Numerical assessments of a nonintrusive surrogate model based on recurrent neural networks and proper orthogonal decomposition: Rayleigh Benard convection

TL;DR

This paper introduces a nonintrusive surrogate modeling framework combining proper orthogonal decomposition, autoencoders, and LSTM neural networks to efficiently simulate Rayleigh-Benard convection.

Contribution

It develops a novel NLPOD approach that integrates linear and nonlinear dimensionality reduction with temporal neural networks for reduced-order modeling.

Findings

The NLPOD model accurately captures convection dynamics.

Hyperparameter sensitivity analysis improves model efficiency.

The approach outperforms traditional linear models.

Abstract

Recent developments in diagnostic and computing technologies offer to leverage numerous forms of nonintrusive modeling approaches from data where machine learning can be used to build computationally cheap and accurate surrogate models. To this end, we present a nonlinear proper orthogonal decomposition (POD) framework, denoted as NLPOD, to forge a nonintrusive reduced-order model for the Boussinesq equations. In our NLPOD approach, we first employ the POD procedure to obtain a set of global modes to build a linear-fit latent space and utilize an autoencoder network to compress the projection of this latent space through a nonlinear unsupervised mapping of POD coefficients. Then, long short-term memory (LSTM) neural network architecture is utilized to discover temporal patterns in this low-rank manifold. While performing a detailed sensitivity analysis for hyperparameters of the LSTM model, the trade-off between accuracy and efficiency is systematically analyzed for solving a canonical Rayleigh-Benard convection system.