Virtual Vibration Testing of External Stores

The difficulty in developing system identification and synthesis methodologies that are valid for broad classes of dynamical systems is due to the well-recognized highly individualistic nature of nonlinear systems which restricts the unifying dynamical features that are amenable to this process. Our proposed technique relies solely on direct time series measurement and post processing, and leads to global as well as local identification of a broad class of dynamical systems. Key to our method is the slow/fast partition of time series measurements which leads to the identification of the dominant fast frequencies in the measured time series (which ultimately govern the dimensionality of the dynamics), to identification of global features of the dynamics in the frequency – energy domain, and to reduced slow flow models. We employ advanced post processing computational algorithms, namely, wavelet transform (WT), Hilbert transform (HT) and empirical mode decomposition (EMD) unified by a solid theoretical framework. Our technique successfully integrates analytical concepts and computational algorithms in a synergistic methodology.