Zpracování signálů z optovláknových mřížkových senzorů pro monitorování dechové aktivity

Abstract

This thesis deals with the processing of signals obtained from the developed sensor composed of optical fiber with implemented Bragg grating. 3 methods of advanced data processing were used for processing. These are the empirical modal decomposition, the Savitzky-Golay filter and the wavelet transform method. The use of signal processing methods plays a crucial role in the subsequent signal analysis, in which the time positions of the recorded breaths are extracted. The resulting values are then compared with reference values obtained by extracting breath time values from a reference signal measured by the chest band using a piezoelectric sensor. The aim of this work is to assess whether the developed fiber optic sensor achieves sufficient results in comparison with the reference meter and to compare the methods of advanced signal processing. The resulting algorithms for signal processing, analysis and subsequent evaluation of results are implemented in a user application that provides a user-friendly and intuitive environment in which the user is able to easily change important parameters on which the functions of individual methods depend. The most efficient method for signal processing in this case was the Savitzky Golay filter, which achieved the best average accuracy results.