Abstract:An adaptive baseline correction method was proposed by combining Particle Swarm Optimization(PSO)and asymmetrically reweighted Penalized Least Squares(arPLS)through fitness function to reduce or eliminate the influence of continuous background radiation,random noise,and sample matrix effect on the characteristic spectrum in the in-situ analysis of long-range Laser-Induced Breakdown Spectroscopy(LIBS).The PSO-arPLS approach is intended to increase the remote LIBS's analytical capacity. The method adds the regular function to the loss function,turning the limited problem into an unconstrained problem,and uses the"asymmetric weighting"approach to accomplish the adaptive baseline correction goal.PSO and arPLS were combined by fitness function and applied to an aluminum-based alloy with trace metal elements as the research sample.Particle swarm automatically discovered the optimal parameters of arPLS fitting baseline to achieve the balance of weight vector and smoothing parameters in baseline signal.The spectral Signal-to-Noise Ratio(SNR)and noise reduction effect of the short-wave and long-wave spectral segments were examined,using the gathered 6061 series aluminum base alloy LIBS as an example.PSO-arPLS was then compared with the conventional airPLS and asPLS calibration methods.Finally,the kernel Support Vector Machine(SVM)model is trained using the original LIBS data set of aluminum base alloy and the data set after baseline correction using the aforementioned three methods,and the confusion matrix of the model is analyzed to confirm the validity of the suggested baseline correction method. The results demonstrate that the suggested PSO-arPLS approach can not only reduce the spectral baseline fluctuation but also increase the spectral SNR and boost the spectral dynamic range when compared to conventional airPLS and asPLS calibration methods.PSO-arPLS can effectively preserve the LIBS signal with spectral characteristics after correction.In contrast,the fitting baselines proposed by airPLS and asPLS methods lead to serious loss of LIBS characteristic peak region and low fitness for eliminating LIBS noise.Therefore,when correcting LIBS spectra with low SNR,the baseline trend can basically overlap with the uncharacterized peak region,and the PSO-arPLS method also shows a good effect in noise elimination.Continuous background radiation is successfully monitored,and the fitted baseline lies within the cross-range of the characteristic spectrum and noise signal.In conclusion,the PSO-arPLS algorithm performs well in noise reduction and corrects the low SNR LIBS spectral baseline region,which nearly overlaps with the non-characteristic peak region. Three different types of aluminum-based alloys doped with comparable trace elements were employed as research objects in this paper.Under the same experimental conditions,the fitting baseline was utilized to analyze and evaluate the variation trend of the spectral SNR addressed by the airPLS and asPLS algorithms.A cubic kernel SVM fine classification model for aluminum-base alloys was created to confirm the efficacy of the suggested approach.The independent test set's confusion matrix classification had 100%accuracy.The kernel SVM model was trained using the original LIBS data set and the suggested baseline corrected data set,and the model's confusion matrix was examined.The outcomes demonstrate that the PSO-arPLS technique could more effectively classify and identify the LIBS data.The 11.8%increase in classification accuracy further demonstrates the PSO-arPLS method's beneficial effects on data analysis.PSO-arPLS approach has strong noise robustness and can overcome the effects of continuous background radiation and LIBS noise at a great distance.Additionally,the proposed method for LIBS adaptive baseline correction can be used in real-world industrial contexts and significantly enhances remote LIBS's capacity for qualitative analysis.

Particle Swarm Optimization (PSO) for Improving the Accuracy of ChemCam LIBS Sub-Model Quantitative Method

Parameter Estimation of Catalytic Cracking Model Using PSO Algorithm

A Novel Cross-Instrument Spectral Harmonization Approach for Mars In-situ LIBS Data

MarSCoDe Martian Material Analysis Based on a PSO-SVR Approach

Control Parameterization‐Based Adaptive Particle Swarm Approach for Solving Chemical Dynamic Optimization Problems

Investigation into the Affect of Chemometrics and Spectral Data Preprocessing Approaches upon Laser-Induced Breakdown Spectroscopy Quantification Accuracy Based on MarSCoDe Laboratory Model and MarSDEEP Equipment

Elastic Particle Swarm Optimization for MarSCoDe Spectral Calibration on Tianwen-1 Mars Rover

Adaptive Baseline Correction Method for Laser-induced Breakdown Spectroscopy

Multi-Parameter Auto-Tuning Algorithm for Mass Spectrometer Based on Improved Particle Swarm Optimization

Calibration Transfer for ChemCam Spectral Data from Different Laser-Induced Breakdown Spectrometers Via a Deep Extreme Learning Machine

Application of Least Square Support Vector Machine Based on Particle Swarm Optimization in Quantitative Analysis of Gas Mixture

Incorporating PLS Model Information into Particle Swarm Optimization for Descriptor Selection in QSAR/QSPR

LIBS Spectrum Baseline Correction Method based on Non-Parametric Prior Penalized Least Squares Algorithm

Simultaneous Ultrasonic Parameter Estimation of a Multi-Layered Material by the PSO-based Least Squares Algorithm Using the Reflection Spectrum

Concentration retrieval in a calibration-free wavelength modulation spectroscopy system using particle swarm optimization algorithm

Estimating Scale Discrepancy in Bayesian Model Calibration for ChemCam on the Mars Curiosity Rover

A novel hybrid feature selection strategy in quantitative analysis of laser-induced breakdown spectroscopy

A calibration framework for the microparameters of the DEM model using the improved PSO algorithm

Scanning Micromirror Calibration Method Based on PSO-LSSVM Algorithm Prediction

Response Surface Methodology for Optimizing LIBS Testing Parameters: A Case to Conduct the Elemental Contents Analysis in Soil

From machine learning to transfer learning in laser-induced breakdown spectroscopy analysis of rocks for Mars exploration