Physical interpretation of the proper orthogonal modes using the singular value decomposition

Proper orthogonal decomposition is a statistical pattern analysis technique for finding the dominant structures, called the proper orthogonal modes, in an ensemble of spatially distributed data. While the proper orthogonal modes are obtained through a statistical formulation, they can be physically interpreted in the field of structural dynamics. The purpose of this paper is thus to provide some insights into the physical interpretation of the proper orthogonal modes using the singular value decomposition. (C) 2002 Academic Press.