부산대학교 도서관

This paper presents the development of adaptive neuro-fuzzy inference systems (ANFIS) model for estimating the modal damping ratio of impact-damped flexible beams. The experimental results obtained from the literature are used as training and testing data for developing the model. Five of six parameters namely the gap, mass, modal amplitude, frequency, and peak value of the imaginary part of the frequency response functions were used as input, whereas the modal damping ratio of impact-damped flexible concrete beams was used as output parameter for the model. The Sugeno-type fuzzy rules were constituted with the Gaussian-type membership functions in the model. In the Sugeno-type fuzzy inference model, the modal damping ratio of impact-damped flexible concrete beams is estimated according to these input parameters. ANFIS model results for modal damping ratio are in good agreement with the experimental results than those obtained from formula presented in the literature. [ABSTRACT FROM AUTHOR]