Variational operator learning: A unified paradigm marrying training neural operators and solving partial differential equations
arXiv:2304.04234 · doi:10.1016/j.jmps.2024.105714
Abstract
Neural operators as novel neural architectures for fast approximating solution operators of partial differential equations (PDEs), have shown considerable promise for future scientific computing. However, the mainstream of training neural operators is still data-driven, which needs an expensive ground-truth dataset from various sources (e.g., solving PDEs' samples with the conventional solvers, real-world experiments) in addition to training stage costs. From a computational perspective, marrying operator learning and specific domain knowledge to solve PDEs is an essential step in reducing dataset costs and label-free learning. We propose a novel paradigm that provides a unified framework of training neural operators and solving PDEs with the variational form, which we refer to as the variational operator learning (VOL). Ritz and Galerkin approach with finite element discretization are developed for VOL to achieve matrix-free approximation of system functional and residual, then direct minimization and iterative update are proposed as two optimization strategies for VOL. Various types of experiments based on reasonable benchmarks about variable heat source, Darcy flow, and variable stiffness elasticity are conducted to demonstrate the effectiveness of VOL. With a label-free training set and a 5-label-only shift set, VOL learns solution operators with its test errors decreasing in a power law with respect to the amount of unlabeled data. To the best of the authors' knowledge, this is the first study that integrates the perspectives of the weak form and efficient iterative methods for solving sparse linear systems into the end-to-end operator learning task.
This version mainly improves the quality of the bitmaps in the results compared to the previous version
References in corpus (27)
- DeepONet: Learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operators
- DGM: A deep learning algorithm for solving partial differential equations
- An Energy Approach to the Solution of Partial Differential Equations in Computational Mechanics via Machine Learning: Concepts, Implementation and Applications
- Scaling Laws for Neural Language Models
- Fourier Neural Operator for Parametric Partial Differential Equations
- Surrogate Modeling for Fluid Flows Based on Physics-Constrained Deep Learning Without Simulation Data
- Integrating Machine Learning and Multiscale Modeling: Perspectives, Challenges, and Opportunities in the Biological, Biomedical, and Behavioral Sciences
- PhyGeoNet: Physics-Informed Geometry-Adaptive Convolutional Neural Networks for Solving Parameterized Steady-State PDEs on Irregular Domain
- A physics-informed variational DeepONet for predicting the crack path in brittle materials
- Stress Field Prediction in Cantilevered Structures Using Convolutional Neural Networks
- Physics-Informed Neural Operator for Learning Partial Differential Equations
- Neural Operator: Graph Kernel Network for Partial Differential Equations
- StressGAN: A Generative Deep Learning Model for 2D Stress Distribution Prediction
- Deep transfer operator learning for partial differential equations under conditional shift
- MgNet: A Unified Framework of Multigrid and Convolutional Neural Network
- FEA-Net: A Physics-guided Data-driven Model for Efficient Mechanical Response Prediction
- Accelerated high-cycle phase field fatigue predictions
- Choose a Transformer: Fourier or Galerkin
- Learning Operators with Coupled Attention
- Transformer for Partial Differential Equations' Operator Learning
- A Helmholtz equation solver using unsupervised learning: Application to transcranial ultrasound
- GNOT: A General Neural Operator Transformer for Operator Learning
- Geometry-Informed Neural Operator for Large-Scale 3D PDEs
- Meta-Auto-Decoder for Solving Parametric Partial Differential Equations
- A Solvable Model of Neural Scaling Laws
- Operator learning with PCA-Net: upper and lower complexity bounds
- A Deep Conjugate Direction Method for Iteratively Solving Linear Systems