Non-Intrusive Reduced Models based on Operator Inference for Chaotic Systems

TL;DR

This paper presents a physics-driven machine learning approach called Operator Inference (OpInf) for non-intrusive reduced modeling of chaotic systems, demonstrating superior forecasting performance over existing neural network methods.

Contribution

It introduces a novel non-intrusive reduced order modeling technique using OpInf for chaotic systems, combining PCA and polynomial operator inference for accurate predictions.

Findings

OpInf models outperform state-of-the-art neural networks in forecasting chaotic systems.

The method achieves high Valid Prediction Time (VPT) in Lorenz 96 and Kuramoto-Sivashinsky equations.

Numerical results show promising accuracy and efficiency of the proposed approach.

Abstract

This work explores the physics-driven machine learning technique Operator Inference (OpInf) for predicting the state of chaotic dynamical systems. OpInf provides a non-intrusive approach to infer approximations of polynomial operators in reduced space without having access to the full order operators appearing in discretized models. Datasets for the physics systems are generated using conventional numerical solvers and then projected to a low-dimensional space via Principal Component Analysis (PCA). In latent space, a least-squares problem is set to fit a quadratic polynomial operator, which is subsequently employed in a time-integration scheme in order to produce extrapolations in the same space. Once solved, the inverse PCA operation is applied to reconstruct the extrapolations in the original space. The quality of the OpInf predictions is assessed via the Normalized Root Mean Squared Error (NRMSE) metric from which the Valid Prediction Time (VPT) is computed. Numerical experiments considering the chaotic systems Lorenz 96 and the Kuramoto-Sivashinsky equation show promising forecasting capabilities of the OpInf reduced order models with VPT ranges that outperform state-of-the-art machine learning methods such as backpropagation and reservoir computing recurrent neural networks [1], as well as Markov neural operators [2].