Novel Computational Methods for Solving High-Dimensional Random Eigenvalue Problems

TITLE:

Novel Computational Methods for Solving High-Dimensional Random Eigenvalue Problems

DATE:

Friday, September 5th, 2014

TIME:

3:30 PM

LOCATION:

GMCS 214

SPEAKER:

Dr. Vaibhav Yadav. Department of Aerospace Engineering at San Diego State University

ABSTRACT:

Random eigenvalue problems (REPs) have far-reaching applications in fields as diverse as quantum physics, number theory, multivariate statistics, signal processing and communication, finance, computational biology and, of course, mechanics. In mechanics, REPs prominently appear in structural dynamics and structural stability, among others. This presentation describes three novel, highly efficient computational methods for solving high-dimensional REPs. Predicated over orthogonal polynomial-based decomposition, these methods proved to be highly effective in efficiently countering the challenges of curse-of-dimensionality, high-nonlinearity and high-input-variability in a REP. Several large and complex REPs occurring in real world engineering applications that were successfully tackled by the proposed methods are also illustrated.

HOST:

Dr. Satchi Venkataraman

DOWNLOAD: