|
doi:10.3850/978-981-08-5118-7_openplenary
On Current/Emerging Approaches for Stochastic Engineering Analyses
Pol D. Spanos
LB Ryon Chair in Engineering, Department of Mechanical Engineering and Materials Science,
Rice University, Houston 77005, USA.
spanos@rice.edu
ABSTRACT
The lecture will focus on certain approaches which are currently available for coping with uncertainty in the process of engineering design. Particular attention will be devoted to addressing the issue of evolving frequency content in the excitation of dynamic systems, and of the determination of their related responses in the presence of nonlinearity. In this regard, the limitations of the traditional Fourier analysis in capturing non-stationary and non-local features in signals will be discussed. This will be used as a departure point for introducing the appealing localization features of modern techniques such as the ones based on wavelets, chirplets, and intrinsic modes. The techniques can be used for the representation both of individual signals and of stochastic processes with evolving frequency content. This perspective on some of the options of modern stochastic engineering analysis will be supplemented with a discussion on efficient algorithms for representing multi-variate random processes in the context of Monte Carlo system reliability studies. Finally, techniques for calculating the response of nonlinear dynamical systems to excitations described by wavelets and other schemes will be discussed. In conjunction with this theme, the use of stochastic dynamics for determining appropriate parameters for risk assessment of nonlinear systems will be discussed. Examples from practical applications will be included.
Keywords: Chirplets, Intrinsic Mods, Wavelets, Nonlinearity, Dynamic System.
|
