In this study we concentrate on neural networks in the form of algorithms, and on their property to learn by examples. The possibility of learning by examples in-cludes also an enormous ability of information extraction. As the input data we have used the regional seismograms which record the ground particle velocities. Application of the networks for evaluation of the two seismic parameters has been demonstrated: first to calculate the magnitude (or seismic moment) of the seismic event directly from the velocity seismograms, without prior converting them to the displacement seismograms, and second, to locate seismic source by using only a single seismogram. Both applications are very convenient due to their speed of cal-culation and the lack of spectral methods - in the first case, and the good precision despite the unusual approach in the second case. To prepare the input data for the neural networks a preprocessing has been applied, which can be described as filter-ing of a seismogram by the filters bank.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.