Model Reduction with Memory and the Machine Learning of Dynamical Systems

Year:    2019

Communications in Computational Physics, Vol. 25 (2019), Iss. 4 : pp. 947–962

Abstract

The well-known Mori-Zwanzig theory tells us that model reduction leads to memory effect. For a long time, modeling the memory effect accurately and efficiently has been an important but nearly impossible task in developing a good reduced model. In this work, we explore a natural analogy between recurrent neural networks and the Mori-Zwanzig formalism to establish a systematic approach for developing reduced models with memory. Two training models — a direct training model and a dynamically coupled training model — are proposed and compared. We apply these methods to the Kuramoto-Sivashinsky equation and the Navier-Stokes equation. Numerical experiments show that the proposed method can produce reduced model with good performance on both short-term prediction and long-term statistical properties.

Journal Article Details

Publisher Name:    Global Science Press

Language:    English

DOI:    https://doi.org/10.4208/cicp.OA-2018-0269

Communications in Computational Physics, Vol. 25 (2019), Iss. 4 : pp. 947–962

Published online:    2019-01

AMS Subject Headings:    Global Science Press

Copyright:    COPYRIGHT: © Global Science Press

Pages:    16

Keywords:    Model reduction Mori-Zwanzig recurrent neural networks.