curve library is obtained through experimental calibration and contains standard curves for 20 common elements such as Fe, Cu, Al, C, and O. The matrix effect correction unit uses the internal standard method combined with a multivariate correction algorithm. The internal standard element is Si. The multivariate correction algorithm uses partial least squares regression (PLSR) to establish a concentration prediction model.

[0006] Furthermore, the laser emitting module also includes: Laser energy monitoring unit: Uses a Thorlabs S120C silicon photodiode to monitor laser energy in real time, with energy stability ≤ ±1% and response time ≤ 1μs; Beam shaping unit: includes a tunable beam expander (Thorlabs BE02-1064) and an aperture (ThorlabsSM1D12), enabling continuous adjustment of the beam diameter within the range of 50-500μm with an adjustment accuracy of 1μm; Laser triggering unit: A high-speed optocoupler is used to synchronize laser triggering and data acquisition, with a synchronization accuracy of ≤10ns.

[0007] Furthermore, the spectral acquisition module is configured with: Fiber optic coupling unit: The fiber used is Thorlabs FT200EMT multimode fiber with a core diameter of 200μm±5μm, a numerical aperture of 0.22±0.02, and a fiber length of 5m±0.1m. Grating beam splitting unit: adopts Richardson Gratings53-*-120R blazed grating with a line density of 1200 lines / mm, blaze wavelength of 500nm±10nm, spectral resolution ≤0.03nm, and diffraction efficiency ≥80%@500nm band; Focusing lens group: The doublet lens is used to achieve efficient focusing of spectral signals, with a focal length of 100mm, a light-transmitting aperture of 30mm, and aberration correction to the diffraction limit.

[0008] Furthermore, the real-time processing unit also includes: Algorithm acceleration module: Hardware acceleration of wavelet transform and PLSR algorithm is implemented using hardware description language, with a processing speed ≥100MSPS; Dynamic parameter adjustment module: Automatically adjusts wavelet transform scale and threshold parameters based on real-time signal-to-noise ratio, with an adjustment range covering scales from 1 to 64 levels; Data compression module: Uses lossless compression algorithm to compress and store spectral data with a compression ratio ≥2:1.

[0009] Furthermore, the quantitative analysis module includes: Matrix effect correction unit: The internal standard method combined with a multivariate correction algorithm is used. The correction coefficients are obtained through experimental calibration. The number of calibration samples is ≥100 and the calibration error is <3%. Dynamic concentration display unit: Equipped with an LCD touch screen model Newhaven Display NHD-7.0-800480EF, which displays elemental concentrations and confidence levels of analysis results in real time. The confidence level is calculated using the Monte Carlo simulation method. Results Validation Module: The cross-validation algorithm is used to assess the reliability of the analysis results, with a cross-validation error of <2%.

[0010] Furthermore, it also includes an environmental parameter monitoring module, which comprises: Temperature sensor: PT1000 platinum resistance thermometer, Heraeus C220, measuring range -20℃ to 80℃, accuracy ±0.2℃, response time ≤1s; Barometric pressure sensor: Bosch BMP280, measuring range 800-1100 hPa, accuracy ±0.5 hPa, response time ≤1 ms; Humidity sensor: Honeywell HIH-5030, measurement range 0-100%RH, accuracy ±2%RH, response time ≤10s; Parameter compensation unit: A polynomial fitting algorithm is used to achieve real-time compensation of environmental parameters for spectral data, and the fitting order is automatically adjusted according to changes in environmental parameters.

[0011] Furthermore, the environmental parameter monitoring module uses the extended Kalman filter algorithm for data fusion to achieve real-time correction of environmental parameters. The extended Kalman filter algorithm achieves optimal estimation by combining historical data and real-time measurement data of environmental parameters with state equations and observation equations. The state transition matrix is ​​automatically updated according to the rate of change of environmental parameters.

[0012] Furthermore, it also includes a remote communication module, which comprises: Wireless transmission module: Supports Ethernet and 4G / 5G wireless transmission protocols. The Ethernet interface uses Intel I210-IT with a transmission rate of 1000Mbps. The 4G / 5G module uses Quectel EG25-G and supports the LTE Cat 4 standard. Configured data encryption unit: AES-256 algorithm is used to achieve data transmission security protection, encryption chip model Microchip ATECC608A, encryption speed ≥200Mbps; Remote control unit: Supports remote parameter configuration and system diagnostics, and uses the SSH protocol to achieve secure remote access.

[0013] Furthermore, a modular design is adopted, with each module enabling plug-and-play functionality through a standard interface. The standard interface adopts the CPCI bus standard, with a data transmission rate of ≥1Gbps, and hot-swappable interfaces support online system maintenance.

[0014] As an improvement, the beneficial effects of the present invention are as follows: 1. Real-time processing capability: Adopting an FPGA parallel processing architecture, it achieves sub-second response speed to meet the real-time detection needs of industrial sites; 2. Improved quantitative accuracy: Through matrix effect correction algorithms, the error of multi-element analysis is controlled to <5%, reaching the international advanced level; 3. Enhanced anti-interference capability: The system employs environmental parameter monitoring and dynamic correction algorithms to effectively suppress environmental interference and improve system stability. 4. Improved system integration: The modular design allows each module to be plug-and-play via standard interfaces, facilitating system maintenance and upgrades; 5. Remote communication capability: Supports Ethernet and 4G / 5G wireless transmission protocols to enable remote transmission and sharing of analysis results. Attached Figure Description

[0015] Figure 1 This is a diagram illustrating the overall architecture of a real-time processing and quantitative analysis system for spectral data based on LIBS technology according to the present invention. Figure 2 This is an architectural diagram of the laser emission module of the present invention; Figure 3This is an architectural diagram of the spectral acquisition module of the present invention; Figure 4 This is an architecture diagram of the real-time processing unit of the present invention; Figure 5 This is an architecture diagram of the quantitative analysis module of the present invention; Figure 6 This is an architecture diagram of the environmental parameter monitoring module of the present invention; Figure 7 This is an architecture diagram of the remote communication module of the present invention; Detailed Implementation

[0016] To make the content of this invention easier to understand, the technical solutions in the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings.

[0017] Example 1: Analysis of Steel Samples The system described in this invention was used to perform elemental analysis on steel samples. The sample was a low-carbon steel standard material GBW01203, and the main element contents were: Fe 98.5%, C 0.5%, Si 0.8%, Mn 0.2%, P 0.04%, S 0.03%.

[0018] System configuration parameters: laser energy 25mJ±1mJ, spot diameter 250μm±5μm, repetition frequency 15Hz±1Hz, integration time 10ms. The spectral acquisition module uses a CMOS detector array with a pixel size of 8μm×8μm, an effective pixel count of 5120×3840, and a quantum efficiency ≥45%@532nm band. The real-time processing unit uses a Xilinx Zynq UltraScale+MPSoC chip, configured with a background subtraction submodule, a feature extraction submodule, a data caching submodule, and an outlier detection submodule.

[0019] The analysis process is as follows: Step 1: The laser emission module generates laser pulses, and the beam shaping unit adjusts the spot diameter to 250μm to irradiate the sample surface and generate plasma; Step 2: The spectral acquisition module acquires the spectral signal emitted by the plasma through the fiber optic coupling unit, and the signal is then split by the grating beam splitting unit and received by the CMOS detector array. Step 3: The real-time processing unit processes the spectral signal in real time. The background subtraction submodule uses an improved wavelet thresholding algorithm with a threshold function of λ=σ√(2logN) / √(1+αt), where σ=0.012, N=2048, and α=0.8. After background subtraction, the signal-to-noise ratio is improved to 45dB. Step 4: The feature extraction submodule uses continuous wavelet transform and adaptive peak position detection algorithm. The scale parameter of the Mexican hat wavelet is set to 32, and the peak position positioning error is <0.05nm. Step 5: The quantitative analysis module establishes a concentration prediction model based on the characteristic peak intensity and peak position information using the partial least squares regression algorithm. The internal standard element is Si, and the matrix effect correction coefficient is obtained through experimental calibration. Step 6: The environmental parameter monitoring module monitors the laboratory environmental parameters in real time: temperature 22℃±0.5℃, air pressure 1013hPa±5hPa, humidity 45%±3%RH, and uses the extended Kalman filter algorithm for data fusion. Step 7: The remote communication module transmits the analysis results to the remote server, encrypting the data using the AES-256 algorithm at a transmission rate of 100Mbps.

[0020] The analysis results showed that the Fe content was 98.4% ± 0.15%, the C content was 0.51% ± 0.03%, the Si content was 0.79% ± 0.02%, the Mn content was 0.21% ± 0.01%, the P content was 0.039% ± 0.002%, and the S content was 0.029% ± 0.001%. The analysis time was 0.8 seconds, and the quantitative error was <2%, meeting the requirements of ASTM E3057-19 standard.

[0021] Example 2: Heavy Metal Analysis of Environmental Soil The system described in this invention was used to analyze heavy metals in farmland soil samples. The samples were collected from a farmland contaminated with heavy metals, and the main heavy metal contents were: Pb 120ppm±10ppm, Cd 35ppm±5ppm, Cr 250ppm±20ppm, and As 50ppm±5ppm.

[0022] System configuration parameters: laser energy 35mJ±2mJ, spot diameter 350μm±10μm, repetition frequency 20Hz±2Hz, integration time 15ms. The spectral acquisition module uses a CMOS detector array, configured with a multimode fiber core diameter of 200μm and a numerical aperture of 0.22.

[0023] The analysis process is as follows: Step 1: The laser emission module generates laser pulses, and the beam shaping unit adjusts the spot diameter to 350μm to irradiate the surface of the soil sample and generate plasma. Step 2: The spectral acquisition module acquires the spectral signal emitted by the plasma through the fiber optic coupling unit, and after being split by the grating beam splitting unit, it is received by the CMOS detector array with a spectral resolution of 0.03 nm. Step 3: The real-time processing unit processes the spectral signal in real time. The background subtraction submodule uses an improved wavelet thresholding algorithm with σ=0.018, N=2048, and α=0.9 in the thresholding function. After background subtraction, the signal-to-noise ratio is improved to 40dB. Step 4: The feature extraction submodule uses continuous wavelet transform and adaptive peak position detection algorithm. The Mexican hat wavelet basis scale parameter is set to 48, and the peak position positioning error is <0.06nm. Step 5: The quantitative analysis module establishes a concentration prediction model based on the characteristic peak intensity and peak position information using the partial least squares regression algorithm. The internal standard element is selected as Al, and the matrix effect correction coefficient is obtained through experimental calibration. Step 6: The environmental parameter monitoring module monitors field environmental parameters in real time: temperature 28℃±1℃, air pressure 1005hPa±3hPa, humidity 60%±5%RH, and uses a polynomial fitting algorithm for parameter compensation. Step 7: The remote communication module transmits the analysis results to the agricultural environment monitoring platform using a 4G network with a speed of 50Mbps.

[0024] The analysis results showed that the Pb content was 118 ppm ± 4 ppm, the Cd content was 34 ppm ± 2 ppm, the Cr content was 245 ppm ± 6 ppm, and the As content was 49 ppm ± 2 ppm. The analysis time was 1.2 seconds, and the quantitative error was <4%, meeting the requirements of the HJ 803-2016 standard for heavy metal detection in soil.

[0025] Example 3: Compositional Analysis of Bronze Artifacts for Cultural Relics Conservation The system described in this invention was used to perform compositional analysis on ancient bronze artifact fragments. The sample was a fragment of a Western Zhou bronze artifact, and the main elemental contents were: Cu 85%, Sn 10%, Pb 3%, Zn 1%, As 0.5%.

[0026] System configuration parameters: laser energy 15mJ±1mJ, spot diameter 150μm±5μm, repetition frequency 5Hz±0.5Hz, integration time 5ms. The spectral acquisition module uses a CMOS detector array with an ultraviolet enhancement coating to improve the quantum efficiency in the ultraviolet band to 40%@300nm.

[0027] The analysis process is as follows: Step 1: The laser emission module generates a laser pulse, which is then adjusted to a spot diameter of 150μm by the beam shaping unit to irradiate the surface of the bronze fragment and generate plasma. Step 2: The spectral acquisition module acquires the spectral signal emitted by the plasma through the fiber optic coupling unit, and after being split by the grating beam splitting unit, it is received by the CMOS detector array. The spectral range is 200-900nm. Step 3: The real-time processing unit processes the spectral signal in real time. The background subtraction submodule uses an improved wavelet thresholding algorithm with σ=0.010, N=2048, and α=0.7 in the threshold function. After background subtraction, the signal-to-noise ratio is improved to 50dB. Step 4: The feature extraction submodule uses continuous wavelet transform and adaptive peak position detection algorithm. The Mexican hat wavelet basis scale parameter is set to 16, and the peak position positioning error is <0.04nm. Step 5: The quantitative analysis module establishes a concentration prediction model using partial least squares regression algorithm based on the characteristic peak intensity and peak position information. The internal standard element is selected as Fe, and the matrix effect correction coefficient is obtained through experimental calibration. Step 6: The environmental parameter monitoring module monitors the environmental parameters of the display case in real time: temperature 20℃±0.2℃, air pressure 1010hPa±2hPa, humidity 40%±2%RH, and uses the extended Kalman filter algorithm for data fusion. Step 7: The remote communication module transmits the analysis results to the cultural relics protection database using a 5G network with a speed of 500Mbps.

[0028] The analysis results showed that the Cu content was 84.8%±0.2%, Sn content was 10.1%±0.1%, Pb content was 3.0%±0.05%, Zn content was 0.95%±0.03%, and As content was 0.48%±0.02%. The analysis time was 0.6 seconds, and the quantitative error was <1.5%, which met the requirements for the analysis of cultural relic protection components.

[0029] Example 4: Analysis of Pharmaceutical Raw Material Components The system described in this invention was used to analyze the components of a pharmaceutical raw material sample. The sample was a cephalosporin antibiotic raw material, and the main element contents were: C 50%, H 5%, N 10%, O 30%, S 2%, Cl 1%.

[0030] System configuration parameters: laser energy 20mJ±1mJ, spot diameter 200μm±5μm, repetition frequency 10Hz±1Hz, integration time 8ms. The spectral acquisition module uses a CMOS detector array and is equipped with a vacuum chamber to reduce atmospheric absorption interference.

[0031] The analysis process is as follows: Step 1: The laser emission module generates laser pulses, and the beam shaping unit adjusts the spot diameter to 200μm to irradiate the surface of the drug sample to generate plasma; Step 2: The spectral acquisition module acquires the spectral signal emitted by the plasma through the fiber optic coupling unit, and after being split by the grating beam splitting unit, it is received by the CMOS detector array with a spectral resolution of 0.02nm. Step 3: The real-time processing unit processes the spectral signal in real time. The background subtraction submodule uses an improved wavelet thresholding algorithm with σ=0.015, N=2048, and α=0.85 in the thresholding function. After background subtraction, the signal-to-noise ratio is improved to 55dB. Step 4: The feature extraction submodule uses continuous wavelet transform and adaptive peak position detection algorithm. The Mexican hat wavelet basis scale parameter is set to 24, and the peak position positioning error is <0.03nm. Step 5: The quantitative analysis module establishes a concentration prediction model based on the characteristic peak intensity and peak position information using the partial least squares regression algorithm. Element B is selected as the internal standard element, and the matrix effect correction coefficient is obtained through experimental calibration. Step 6: The environmental parameter monitoring module monitors the laboratory environmental parameters in real time: temperature 23℃±0.3℃, air pressure 1008hPa±4hPa, humidity 50%±3%RH, and uses a polynomial fitting algorithm for parameter compensation. Step 7: The remote communication module transmits the analysis results to the drug quality monitoring platform via Ethernet at a rate of 1000Mbps.

[0032] The analytical results showed: C content 49.9%±0.1%, H content 4.98%±0.05%, N content 9.95%±0.05%, O content 29.8%±0.1%, S content 1.98%±0.03%, and Cl content 0.99%±0.02%. The analysis time was 0.9 seconds, and the quantitative error was <1%, meeting the requirements of the USP. <467> Drug component analysis standards requirements.

[0033] Example 5: Analysis of Geological Core Samples Elemental analysis of geological core samples was performed using the system described in this invention. The sample was a granite core, and the main element contents were: SiO2 72%, Al2O3 14%, K2O 4%, Na2O 3%, CaO 2%, and MgO 1%.

[0034] System configuration parameters: laser energy 40mJ±2mJ, spot diameter 400μm±10μm, repetition frequency 25Hz±2Hz, integration time 20ms. The spectral acquisition module uses a CMOS detector array and is equipped with a high-temperature resistant protective cover to adapt to field operation environments.

[0035] The analysis process is as follows: Step 1: The laser emission module generates laser pulses, and the beam shaping unit adjusts the spot diameter to 400μm to irradiate the surface of the core sample to generate plasma; Step 2: The spectral acquisition module acquires the spectral signal emitted by the plasma through the fiber optic coupling unit, and after being split by the grating beam splitting unit, it is received by the CMOS detector array. The spectral range is 200-1000nm. Step 3: The real-time processing unit processes the spectral signal in real time. The background subtraction submodule uses an improved wavelet thresholding algorithm with σ=0.020, N=2048, and α=0.95 in the thresholding function. After background subtraction, the signal-to-noise ratio is improved to 38dB. Step 4: The feature extraction submodule uses continuous wavelet transform and adaptive peak position detection algorithm. The Mexican hat wavelet basis scale parameter is set to 64, and the peak position positioning error is <0.07nm. Step 5: The quantitative analysis module establishes a concentration prediction model based on the characteristic peak intensity and peak position information using the partial least squares regression algorithm. The internal standard element is Ti, and the matrix effect correction coefficient is obtained through experimental calibration. Step 6: The environmental parameter monitoring module monitors the field environmental parameters in real time: temperature 30℃±1℃, air pressure 980hPa±5hPa, humidity 70%±5%RH, and uses the extended Kalman filter algorithm for data fusion. Step 7: The remote communication module transmits the analysis results to the geological exploration data center via satellite communication at a rate of 10Mbps.

[0036] The analysis results showed that the SiO2 content was 71.9%±0.3%, the Al2O3 content was 13.9%±0.2%, the K2O content was 3.98%±0.05%, the Na2O content was 2.95%±0.04%, the CaO content was 1.98%±0.03%, and the MgO content was 0.99%±0.02%. The analysis time was 1.5 seconds, and the quantitative error was <3%, meeting the requirements of the DZ / T 0279.1-2016 standard for geological sample analysis.

[0037] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A real-time processing and quantitative analysis system for spectral data based on LIBS technology, characterized in that, include: Laser emission module: Employs an Nd:YAG solid-state laser with a pulse width ≤3ns, model Spectra-PhysicsQuanta-Ray PRO-Series, operating wavelength 1064nm±1nm, repetition rate continuously adjustable from 1-1000Hz via FPGA control, single-pulse energy stability ≤±1.5%, beam quality M 2 ≤1.2; Spectrum acquisition module: includes a CMOS detector array, specifically using a Sony IMX250 sensor with a pixel size of 8μm×8μm, an effective pixel count of 5120×3840, a quantum efficiency of ≥45%@532nm band, a dynamic range of 72dB, and a dark current of ≤0.1e- / s; Real-time processing unit: Integrates a Xilinx Zynq UltraScale+ MPSoC chip to achieve parallel data processing, and is configured with: (1) Background subtraction submodule: An improved wavelet thresholding algorithm is adopted, with a threshold function λ=σ√(2logN) / √(1+αt), where σ is the noise standard deviation calculated in real time by the median absolute deviation method, N is the data length of 2048 points, and α is the dynamic adjustment coefficient that is automatically adjusted from 0.1 to 1.0 according to the signal-to-noise ratio. (2) Feature extraction submodule: Combining continuous wavelet transform and adaptive peak position detection algorithm, the continuous wavelet transform adopts Mexican hat wavelet basis, and the scale parameter is dynamically adjusted in the range of 1-64 according to the signal-to-noise ratio. The adaptive peak position detection achieves peak position positioning error <0.05nm by calculating the zero crossing point of the first derivative and using cubic spline interpolation. (3) Data caching submodule: A dual-port RAM with a capacity of 256KB is used to achieve seamless connection of data streams; (4) Outlier detection submodule: Based on the improved Grubbs criterion, it dynamically identifies and removes outlier data points. The improved Grubbs criterion judges the outlier by calculating the deviation of the data points and combining it with the 95% confidence threshold. Quantitative Analysis Module: Includes a multi-element standard curve library and a matrix effect correction unit. The multi-element standard curve library is obtained through experimental calibration and contains standard curves for 20 common elements such as Fe, Cu, Al, C, and O. The matrix effect correction unit uses the internal standard method combined with a multivariate correction algorithm. The internal standard element is Si. The multivariate correction algorithm uses partial least squares regression (PLSR) to establish a concentration prediction model.

2. The real-time processing and quantitative analysis system for spectral data based on LIBS technology according to claim 1, characterized in that, The laser emitting module also includes: Laser energy monitoring unit: Uses a Thorlabs S120C silicon photodiode to monitor laser energy in real time, with energy stability ≤ ±1% and response time ≤ 1μs; Beam shaping unit: includes a tunable beam expander (Thorlabs BE02-1064) and an aperture (ThorlabsSM1D12), enabling continuous adjustment of the beam diameter within the range of 50-500μm with an adjustment accuracy of 1μm; Laser triggering unit: A high-speed optocoupler is used to synchronize laser triggering and data acquisition, with a synchronization accuracy of ≤10ns.

3. The real-time processing and quantitative analysis system for spectral data based on LIBS technology according to claim 1, characterized in that, The spectral acquisition module is configured with: Fiber optic coupling unit: The fiber used is Thorlabs FT200EMT multimode fiber with a core diameter of 200μm±5μm, a numerical aperture of 0.22±0.02, and a fiber length of 5m±0.1m. Grating beam splitting unit: adopts Richardson Gratings 53-*-120R blazed grating with a line density of 1200 lines / mm, blaze wavelength of 500nm±10nm, spectral resolution ≤0.03nm, and diffraction efficiency ≥80%@500nm band; Focusing lens group: The doublet lens is used to achieve efficient focusing of spectral signals, with a focal length of 100mm, a light-transmitting aperture of 30mm, and aberration correction to the diffraction limit.

4. The real-time processing and quantitative analysis system for spectral data based on LIBS technology according to claim 1, characterized in that, The real-time processing unit further includes: Algorithm acceleration module: Hardware acceleration of wavelet transform and PLSR algorithm is implemented using hardware description language, with a processing speed ≥100MSPS; Dynamic parameter adjustment module: Automatically adjusts wavelet transform scale and threshold parameters based on real-time signal-to-noise ratio, with an adjustment range covering scales from 1 to 64 levels; Data compression module: Uses lossless compression algorithm to compress and store spectral data with a compression ratio ≥2:

1.

5. The real-time processing and quantitative analysis system for spectral data based on LIBS technology according to claim 1, characterized in that, The quantitative analysis module includes: Matrix effect correction unit: The internal standard method combined with a multivariate correction algorithm is used. The correction coefficients are obtained through experimental calibration. The number of calibration samples is ≥100 and the calibration error is <3%. Dynamic concentration display unit: Equipped with an LCD touch screen model Newhaven Display NHD-7.0-800480EF, which displays elemental concentrations and confidence levels of analysis results in real time. The confidence level is calculated using the Monte Carlo simulation method. Results Validation Module: The cross-validation algorithm is used to assess the reliability of the analysis results, with a cross-validation error of <2%.

6. The real-time processing and quantitative analysis system for spectral data based on LIBS technology according to claim 1 further includes an environmental parameter monitoring module, wherein the environmental parameter monitoring module includes: Temperature sensor: PT1000 platinum resistance thermometer, Heraeus C220, measuring range -20℃ to 80℃, accuracy ±0.2℃, response time ≤1s; Barometric pressure sensor: Bosch BMP280, measuring range 800-1100 hPa, accuracy ±0.5 hPa, response time ≤1 ms; Humidity sensor: Honeywell HIH-5030, measurement range 0-100%RH, accuracy ±2%RH, response time ≤10s; Parameter compensation unit: A polynomial fitting algorithm is used to achieve real-time compensation of environmental parameters for spectral data, and the fitting order is automatically adjusted according to changes in environmental parameters.

7. The real-time processing and quantitative analysis system for spectral data based on LIBS technology according to claim 1, characterized in that, The environmental parameter monitoring module uses the extended Kalman filter algorithm for data fusion to achieve real-time correction of environmental parameters. The extended Kalman filter algorithm achieves optimal estimation by combining historical data and real-time measurement data of environmental parameters with state equations and observation equations. The state transition matrix is ​​automatically updated according to the rate of change of environmental parameters.

8. The real-time processing and quantitative analysis system for spectral data based on LIBS technology according to claim 1 further includes a remote communication module, the remote communication module comprising: Wireless transmission module: Supports Ethernet and 4G / 5G wireless transmission protocols. The Ethernet interface uses Intel I210-IT with a transmission rate of 1000Mbps. The 4G / 5G module uses Quectel EG25-G and supports the LTE Cat 4 standard. Configured data encryption unit: AES-256 algorithm is used to achieve data transmission security protection, encryption chip model Microchip ATECC608A, encryption speed ≥200Mbps; Remote control unit: Supports remote parameter configuration and system diagnostics, and uses the SSH protocol to achieve secure remote access.

9. The real-time processing and quantitative analysis system for spectral data based on LIBS technology according to any one of claims 1 to 8 adopts a modular design, and each module realizes plug-and-play functionality through a standard interface. The standard interface adopts the CPCI bus standard, with a data transmission rate of ≥1Gbps, and the interface hot-swappable supports online system maintenance.

Images