In this work simulated ultrasonic waveforms in a concrete specimen obtained by a software based on finite element method were used to develop an automatic inspection method. A piezoelectric transducer is used to generate stress waves that are reflected by voids. Then the waves are received by another transducer set at a fixed distance from the first one on the same specimen surface. Time and frequency features has been extracted from the waveforms, the most significant features have been chosen by a genetic feature selection and the classification performances were estimated referring to a k-NN classifier.
Genetic Feature Selection and Statistical Classification of Voids in Concrete Structure / Acciani, Giuseppe; G., Fornarelli; D., Magarielli; D., Maiullari. - Volume AIDSS: Artificial intelligence and decision support systems:(2008), pp. 231-234. (Intervento presentato al convegno 10th International Conference on Enterprise Systems Information, ICEIS 2008 tenutosi a Barcelona, Spain nel June 12-16, 2008) [10.5220/0001692802310234].
Genetic Feature Selection and Statistical Classification of Voids in Concrete Structure
ACCIANI, Giuseppe;
2008-01-01
Abstract
In this work simulated ultrasonic waveforms in a concrete specimen obtained by a software based on finite element method were used to develop an automatic inspection method. A piezoelectric transducer is used to generate stress waves that are reflected by voids. Then the waves are received by another transducer set at a fixed distance from the first one on the same specimen surface. Time and frequency features has been extracted from the waveforms, the most significant features have been chosen by a genetic feature selection and the classification performances were estimated referring to a k-NN classifier.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.