Correlation analysis of periodic signals in the noise with the use of the Arduino platform

摘要

A correlation analysis of the periodic signal with noise received from the impulse sensor “PulseSensor”, using the Arduino platform and a signal frequency is determined. The novelty is that the proposed mathematical tool, in contrast to the classical Fourier analysis (development in the Fourier series of the investigated signal for each particular patient), makes it possible to determine the statement of its disease. A simple and reliable methodology (without attempt to present simulated process by a system of nonlinear differential equations that does not apriori accurately reflect the process itself and the exact solution of which in the general formulation is impossible to receive) finding of useful (determinant) signal by the maximum of its correlation coefficient is proposed. Phase discretization is necessary because the starting point of measurement is random, and the correlation coefficient of the investigated and reference signal with a minimum difference of their phases is informally valuable. This approach makes possible to highlight the useful signal at high noise levels which exceeds the level of useful signal.