Multi-frequency wavefield solutions for variable velocity models using meta-learning enhanced low-rank physics-informed neural network

TL;DR

This paper introduces Meta-LRPINN, a novel physics-informed neural network framework that combines low-rank SVD parameterization, meta-learning, and frequency embedding to efficiently model multi-frequency wavefields in complex velocity models, achieving faster convergence and better generalization.

Contribution

The paper presents a new framework that integrates low-rank SVD, meta-learning, and frequency embedding to improve modeling of multi-frequency wavefields with enhanced efficiency and generalization capabilities.

Findings

Meta-LRPINN converges faster than baseline methods.

It achieves higher accuracy in wavefield modeling.

The framework generalizes well to out-of-distribution frequencies.

Abstract

Physics-informed neural networks (PINNs) face significant challenges in modeling multi-frequency wavefields in complex velocity models due to their slow convergence, difficulty in representing high-frequency details, and lack of generalization to varying frequencies and velocity scenarios. To address these issues, we propose Meta-LRPINN, a novel framework that combines low-rank parameterization using singular value decomposition (SVD) with meta-learning and frequency embedding. Specifically, we decompose the weights of PINN's hidden layers using SVD and introduce an innovative frequency embedding hypernetwork (FEH) that links input frequencies with the singular values, enabling efficient and frequency-adaptive wavefield representation. Meta-learning is employed to provide robust initialization, improving optimization stability and reducing training time. Additionally, we implement adaptive rank reduction and FEH pruning during the meta-testing phase to further enhance efficiency. Numerical experiments, which are presented on multi-frequency scattered wavefields for different velocity models, demonstrate that Meta-LRPINN achieves much fast convergence speed and much high accuracy compared to baseline methods such as Meta-PINN and vanilla PINN. Also, the proposed framework shows strong generalization to out-of-distribution frequencies while maintaining computational efficiency. These results highlight the potential of our Meta-LRPINN for scalable and adaptable seismic wavefield modeling.