An Efficient Wavelet-based Physics Informed Residual Neural Networks for Flow Field Reconstruction with Extremely Sparse Data

TL;DR

This paper presents wavelet-physics-informed residual neural networks (W-PIRNNs) that effectively reconstruct complex flow fields from extremely sparse data, integrating wavelet activations and residual connections to overcome traditional PINN limitations.

Contribution

The introduction of W-PIRNNs combining wavelet activations with residual networks enables accurate flow reconstruction from minimal data, addressing both forward and inverse PDE problems.

Findings

Reconstructed flow fields from only 0.05% velocity data.

Achieved high accuracy in solving PDEs like Burger's and Schrödinger equations.

Effectively addressed both forward and inverse flow problems.

Abstract

This paper introduces wavelet-physics-informed residual neural networks (W-PIRNNs) to study complex fluid flow problems by reconstructing the flow field from highly sparse, supervised data. Our W-PIRNNs fundamentally integrate ResNet and employ the wavelet $W(t) = w_1 \sin(t) + w_2 \cos(t)$ as an activation function. Due to the vanishing and ballooning gradient problems associated with typical PINNs' deep networks, we implemented residual-based skip connections. Our W-PIRNNs, which integrate supervised data with physical principles, demonstrate efficacy even in scenarios of sparse or partial data, enabling the reconstruction of flow fields using merely $0.05\%$ velocity data for training. The wake flow around a circular cylinder served as the test case for our proposed technique, which depends exclusively on velocity data for training. This technique facilitates the precise reconstruction of velocity, pressure, streamlines, and vorticity, requiring fewer epochs and less processing time. Significantly, our proposed W-PIRNNs effectively resolve PDEs in both forward and inverse contexts. Burger's equation served as a test case for both the forward and inverse problem configurations. Our network calculates the diffusion or viscosity coefficient ($\lambda_2$) with an absolute error of $0.065\%$ and the convection coefficient ($\lambda_1$) with an absolute error of $0.002\%$. Furthermore, the Schr\"odinger equation is examined in the forward setting to assess the framework's ability to handle periodic boundary conditions. To the best of our knowledge, W-PIRNNs represent the first method capable of flow reconstruction using highly sparse supervised data, as well as reconstructing streamline and vorticity, and they effectively address both forward and inverse problems with high accuracy.