Long-time prediction of nonlinear parametrized dynamical systems by deep learning-based ROMs

Duration: 31 mins 55 secs

Share this media item:

Embed this media item:

<iframe width="_width_" height="_height_" src="https://sms.cam.ac.uk/media/3708942/embed" frameborder="0" scrolling="no" allowfullscreen></iframe>

Choose size:

About this item

Available Formats

About this item

Description:	Professor Andrea Manzoni (Politecnico di Milano) 19th November 2021 \| 14:30 - 15:00


Created:	2021-11-22 14:36
Collection:	Modelling Behaviour to Inform Policy for Pandemics
Publisher:	Isaac Newton Institute for Mathematical Sciences
Copyright:	Professor Andrea Manzoni
Language:	eng (English)


Abstract:	Deep learning-based reduced order models (DL-ROMs) have been recently proposed to overcome common limitations shared by conventional ROMs - built, e.g., through proper orthogonal decomposition (POD) - when applied to nonlinear time-dependent parametrized PDEs. Although extremely efficient at testing time, when evaluating the PDE solution for any new testing-parameter instance, DL-ROMs require an expensive training stage. To avoid this latter, a prior dimensionality reduction through POD, and a multi-fidelity pretraining stage, are introduced, yielding the POD-DL-ROM framework, which allows to solve time-dependent PDEs even faster than in real-time. A further step has led us to introduce LSTM networks instead than convolutional autoencoders, ultimately obtaining POD-LSTM-ROMs that better grasp the time evolution of the PDE system, to enhance predictions for new testing-parameter instances. Equipped with LSTMs, POD-LSTM-ROMs also allow us to perform extrapolation of the PDE solution forward in time, that is, on a (much) larger time domain than the one used to train the ROM, for unseen values of the input parameters - a task often missed by traditional projection-based ROMs. To this aim, we train a POD-LSTM-ROM on snapshots acquired on a given time interval, and then approximate the solution on a much longer time window, taking advantage of a LSTM architecture in a different form besides the POD-LSTM-ROM introduced before. Building predictions for future times based on the past, these coupled architectures mimic the behavior of common numerical solvers for dynamical systems. We assess the performance of the proposed framework on several examples, ranging from low-dimensional, nonperiodic systems to applications in structural mechanics dealing with micro electromechanical systems, obtaining faster than real-time simulations that are able to preserve a remarkable accuracy across the entire time domain considered during training.

Abstract:

Deep learning-based reduced order models (DL-ROMs) have been recently proposed to overcome common limitations shared by conventional ROMs - built, e.g., through proper orthogonal decomposition (POD) - when applied to nonlinear time-dependent parametrized PDEs. Although extremely efficient at testing time, when evaluating the PDE solution for any new testing-parameter instance, DL-ROMs require an expensive training stage. To avoid this latter, a prior dimensionality reduction through POD, and a multi-fidelity pretraining stage, are introduced, yielding the POD-DL-ROM framework, which allows to solve time-dependent PDEs even faster than in real-time. A further step has led us to introduce LSTM networks instead than convolutional autoencoders, ultimately obtaining POD-LSTM-ROMs that better grasp the time evolution of the PDE system, to enhance predictions for new testing-parameter instances.

Equipped with LSTMs, POD-LSTM-ROMs also allow us to perform extrapolation of the PDE solution forward in time, that is, on a (much) larger time domain than the one used to train the ROM, for unseen values of the input parameters - a task often missed by traditional projection-based ROMs. To this aim, we train a POD-LSTM-ROM on snapshots acquired on a given time interval, and then approximate the solution on a much longer time window, taking advantage of a LSTM architecture in a different form besides the POD-LSTM-ROM introduced before. Building predictions for future times based on the past, these coupled architectures mimic the behavior of common numerical solvers for dynamical systems. We assess the performance of the proposed framework on several examples, ranging from low-dimensional, nonperiodic systems to applications in structural mechanics dealing with micro electromechanical systems, obtaining faster than real-time simulations that are able to preserve a remarkable accuracy across the entire time domain considered during training.

Available Formats

Format	Quality	Bitrate	Size
MPEG-4 Video	1280x720	2.91 Mbits/sec	696.68 MB	View	Download
MPEG-4 Video	640x360	1.36 Mbits/sec	327.42 MB	View	Download
WebM	1280x720	2.35 Mbits/sec	564.17 MB	View	Download
WebM	640x360	560.21 kbits/sec	131.03 MB	View	Download
iPod Video	480x270	479.19 kbits/sec	112.02 MB	View	Download
MP3	44100 Hz	249.79 kbits/sec	58.45 MB	Listen	Download
Auto *	(Allows browser to choose a format it supports)