mLaSDI: Multi-stage latent space dynamics identification
By: William Anderson, Seung Whan Chung, Youngsoo Choi
Potential Business Impact:
Makes computer models of science faster and better.
Determining accurate numerical solutions of partial differential equations (PDEs) is an important task in many scientific disciplines. However, solvers can be computationally expensive, leading to the development of reduced-order models (ROMs). Recently, Latent Space Dynamics Identification (LaSDI) was proposed as a data-driven, non-intrusive ROM framework. LaSDI compresses the training data using an autoencoder and learns a system of user-chosen ordinary differential equations (ODEs), which govern the latent space dynamics. This allows for rapid predictions by interpolating and evolving the low-dimensional ODEs in the latent space. While LaSDI has produced effective ROMs for numerous problems, the autoencoder can have difficulty accurately reconstructing training data while also satisfying the imposed dynamics in the latent space, particularly in complex or high-frequency regimes. To address this, we propose multi-stage Latent Space Dynamics Identification (mLaSDI). With mLaSDI, several autoencoders are trained sequentially in stages, where each autoencoder learns to correct the error of the previous stages. We find that applying mLaSDI with small autoencoders results in lower prediction and reconstruction errors, while also reducing training time compared to LaSDI.
Similar Papers
Sequential decoder training for improved latent space dynamics identification
Machine Learning (CS)
Makes computer simulations faster and more accurate.
Thermodynamically Consistent Latent Dynamics Identification for Parametric Systems
Machine Learning (CS)
Makes computer models run much faster and smarter.
Rollout-LaSDI: Enhancing the long-term accuracy of Latent Space Dynamics
Machine Learning (CS)
Makes computer predictions of science problems last longer.