A parallel and streaming dynamic mode decomposition algorithm with finite precision error analysis for large data

DOI10.1016/J.JCP.2018.12.012MaRDI QIDQ2214545zbMATH OpenOpenAlexFDO

Authors Sreevatsa Anantharamu, Krishnan Mahesh

Publication date 9 December 2020

Published in Journal of Computational Physics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1805.05821

zbMATH Keywords

error analysislarge dataparalleldynamic mode decompositionfull-orthogonalization-Arnoldi

Mathematics Subject Classification ID

Parallel numerical computation (65Y05) Complexity and performance of numerical algorithms (65Y20) Numerical computation of eigenvalues and eigenvectors of matrices (65F15) Basic methods in fluid mechanics (76M99)

Abstract: A novel technique based on the Full Orthogonalization Arnoldi (FOA) is proposed to perform Dynamic Mode Decomposition (DMD) for a sequence of snapshots. A modification to FOA is presented for situations where the matrix $A$ is unknown, but the set of vectors ${A^{i-1}{v}_1}_{i=1}^{N-1}$ are known. The modified FOA is the kernel for the proposed projected DMD algorithm termed, FOA based DMD. The proposed algorithm to compute DMD modes and eigenvalues i) does not require Singular Value Decomposition (SVD) for snapshot matrices $X$ with $kapp{a}_2(X)ll1/epsilo{n}_m$, where $kapp{a}_2(X)$ is the 2-norm condition number of the snapshot matrix and $epsilo{n}_m$ is the relative round-off error or machine epsilon, ii) has an optional rank truncation step motivated by round off error analysis for snapshot matrices $X$ with $kapp{a}_2(X)approx1/epsilo{n}_m$, iii) requires only one snapshot at a time, thus making it a 'streaming' method even with the optional rank truncation step, iv) consumes less memory and requires less floating point operations to obtain the projected matrix than existing projected DMD methods and v) lends itself to easy parallelism as the main computational kernel involves only vector additions, dot products and matrix vector products. The new technique is therefore well-suited for DMD of large datasets on parallel computing platforms. We show both theoretically and using numerical examples that for FOA based DMD without rank truncation, the finite precision error in the computed projection of the linear mapping is $O(epsilo{n}_{m}kapp{a}_2(X))$. The proposed method is also compared to existing projected DMD methods for computational cost, memory consumption and relative round off error. Error indicators are presented that are useful to decide when to stop acquiring new snapshots. The proposed method is applied to several examples of numerical simulations of fluid flow.

Recommendations

• Randomized Dynamic Mode Decomposition
• Tensor-based dynamic mode decomposition
• Dynamic mode decomposition -- a numerical linear algebra perspective
• Variants of dynamic mode decomposition: boundary condition, Koopman, and Fourier analyses
• Consistent Dynamic Mode Decomposition

Cites work

• scientific article; zbMATH DE number 6678650 (Why is no real title available?)
• scientific article; zbMATH DE number 1012640 (Why is no real title available?)
• A data-driven approximation of the koopman operator: extending dynamic mode decomposition
• A numerical study of shear layer characteristics of low-speed transverse jets
• Accuracy and Stability of Numerical Algorithms
• Analysis of Fluid Flows via Spectral Properties of the Koopman Operator
• Communication-optimal parallel and sequential QR and LU factorizations
• Dynamic mode decomposition of numerical and experimental data
• Experimental investigation of nonlinear instabilities in annular liquid sheets
• Loss and Recapture of Orthogonality in the Modified Gram–Schmidt Algorithm
• Model Reduction for Flow Analysis and Control
• Numerical methods for large eigenvalue problems
• On dynamic mode decomposition: theory and applications
• Optimal mode decomposition for unsteady flows
• Recursive dynamic mode decomposition of transient and post-transient wake flows
• Rounding error analysis of the classical Gram-Schmidt orthogonalization process
• Spectral analysis of nonlinear flows
• Variants of dynamic mode decomposition: boundary condition, Koopman, and Fourier analyses

Cited in

(7)

• Effect of tabs on the shear layer dynamics of a jet in cross-flow
• Airfoil-based convolutional autoencoder and long short-term memory neural network for predicting coherent structures evolution around an airfoil
• Global stability analysis and direct numerical simulation of boundary layers with an isolated roughness element
• Data‐driven identification of the spatiotemporal structure of turbulent flows by streaming dynamic mode decomposition
• Numerical study of cavitation regimes in flow over a circular cylinder
• Higher-order dynamic mode decomposition on-the-fly: a low-order algorithm for complex fluid flows
• Tracking Distributed Parameters System Dynamics with Recursive Dynamic Mode Decomposition with Control

Describes a project that uses

Uses Software

• mctoolbox

This page was built for publication: A parallel and streaming dynamic mode decomposition algorithm with finite precision error analysis for large data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2214545)