Characterizing the Predictive Accuracy of Dynamic Mode Decomposition for Data-Driven Control

Dynamic mode decomposition (DMD) is a versatile approach that enables the construction of low-order models from data. Controller design tasks based on such models require estimates and guarantees on predictive accuracy. In this work, we provide a theoretical analysis of DMD model errors that reveals impact of model order and data availability. The analysis also establishes conditions under which DMD models can be made asymptotically exact. We verify our results using a 2D diffusion system.

[1]  I. Mezić,et al.  Analysis of Fluid Flows via Spectral Properties of the Koopman Operator , 2013 .

[2]  Damon Honnery,et al.  An error analysis of the dynamic mode decomposition , 2011, Experiments in Fluids.

[3]  Daniel M. Tartakovsky,et al.  Predictive Accuracy of Dynamic Mode Decomposition , 2019, 1905.01587.

[4]  Jennifer Annoni,et al.  Wind farm flow modeling using an input-output reduced-order model , 2016, 2016 American Control Conference (ACC).

[5]  J. Nathan Kutz,et al.  Dynamic mode decomposition for financial trading strategies , 2015, 1508.04487.

[6]  Clarence W. Rowley,et al.  Evaluating the accuracy of the dynamic mode decomposition , 2016, Journal of Computational Dynamics.

[7]  Steven L. Brunton,et al.  On dynamic mode decomposition: Theory and applications , 2013, 1312.0041.

[8]  Igor Mezic,et al.  Linear predictors for nonlinear dynamical systems: Koopman operator meets model predictive control , 2016, Autom..

[9]  Stefan Volkwein,et al.  POD a-posteriori error estimates for linear-quadratic optimal control problems , 2009, Comput. Optim. Appl..

[10]  Bernd R. Noack,et al.  Model reduction using Dynamic Mode Decomposition , 2014 .

[11]  P. Holmes,et al.  The Proper Orthogonal Decomposition in the Analysis of Turbulent Flows , 1993 .

[12]  Clarence W. Rowley,et al.  Online dynamic mode decomposition for time-varying systems , 2017, SIAM J. Appl. Dyn. Syst..

[13]  P. Schmid,et al.  Dynamic mode decomposition of numerical and experimental data , 2008, Journal of Fluid Mechanics.

[14]  Igor Mezic,et al.  On Convergence of Extended Dynamic Mode Decomposition to the Koopman Operator , 2017, J. Nonlinear Sci..

[15]  Steven L. Brunton,et al.  Dynamic Mode Decomposition with Control , 2014, SIAM J. Appl. Dyn. Syst..

[16]  B. Moore.  Principal component analysis in linear systems: Controllability, observability, and model reduction , 1981 .

[17]  El-hadi Zahzah,et al.  Dynamic Mode Decomposition for Robust PCA with Applications to Foreground/Background Subtraction in Video Streams and Multi-Resolution Analysis , 2016 .