Devising a computer method to recognize and analyze spectrometric signals parameters

The object of this study is computerized systems for measuring the parameters of spectrometric signals digitized using special hardware. The task addressed in the research is to improve the process of filtering the usable pulse signal from noise and increase the accuracy of measuring pulse parameters by devising a new method of analysis. In order to verify the performance of the new method in comparison with several already known ones, input data arrays with predetermined parameters were prepared using computer simulation. A special algorithm was also developed to verify each detected pulse. As a result, the main characteristics of the methods, such as signal recognition accuracy and data processing speed, were obtained for several scenarios with different durations of modeling process and different pulse generation intensities. Comparative performance metrics were provided for all described software analysis methods. Ultimately, in the studied scenarios, the devised method showed better recognition ability than the considered alternative methods. The key features of the proposed method are the use of software filters built on the basis of the application of Fast Discrete Fourier Transform (FDFT) algorithms and further computer processing of the signal using a mechanism for correcting the amplitudes of superimposed pulses. This makes it possible to filter the signal from noise without significantly changing the usable component and to more accurately determine the amplitudes in case of their frequent superposition. In practice, the devised method could be used to improve existing and design new computer systems of spectral analysis