PI-VEGAN: Physics Informed Variational Embedding Generative Adversarial Networks for Stochastic Differential Equations

Ruisong Gao,Yufeng Wang,Min Yang,Chuanjun Chen

2023-07-21

Abstract:We present a new category of physics-informed neural networks called physics informed variational embedding generative adversarial network (PI-VEGAN), that effectively tackles the forward, inverse, and mixed problems of stochastic differential equations. In these scenarios, the governing equations are known, but only a limited number of sensor measurements of the system parameters are available. We integrate the governing physical laws into PI-VEGAN with automatic differentiation, while introducing a variational encoder for approximating the latent variables of the actual distribution of the measurements. These latent variables are integrated into the generator to facilitate accurate learning of the characteristics of the stochastic partial equations. Our model consists of three components, namely the encoder, generator, and discriminator, each of which is updated alternatively employing the stochastic gradient descent algorithm. We evaluate the effectiveness of PI-VEGAN in addressing forward, inverse, and mixed problems that require the concurrent calculation of system parameters and solutions. Numerical results demonstrate that the proposed method achieves satisfactory stability and accuracy in comparison with the previous physics-informed generative adversarial network (PI-WGAN).

Machine Learning

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve This paper aims to address the forward, backward, and mixed problems in Stochastic Differential Equations (SDEs). Specifically: 1. **Problem Background**: - Stochastic Differential Equations are widely used in fields such as finance, physics, and engineering, typically involving random fluctuations in systems. - Solving these equations requires handling variations in space and time as well as the randomness of the solutions. 2. **Limitations of Existing Methods**: - Traditional numerical methods (such as Monte Carlo methods, spectral methods, and stochastic Galerkin methods) are inefficient in high-dimensional stochastic partial differential equations and suffer from the "curse of dimensionality." - Physics-Informed Neural Networks (PINNs) incorporate physical knowledge as soft constraints into the empirical loss function, combining automatic differentiation and stochastic optimization techniques to train models, and have shown excellent performance in solving fundamental partial differential equations. 3. **Proposed Method in the Paper**: - A new method called Physics-Informed Variational Embedding Generative Adversarial Network (PI-VEGAN) is proposed to solve the forward, backward, and mixed problems of stochastic differential equations. - By introducing a variational encoder to approximate the latent variables of the actual measurement distribution, the learning stability and accuracy of the generator are improved. - The model consists of three parts: encoder, generator, and discriminator, and these components are trained by alternately updating them. 4. **Objectives**: - Accurately learn the characteristics of stochastic partial differential equations with limited sensor measurement data. - Improve the stability and accuracy of the model in solving forward, backward, and mixed problems, especially in the case of high-dimensional stochastic partial differential equations. Through the above methods, this paper aims to overcome the limitations of traditional methods in high-dimensional stochastic partial differential equations and provide an efficient and stable solution.

PI-VEGAN: Physics Informed Variational Embedding Generative Adversarial Networks for Stochastic Differential Equations

PI-VAE: Physics-Informed Variational Auto-Encoder for stochastic differential equations

Physics-Informed Generative Adversarial Networks for Stochastic Differential Equations

Physics-Informed Generator-Encoder Adversarial Networks with Latent Space Matching for Stochastic Differential Equations

Data-driven forward-inverse problems of the 2-coupled mixed derivative nonlinear Schrödinger equation using deep learning

Physics-Informed Generative Adversarial Network Solution to Buckley–Leverett Equation

Variational Physics-Informed Neural Networks For Solving Partial Differential Equations

Cv-Pinn: Efficient Learning of Variational Physics-Informed Neural Network with Domain Decomposition

VI-PINNs: Variance-involved Physics-informed Neural Networks for Fast and Accurate Prediction of Partial Differential Equations

Physics-guided generative adversarial network to learn physical models

Physics-integrated generative modeling using attentive planar normalizing flow based variational autoencoder

PIG: Physics-Informed Gaussians as Adaptive Parametric Mesh Representations

Learning Physics-Informed Neural Networks without Stacked Back-propagation

VC-PINN: Variable coefficient physics-informed neural network for forward and inverse problems of PDEs with variable coefficient

Physics-informed variational inference for uncertainty quantification of stochastic differential equations

Variational integrator graph networks for learning energy-conserving dynamical systems

Combining physics-informed graph neural network and finite difference for solving forward and inverse spatiotemporal PDEs

VC-PINN: Variable Coefficient Physical Information Neural Network For Forward And Inverse PDE Problems with Variable Coefficient

Recipes for when physics fails: recovering robust learning of physics informed neural networks

SVD-PINNs: Transfer Learning of Physics-Informed Neural Networks via Singular Value Decomposition