Multivariate Temperature Calibrations Based on Upconversion Fluorescence Spectra of Holmium and Ytterbium-Doped Alumofluoride Glasses

Abstract

Machine learning methods (principal component analysis, support vector regression and partial least squares) are used to calibrate the temperature based on the spectra of Ho³⁺ upconversion fluorescence in alumofluoride glasses in the wavelength ranges of 505–560 nm and 605–670 nm for the temperature range from 30 to 105°C along with the traditional ratiometric method. Among the methods considered, the best calibration models were obtained by the partial least squares with searching the combination of moving window partial least squares with the fluorescence spectra of 95MgCaSrBaYAl₂F₁₄-5Ba(PO₃)₂:5%Yb³⁺,0.1%Ho³⁺ glass. The values of absolute sensitivity of 0.021 K^–1 and temperature uncertainty of 0.12 K achieved by machine learning methods significantly exceed the characteristics of the ratiometric method.