A High-Efficient Hybrid Physics-Informed Neural Networks Based on Convolutional Neural Network

• Published in: IEEE transaction on neural networks and learning systems Vol. 33; no. 10; pp. 5514 - 5526
• Main Author: Fang, Zhiwei
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.10.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Approximation; Artificial neural networks; Convergence; Convolutional neural network (CNN); Differential equations; Finite volume method; finite-difference method; Geometry; Handheld computers; hybrid physics-informed neural network (hybrid PINN); Inverse problems; Learning systems; local fitting method; Machine learning; Mathematical analysis; Mathematical models; Neural networks; Numerical methods; Operators (mathematics); Partial differential equations; Physics; Training

In this article, we develop a hybrid physics-informed neural network (hybrid PINN) for partial differential equations (PDEs). We borrow the idea from the convolutional neural network (CNN) and finite volume methods. Unlike the physics-informed neural network (PINN) and its variations, the method pro...