Data processing and estimation of measurements errors in intracavity laser spectroscopy

摘要

Methods of absorption spectra processing in intracavity laser spectroscopy (ICLS) are investigated for the case of very low quantities of cesium in water solutions. Common procedures in such cases require fitting or filtering of the laser generation broadband spectrum, search for the investigated atomic or molecular absorption lines, calculating the examined substance concentration in quantitative analysis or its state determination in quality analysis. In this study various methods of absorption spectra processing in the case of atomic cesium line λ = 455.531 nm in the intracavity laser spectrometer are examined. Methods based on digital filtering and approximation were tested. The artificial neural network algorithm was tried as a tool to reduce the apparatus error. For the trace concentration definition the mean-square error was decreased to a value less then 9%.