PERIODIC SIGNAL FILTRATION USING DIGITAL FILTERING SYSTEM CALCULATION OPTIMIZED BY APPROXIMATION

Abstract

The article assesses the filtration of periodic signals of different shapes with the use of optimized calculation of discrete low pass filtering system (LPFS) orders. The calculation is based on the minimization of the difference between approximated values and filtered values of a noisy signal. To assess the feasibility of the calculation of the optimized discrete filtering system order, a series of experiments has been carried out on generation and filtration of noisy periodic signals with different signal shapes and various signal-to-noise ratios. For all of the noisy signals, optimal (relative to the original noise-free signals) and optimized (relative to the approximated signals) LPFS orders based on a minimal mean squared error (MSE) have been calculated.  After calculation of MSE between filtered and original signals, it was observed that the optimal and optimized filters allow us to get comparable noisy signal filtration quality. The article is completed by the conclusions about the

possibility of using the discrete filtering system order calculation optimized by approximation.