Stabilization and regularization in time-domain inverse filtering
The aim of the paper is to study two important problems involving stabilization and regularization. For non-minimum phase systems, a correct stabilization becomes an important part of filtering and contributes significantly to the quality of the input signal estimation. The paper proposes two algorithms for inverse filter stabilization, both leading to the noncausality of the inverse filter. One of the stabilization algorithms, involving pole transposition and phase correction shows the most promising properties. Appropriate regularization is proved to be necessary in order to reduce the undesirable effect of ill-conditioning on the result of inverse filtering. The proposed method of regularization corrects the location of the poles of the inverse filter transmittance and yields promising results. For moderately low signal/noise ratios, its effectiveness is only slightly lower, while for high signal/noise ratio values, the proposed regularization effectiveness is slightly higher than the one of the classical algorithm. An inherent advantage of the method is that regularization is performed at the stage of inverse filtering, thus enabling elimination of an additional regularization step.
- Regularization theory in image restoration-the stabilizing functional approach
- Filtering and inverse problems solving
- On regularization of algorithms of information processing in control problems
- Stabilization of spectral methods for the analysis of singular systems using piecewise constant basis functions
- Filtered inverse systems
This page was built for publication: Stabilization and regularization in time-domain inverse filtering
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1269841)
