Uncertainties Consideration in Empirical Frequency Response Function Data for Damage Identification Based On Artificial Neural Network

The modern application of frequency response function (FRF) with artificial neural networks (ANN) has become one the leading methods in vibration-based damage detection approach. However, since full-size empirically obtained FRF data is used as ANN input, a broad composition input layer series would occur. Consequently, principal component analysis (PCA) adopted to compress magnitude. Despite this, PCA alone unable select important features effectively, due exceedingly size addition existing uncertainties. Therefore, this study proposed merger non-probabilistic and approach by considering uncertainties effect inefficiency using empirical data. from steel truss bridge structure. results show that PoDE values above 95% are measured at particular executed locations DMI severity actual locations. Overall, method capable on for structural identification.