A neural network based procedure is proposed for the crosstalk prediction in twisted bundles. The nonuniform bundle is approximated as a cascade of uniform sections and the neural network is suitably trained with few cross section configurations to learn the behavior of the per unit length parameters. The procedure allows a fast and accurate prediction tool.
Crosstalk Prediction in Twisted Bundles by a Neural Approach / Cannas, B.; Fanni, A.; Maradei, Francescaromana. - STAMPA. - (2002), pp. 638-641. (Intervento presentato al convegno 3rd International Symposium on Electromagnetic Compatibility tenutosi a Beijing, China nel May 21-24, 2002) [10.1109/ELMAGC.2002.1177512].
Crosstalk Prediction in Twisted Bundles by a Neural Approach
MARADEI, Francescaromana
2002
Abstract
A neural network based procedure is proposed for the crosstalk prediction in twisted bundles. The nonuniform bundle is approximated as a cascade of uniform sections and the neural network is suitably trained with few cross section configurations to learn the behavior of the per unit length parameters. The procedure allows a fast and accurate prediction tool.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.