Abstract
In this paper, terahertz attenuated total reflection (THz-ATR) technology is used to measure formaldehyde solutions with a concentration of 0% to 10%, and the refractive index and absorption coefficient of the formaldehyde solution in the range of 0.2~1.2THz are selected to establish partial least squares (PLS) quantitative detection model and back propagation neural network (BPNN) quantitative detection model. Among them, the correlation coefficient (R) and root mean square error (RMSE) of the PLS model based on the refractive index are 98.99% and 0.4602, respectively; the R and RMSE of the PLS model based on the absorption coefficient are 99.05% and 0.4464, respectively; the R of the BPNN model based on the refractive index is 99.80%, and the RMSE is 0.2407; the R of the BPNN model based on the absorption coefficient is 99.93%, and the RMSE is 0.1182.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have