Joint Kalman–Haar Algorithm Applied to Signal Processing

Under the analysis of signals disturbed by noise, in this paper we propose a working methodology aimed to seize the best estimate of combining Kalman filtering with the characterization that is achieved by applying a multiresolution analysis (MRA) using wavelets. From the standpoint of Kalman filter...

Descripción completa

Autores Principales: Viegener, Alejandro, Sirne, Ricardo O., Serrano, Eduardo P., Fabio, Marcela, D'Attellis, Carlos E.
Formato: Artículo
Idioma: Español
Publicado: 2015
Acceso en línea: http://revistas.ucr.ac.cr/index.php/matematica/article/view/2103
http://hdl.handle.net/10669/13005
Sumario: Under the analysis of signals disturbed by noise, in this paper we propose a working methodology aimed to seize the best estimate of combining Kalman filtering with the characterization that is achieved by applying a multiresolution analysis (MRA) using wavelets. From the standpoint of Kalman filtering this combined procedure is quasi-optimal, but the change to be made allows the simultaneous implementation of a scheme of wavelet denoising; with this decreases the computational cost of applying both procedures separately. Our proposal is to process the signal by successive non-overlapping intervals, combining the process for calculating the optimal filter with a MRA using the Haar wavelet. The method takes advantage of the combined use of both tools (Kalman-Haar) and is free from edge problems related to the signal segmentation.