The Canonical Variate Analysis has been applied in many circumstances as a powerful dynamic identification tool. Its capability of overcoming the high modal density matter is also improved by implemented special tools such as Probability Density Function and Modal Assurance Criterion stabilisation, making the CVA technique very reliable to monitor systems with unknown imput. Such a situation frequently occours due to the practical impossibility of measuring the imput dynamic load, as in the case of a bridge excited by traffic or in the case of a car running over a rough pavement. In this paper it is shown how the data processing can be improved by following some guidelines, based on an analytical approach to the model order choice, problem coupled with the already mentioned tools, i.e the PDF and the MAC functions.
An error estimate for CVA / Fasana, A.; Garibaldi, L.; Marchesiello, S.; Sorrentino, Silvio. - STAMPA. - 204-2:204-205(2001), pp. 261-270. (Intervento presentato al convegno 4th International Conference on Damage Assessment of Structures (DAMAS 2001) tenutosi a CARDIFF, WALES nel JUN 25-28, 2001) [10.4028/www.scientific.net/KEM.204-205.261].
An error estimate for CVA
SORRENTINO, Silvio
2001
Abstract
The Canonical Variate Analysis has been applied in many circumstances as a powerful dynamic identification tool. Its capability of overcoming the high modal density matter is also improved by implemented special tools such as Probability Density Function and Modal Assurance Criterion stabilisation, making the CVA technique very reliable to monitor systems with unknown imput. Such a situation frequently occours due to the practical impossibility of measuring the imput dynamic load, as in the case of a bridge excited by traffic or in the case of a car running over a rough pavement. In this paper it is shown how the data processing can be improved by following some guidelines, based on an analytical approach to the model order choice, problem coupled with the already mentioned tools, i.e the PDF and the MAC functions.
Pubblicazioni consigliate
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris
