Search

The eigensystem realization algorithm (ERA) is a data-driven approach for subspacesystem identification and is widely used in many areas of engineering. However, the computationalcost of the ERA is dominated by a step that involves the singular value decomposition (SVD) ofa large, dense matrix with block Hankel structure. This paper develops computationally efficientalgorithms for reducing the computational cost of the SVD step by using randomized subspaceiteration and exploiting the block Hankel structure of the matrix. We provide a detailed analysis ofthe error in the identified system matrices and the computational cost of the proposed algorithms.We demonstrate the accuracy and computational benefits of our algorithms on two test problems:the first involves a partial differential equation that models the cooling of steel rails, and the secondis an application from power systems engineering. [ABSTRACT FROM AUTHOR]