ABSTRACT
To improve prediction performance and reduce artifacts in Raman spectra, we developed an eXtreme Gradient Boosting (XGBoost) preprocessing method to preprocess the Raman spectra of glucose, glycerol and ethanol mixtures. To ensure the robustness and reliability of the XGBoost preprocessing method, an X-LR model (which combined XGBoost preprocessing and a linear regression (LR) model) and a X-MLP model (which combined XGBoost preprocessing and a multilayer perceptron (MLP) model) were developed. These two models were used to quantitatively analyze the concentrations of glucose, glycerol and ethanol in the Raman spectra of mixed solutions. The proportion map of hyperparameters was firstly used to narrow down the search range of hyperparameters in the X-LR and the X-MLP models. Then the correlation coefficients (R2), root mean square of calibration (RMSEC), and root mean square error of prediction (RMSEP) were used to evaluate the models' performance. Experimental results indicated that the XGBoost preprocessing method achieved higher accuracy and generalization capability, with better performance than those of other preprocessing methods for both LR and MLP models.
ABSTRACT
Photoacoustic spectroscopy (PAS) has been rapidly developed and applied to different detection scenarios. The acoustic pressure detection is an important part in the PAS system. In this paper, an ultrahigh sensitivity Fabry-Perot acoustic sensor with a T-shaped cantilever was proposed. To achieve the best acoustic pressure effect, the dimension of the cantilever structure was designed and optimized by finite element analysis using COMSOL Multiphysics. Simulation results showed that the sensitivity of such T-shaped cantilever was 1.5 times higher than that based on a rectangular cantilever, and the resonance frequency of T-shaped cantilever were able to modulate from 800â¯Hz to 1500â¯Hz by adjusting the multi-parameter characteristics. Experimental sensing results showed that the resonance frequency of T-shaped Fabry-Perot acoustic sensor was 1080â¯Hz, yielding a high sensitivity of 1.428 µm/Pa, with a signal-to-noise ratio (SNR) of 84.8â¯dB and a detectable pressure limit of 1.9 µPa/Hz1/2@1â¯kHz. We successfully used such acoustic sensor to measure acetylene (C2H2) concentration in the PAS. The sensitivity of PAS for C2H2 gas was 3.22â¯pm/ppm with a concentration range of 50â¯ppm â¼100â¯ppm, and the minimum detection limit was 24.91ppb.