Randomized model order reduction
- 65 Downloads
The singular value decomposition (SVD) has a crucial role in model order reduction. It is often utilized in the offline stage to compute basis functions that project the high-dimensional nonlinear problem into a low-dimensional model which is then evaluated cheaply. It constitutes a building block for many techniques such as the proper orthogonal decomposition (POD) and dynamic mode decomposition (DMD). The aim of this work is to provide an efficient computation of low-rank POD and/or DMD modes via randomized matrix decompositions. This is possible due to the randomized singular value decomposition (rSVD) which is a fast and accurate alternative of the SVD. Although this is considered an offline stage, this computation may be extremely expensive; therefore, the use of compressed techniques drastically reduce its cost. Numerical examples show the effectiveness of the method for both POD and DMD.
KeywordsNonlinear dynamical systems Proper orthogonal decomposition Dynamic mode decomposition Randomized linear algebra
Mathematics Subject Classification (2010)65L02 65M02 37M05 62H25
Unable to display preview. Download preview PDF.
This study is supported by the Department of Energy (grant no. DE-SC0009324) and the U.S. Air Force Office of Scientific Research (FA9550-15-1-0385).
- 2.Barrault, M., Maday, Y., Nguyen, N.C., Patera, A.T.: An empirical interpolation method: application to efficient reduced-basis discretization of partial differential equations Comptes Rendus Mathematique, 339, pp. 667–672 (2004)Google Scholar
- 8.Duersch, J., Gu, M. (2015)Google Scholar
- 9.Erichson, N.B., Voronin, S., Brunton, S.L., Kutz, J.N.: Randomized matrix decompositions using R, arXiv:1608.02148 (2016)
- 14.Isaac, T., Petra, N., Stadler, G., Ghattas, O.: Scalable and efficient algorithms for the propagation of uncertainty from data through inference to prediction for large-scale problems, with application to flow of the Antarctic ice sheet. J. Comp. Phys. 296, 348–368 (2015)MathSciNetCrossRefzbMATHGoogle Scholar
- 16.Kutz, J.N., Brunton, S., Brunton, B., Proctor, J.: Dynamic mode decomposition: Data-driven modeling of complex systems. SIAM Press (2016)Google Scholar
- 19.Martinsson, P.-G.: factorizations, blocked rank-revealing QR: how randomized sampling can be used to avoid single-vector pivoting. arXiv:1505.08115 (2015)
- 21.Martinsson, P.-G.: Randomized methods for matrix computations and analysis of high dimensional data, arXiv:1607.01649 (2016)
- 22.Martinsson, P.-G., Quintana-Orti, G., Heavner, N.: randUTV: A blocked randomized algorithm for computing a rank-revealing UTV factorization, arXiv:1703.00998 (2017)
- 27.Szlam, A., Kluger, Y., Tygert, M.: An implementation of a randomized algorithm for principal component analysis, arXiv:1412.3510(2014)
- 29.Volkwein, S.: Model Reduction Using Proper Orthogonal Decomposition. Lecture Notes, University of Konstanz (2013)Google Scholar
- 30.Voronin, S., Martinsson, P.-G.: RSVDPACK: Subroutines for computing partial singular value decompositions via randomized sampling on single core, multi core, and GPU architectures, arXiv:1502.05366 (2015)