Affiliation:
1. Free University of Colombia
2. University of Quindío
Abstract
Abstract
A characterization of seismic signals was carried out, using signal analysis techniques in the time-frequency domain, in order to identify the relevant characteristics of five seismic sources, excluding any knowledge of the seismogenic source and extracting and recognizing characteristic patterns that would allow classification of the earthquakes and to allow any possible qualitative explanation of a likely source that originated the event. For this purpose, a database consisting of 293 signals that were previously located and assigned to mentioned seismic sources. Finally, a basic classification system was designed through a neural network multilayer perceptron that assigns the earthquake to one of the seismic sources established according to the features and patterns that have been detected using time-frequency analysis.
Publisher
Research Square Platform LLC
Reference22 articles.
1. Large-scale seismic signal analysis with Hadoop;Addair TG;Computers and Geosciences,2014
2. Application of wavelet for seismic wave analysis in Kathmandu Valley after the 2015 Gorkha earthquake, Nepal;Adhikari B;Geoenvironmental Disasters,2020
3. Time-frequency analysis via deconvolution with sparsity constraints;Bonar D;GeoCanada, Expanded Abstracts,2010
4. The marginals and time-frequency distributions;Cohen L;Advanced Signal Processing Algorithms, Architectures, and Implementations XIII,2003
5. Linear perturbations of the Wigner distribution and the Cohen class;Cordero E;Analysis and Applications,2020