Waveformer for modelling dynamical systems
This addresses a bottleneck in neural operators for time-dependent problems, offering improved accuracy for applications like fluid dynamics and PDE modeling.
The authors tackled the problem of accurate long-term predictions for time-dependent dynamical systems by proposing Waveformer, a novel operator learning approach that combines wavelet transforms and transformers. Results show Waveformer outperforms existing state-of-the-art operator learning algorithms by up to an order of magnitude in accuracy, particularly in extrapolation tasks.
Neural operators have gained recognition as potent tools for learning solutions of a family of partial differential equations. The state-of-the-art neural operators excel at approximating the functional relationship between input functions and the solution space, potentially reducing computational costs and enabling real-time applications. However, they often fall short when tackling time-dependent problems, particularly in delivering accurate long-term predictions. In this work, we propose "waveformer", a novel operator learning approach for learning solutions of dynamical systems. The proposed waveformer exploits wavelet transform to capture the spatial multi-scale behavior of the solution field and transformers for capturing the long horizon dynamics. We present four numerical examples involving Burgers's equation, KS-equation, Allen Cahn equation, and Navier Stokes equation to illustrate the efficacy of the proposed approach. Results obtained indicate the capability of the proposed waveformer in learning the solution operator and show that the proposed Waveformer can learn the solution operator with high accuracy, outperforming existing state-of-the-art operator learning algorithms by up to an order, with its advantage particularly visible in the extrapolation region